ISFDB:Community Portal/Archive/Archive19


Jump to: navigation, search

This is an archive page for the Community Portal. Please do not edit the contents. To start a new discussion, please click here.
This archive includes discussions from

Archive Quick Links
Archives of old discussions from the Community Portal.

1 · 2 · 3 · 4 · 5 · 6 · 7 · 8 · 9 · 10 · 11 · 12 · 13 · 14 · 15 · 16 · 17 · 18 · 19 · 20 · 21 · 22 · 23 · 24 · 25 · 26 · 27 · 28 · 29 · 30 · 31 · 32 · 33 · 34 · 35 · 36 · 37 · 38 · 39 · 40 · 41 · 42 · 43 · 44 · 45 · 46 · 47 · 48 · 49 · 50 · 51

Archive of the Community Portal - March-May 2010


Design discussion of Artist / Coverart handling in Pub Screen

I have started a discussion about the the Pub Screen's handling of the manipulation of artist credits and coverart information at ISFDB:Proposed_Design_Changes#Artist_and_Coverart_handling_in_Pub_Screen. Contributions welcome! --MartyD 13:07, 28 February 2010 (UTC)

Highlight Verified Magazines in Grid

I would like to request a new feature that would Highlight Magazines with Primary Verification in the Grid. This would allow editors to identify all unverified magazines at a single glance, rather than stepping through the editor records year by year. Maybe light green background for "verified," yellow for "unverified," and red for "no data."--Rkihara 19:02, 13 March 2010 (UTC)

At this time magazine grids are maintained in the Wiki, so any changes would have to be done by editors using Wiki codes. For example, I have changed the color coding of a single row for Magazine:Orbit to show what it may look like. I don't know how to get the same effect for individual cells, but I have seen it done within Wikipedia, so I am sure we could identify the right codes after a little research.
A longer term solution would be to create an "issue grid" view for magazine series within the core ISFDB application. It wouldn't be hard to do for well behaved magazines with dates like "1953-10-00", "1953-12-00", "1954-06-00", and so on. It would be harder to do for quarterly magazines which often do not have month information in the date field or use the "1931-00-01", "1931-00-2", etc convention. Still, its probably worth going after the "low hanging fruit" and see how many magazines it handles. Ahasuerus 19:47, 13 March 2010 (UTC)
I followed Ahasuerus's example for Weird Tales to see what a larger sample looks like. It is certainly possible to highlight individual cells. I used the same color for the "no data" cells, though we may want to go with Rkihara's suggestion of a different color. I could go either way on that. This is still an experiment, so feel free to alter or revert my edits to that page depending on how we decide to proceed. --Ron ~ RtraceTalk 21:33, 13 March 2010 (UTC)
That looks great! Looking at your example, I wonder if it would be easier when most of the entries are verified to set the whole grid to green and flip the backgrounds individually on the unverified issues?--Rkihara 22:43, 13 March 2010 (UTC)
I found it's easier to highlight the unverified magazines instead of the verified magazines, as there are more of the former than the latter. For example, Thrilling Wonder Stories vs Startling Stories. The use of green highlighting seemed a little understated in the latter case, but gold seems to be a little hard on the eye. If no one objects, I'm thinking of highlighting the grids that I'm presently working on (Amazing, Astounding/Analog, etc.).--Rkihara 17:40, 16 March 2010 (UTC)
Perhaps this will be of temporary use. Based upon there being a primary verifier. Current as of 4/13/2010.--swfritter 22:39, 16 March 2010 (UTC)
There are also about 50 issues from the Campbell era Astounding/Analog that have not been verified. It looks like most of these have had a great deal of work done on them.--swfritter 22:51, 16 March 2010 (UTC)
Those lists will speed things up considerably. I finished highlighting Amazing this afternoon before seeing your posting, tediously stepping through the editor records year-by-year. I'll do Analog next. I've noticed the records you mentioned before and wondered if the editor was going to come back and verify them.--Rkihara 02:55, 17 March 2010 (UTC)
The entries all seem to be very detailed. I have all the US issues, as you probably do, so it should be a quick job to verify them if the editor who entered the data doesn't claim them shortly; there should be no need to change the entries in the wiki.--swfritter 03:40, 17 March 2010 (UTC)
I've completed the markup of the Analog Grid, and it looks like we could finish off the grid in a couple of days, if no one claims the unverified issues. Should we put up a notice of intent on the Community Portal? One editor to check with might be "Tpi" who has been slowly filling in the blanks, but maybe not always verifying.--Rkihara 18:06, 17 March 2010 (UTC)
I figured we might get a response from this discussion; but it might be a good idea to post a separate notice. Analog issues have been verified by many hands so it might be hard to figure out who did these. Maybe if you leave a notice up over the weekend.--swfritter 20:50, 17 March 2010 (UTC)

Anyone Attending the Windy City Pulp and Paper Convention?

I'll be attending the Windy City Pulp and Paper Convention this year and I was wondering if anyone else in this group will be there, and would be interested in meeting up?--Rkihara 19:07, 13 March 2010 (UTC)

Finishing the Verification of the Analog SF Magazines

I recently highlighted the unverified magazines in the Analog grid for identification, and it's apparent that the majority have been verified. Data entry appears complete in most of the remaining issues, but the editors left these unverified. Swfritter and I would like to complete the Primary Verification of the magazines to the present, but we don't wish to interfere with editors who may still be working on the unverified issues. Please let us know your thoughts on this.--Rkihara 04:31, 19 March 2010 (UTC)

Fair question. What about this. Anyone who is working on Analog issues please contact the above and help complete the series? As for reality. Some editors apparently check but do not verify. Sometimes you complete something, but someone else checks and verifies. In neither case is there a foul. In this case, you guys, have the material and can complete the comprehensive check of them. Go for it, and TG you have the material, because until it is done we have to await someone who has it. So, no problem that I see. Thanks, Harry. --Dragoondelight 12:58, 19 March 2010 (UTC)
I've entered complete contents for magazine issues (not Analog) and forgot to verify them, so I can see how this sort of thing happens. (I come across several records a week that I've entered without verifying the first time.) Feel free to go ahead and check the data on these Analogs, then do a primary verification. If the editor who did the original data entry comes back, he can always do a 2nd or 3rd verification. No personal property around here. (It may have been absent moderator and editor Bob Hall (Hall3730) who's not responded to his talk page in more than a year.) Mhhutchins 14:23, 19 March 2010 (UTC)
Magazines I forgot to verify are also coming to my attention. The Primary Verifier is usually the first person to get a query about a magazine and is most likely to be responsible for entries that may require explanation; notifying the FBI that we need track down the perps is probably not necessary so I would think Ron and I can work on these next week. If this was the year 2012 I would probably do them immediately just to make sure they get done before the world comes to an end.--swfritter 14:38, 19 March 2010 (UTC)
I usually try to verify those issues I have checked. Often I'll go through one year (slowly, as there are quite many other things in my life which take time), and when it is ready I'll make all the verifications in one sweep. I just started on 1998. I have now about 97% of all issues between 1949-2010, and if anyone else doesn't do it, I will do all those I have - but admidtably fairly slowly. If/when some has more time they of course more than welcome to do the verifications... Tpi 16:41, 19 March 2010 (UTC)
The only reason it would be preferable for the issues involved is if you had done the expanded data entry. They appear to have been done in a very competent and complete manner which is certainly a good way to describe the way in which your work is done. I, and I assume Ron, will be glad to do them. They will be easy work.--swfritter 17:45, 19 March 2010 (UTC)
Is there any good reason not to make Tpi a moderator for such, if he's done that well? (Not really my area, I mostly restrict my magazine/fanzine work to things that nobody else will touch usually, and break a few rules in doing so.) BLongley 21:13, 19 March 2010 (UTC)
Like minded about Tpi. Shame on you Bill for breaking rules (Happy face emoticon if I were an emoticon kind of guy would go here). Although, I think rules is not the correct term. Perhaps Help should better be termed as Guidelines or maybe even Suggestions?--swfritter 21:36, 19 March 2010 (UTC)
1998 is now verified - it took less work than I thought. About moderator status: Well, I think would be able to moderate my own submissions, at least concerning magazines, sometimes it would make the work a bit easier. (not having to wait for moderation, before assigning something to a proper series would be nice) It is ok for me, but I also can go on like I have gone so far, and wait for the verifications. Tpi 15:42, 22 March 2010 (UTC)

Nominating editor Tpi for moderator

Tpi has worked methodically and efficiently through the more recent issues of Analog. Tpi's talk page is conspicuously absent any recent queries concerning his submissions. Tpi's editing time seems to be somewhat limited and being a moderator would substantially facilitate data entry. Tpi is also a very responsible editor so any lack of experience in other areas is likely to be preceded by careful study of the issues involved. Tpi is willing and I think meets the Moderator Requirements.--swfritter 14:18, 24 March 2010 (UTC)


  1. Support, as nominator.--swfritter 14:18, 24 March 2010 (UTC)
  2. Support, nomination says it all.--Rkihara 16:30, 24 March 2010 (UTC)
  3. Support, OK by me. BLongley 19:58, 24 March 2010 (UTC)
  4. Support, Good worker, sound judgment. Hope he enjoys! --Dragoondelight 22:18, 24 March 2010 (UTC)
  5. Support, with the hope that his experience with Analog will help broaden this magazine entry expertise into other areas of the database. Mhhutchins 03:58, 25 March 2010 (UTC)
  6. Support, contributions look good (and I'll no longer be the junior moderator). --Ron ~ RtraceTalk 11:38, 25 March 2010 (UTC)
  7. Support, Could as well be neutral, as I never clashed with Tpi, but I consider that to be a good sign, and it feels good to have more European moderators. --Willem 20:07, 25 March 2010 (UTC)
  8. Support, No complaints here. --MartyD 11:09, 26 March 2010 (UTC)
  9. Support. I never ran into any problems with Tpi's submissions when I worked with them. The only thing that I would recommend is being careful outside of your immediate areas of expertise since there are quite a few obscure catches and caveats in our data entry rules. Ahasuerus 14:21, 26 March 2010 (UTC)
  10. Support. Have virtually no contact with magazine entries, thus none with TPI, but I trust the judgment of the Mods who have. ~Bill. --Bluesman 03:25, 30 March 2010 (UTC)




Nomination passes; moderator flag has been set. Congratulations to Tpi! Ahasuerus 04:16, 30 March 2010 (UTC)

Thank you for trusting me. I'll try to do my best. Tpi 19:25, 30 March 2010 (UTC)

Automating repetitive tasks

In case some of us are not aware of its existence, there is a handy utility called MemoKeys which lets you program your function keys to enter arbitrary strings of text. For example, I have "F2-D" set to "Data from as of 2010-03-" so that I don't have to type the same string over and over again. The only caveat is that Web browsers typically use a number of function keys like F1, F3 and F5 for their own nefarious purposes, but F2 seems to be available, at least under Firefox. Ahasuerus 01:31, 28 March 2010 (UTC)

Another product is AutoHotkey which is more powerful but probably a little more complicated to use.--swfritter 12:57, 28 March 2010 (UTC)

Merging/making variant titles of interiorart

Once again this thorny issue causes intense bleeding. As simple as this issue appears at first, it becomes very difficult when figuring a way to implement. Short of limiting entry of artwork to one entry per story no matter how many pieces of art, every solution is ugly. Of course, the one entry per story solution would require massive amounts of work for existing data; it would also be desirable to retain data indicating how many illustrations there are for a story. For now I think it is wiser just to indicated that illustrations are "(reprints)". I am basically taking myself out of this discussion from now on. I will release the submissions from HOLD status in a couple of days or once there is some sort of consensus as to how to proceed.--swfritter 17:09, 30 March 2010 (UTC)

I don't like the reprint as variant of suspected original INTERIORART: those really can't be as sure as the more definite Interiorart reprints of COVERART, and they'd be ugly. If there's not enough certainty to merge, I would leave them alone. This isn't the ISADB, I see no reason to de-clutter artist pages as much as author pages. BLongley 18:00, 30 March 2010 (UTC)
Creating more records elsewhere is only an illusory de-cluttering, anyway. I am ambivalent about treating art as fiction [regardless of subject matter], especially to creating contents records for collections of art works. Using such works to verify artists is great, but to create more records, the variants, then who knows what else seems way over the top. Imagine if the Mona Lisa ever graced an SF cover. Would every art book ever to have that image in it now be fair game to add to the DB and that each would have a content record which would then become a variant......... As Bill points out, this is not the ISADB. Art as adjunct, yes; as equivalent, no. ~Bill, --Bluesman 19:10, 30 March 2010 (UTC)
Sorry about this, but this conversation seems to be off on a tangent. The issue simply put is that the interior art has been reprinted (in undoubted much better condition) within a layout showing to which story, in which year, and it was created by the artist. The No-Go opinion is that it can not be matched to a specific piece of art for two reasons, splitting of contents and either no common place to check art or captured Image of the interior art (remember no fairness of display as the art itself is generally owned by the artist or heirs, not usually the publishers). The reason to display it as a variant is that it connects to specific fiction at specific times. The reason would still be good if it was a Mona Lisa variant. (reprint) begs the question of what to where and you are all catalogers of items, including art which is specified for a publication, thus you know that an item safely ensconced in it's niche is more informative than a collection of titles without differentiation (even if they are reprint). To not connect is to make the question, "Why bother to catalog the art at all" (not to mention why parse it to second, third, etc). Thou it might be fun to say it is not the ISADB, you opened that door long ago and bemoaning that is not facing up to the idea that you have a functional solution to the problem. Try breaking the paradigm that visual cover art is held to a higher standard than published words. Then establish a working pattern of instructions to properly catalog the problem, as denouncing it as minor is obviously not working. First point) The DB can not store images that are not properly open to the fairness of display criteria, therefore you can never store images or have links to show the interior art or any variant of the art not clearly showing some cover attribute, i.e. if you found a book cover with just the art, you could never establish a fairness level to display it. Shawl-Shimying and veil-weaping will not solve the issue. My contention is that a hard standard of written connection of the printed art must be used. Thus the art most be attributed to a specific book or story, it must have a year (minimum dating) and it must have a specific artist claiming it's creation. The connected to does NOT have to be the specific one in the art, most reprinted art is a variant anyway, thus if it is variant link to the first instance for the published story/book should suffice. If any one, care, and obviously some do, then they should be satisfied or feel free to obtain the 'birth' material. In any case dodging the issue that art did not drive the success and influence of many stories is beyond reasonableness. In a future that reflects upon the past this issue will have more importance as time goes on. The human mind visuals as much as it meditates on the contents of a story, thus the accompanying art can be more relevant than a baseball card sold with gum. If I had known there was a 'hard rule' against I would not have connected in the first place, but I did it and made no differencing between cover art and strictly interior art in the "Broecker" book. I did not strive to create a new contest of wills or dissension, but if others have done so before, then obviously the issue needs serious treatment. When I finally read the 'hard rule' it seemed even then to be written to not cross connect the most parsed level of the art to each other, and even that made no real sense to me. If an artist creates a set of 2,3,4 or 7 more of interior art pieces for a subject title, then why those would not be cross connected can only be to preserve the idea that each is individual, but it defeats the idea that the set has a common basis. Parsing of art credit in a story makes most sense, when different artists have different contributions, otherwise it is hard to figure out what parsing of it does. I have parsed multiple art to different artist in a reprint omnibus, but I parsed it to the story level, thus assuring the reader/user could become aware that viewpoint had changed in that art. Believe it or not I can drop the issue, I was solidly trained in watch the train wreck itself, because we will not see it ideology, but I will give you the caveat that after the actual wreck everyone seems to have had a better idea. Why not work it out versus staving it off with unclear absolutes? On the record, I wished someone to object when I brought my Broecker art to the general attention. I could foresee objection, but I could and did not foresee it would not be seriously thought out. Thank, Harry. --Dragoondelight 20:24, 30 March 2010 (UTC)
"Shawl-Shimying and veil-weaping" [???]. A self-descriptive? "...a hard standard of written connection of the printed art must be used." Or else what? Art ≠ spec-fic. --Bluesman 21:13, 30 March 2010 (UTC)
Frankly the words are bad to use, my apologies. My intent is that people try to retain the points of the discussion to the requested level of attention, by sliding off subject, people show they are not paying attention to the central object of the discussion. My commentary was therefore out of place, but it was 'stimulated'. That people do not care a 'whit' for what I say is obvious, they are not in my boots, but if they try to imagine what I am saying then they show the common respect required of discussion. The bane of any discussion is it going of topic, though not paying attention to the details is probably worse. Believe me I understand the 'negative' position, most thoroughly, I also see the pitfalls that accompany it. My position is/has always been that 'Do Not' rules fail continually, while the pitfall of 'Can Do' rules is they expand. If you care to read further you will see my position is that the rule needs to change to encourage growth of research. The NO position makes it easy to do less, and thus miss the possibility that material written and artistic is missed. Sidebar issues about artist DB's is a past tense issue, this DB preserves art and the NO position is basically a protection for 100% validation that under current law and free use restrictions will never be obtained, therefore I see that variant link connection can create stimulation to obtain both the original and probably the best of the reprint, not to mention the reading of the written component. So forgive my snarking, but please give it an open mind. Too many have already found themselves 'standing to defend' the impossible to obtain. I am standing to 'obtain' the greatest common interest, which rather feeblemindedly I hope will create the desire to get all material available to the greater mass of people. To all I say that I have seen many things attributed in this DB which required much less work to find it's verity than what I have proposed. Thanks, Harry. --Dragoondelight 14:05, 31 March 2010 (UTC)
Harry, I approved most of the Broecker variants, but I'm not aware that any of them were interior work. All were cover art if I remember correctly. It seems you're not addressing the situation that Stephen (Swfritter) presented: the db handles interior art differently from cover art. The two methods used to record interior art can not be reconciled when it comes to reprints the same way cover art records were in the Broecker situation. Adding the "reprint" to the title record establishes the connection between artwork and fiction. No, there's not a direct connection between the two art records (variant or merge), but it's presently the best way to handle the situation. If there comes a time when the methods can be reconciled, it would be easy to find those records with "reprint" in the title and repair them. We did it with serials and that involved thousands of records. This would be a walk in the park in comparison. Mhhutchins 21:40, 30 March 2010 (UTC)
Apologies, Mike. To me this was not a 'separate' issue, and it was 'all' part of the same problem of whether to connect or not. There were only two interior art that connected in the manner I have been trying to attain. The problem is that I am seeing the variant link in one way and everyone else seems to believe the variant link attains something like 100% similarity. I see the link as a 'pointer' to creation. My problem is that if you have an art source and you check it out to it's origin, if you can NOT link then you have no reason to do so. At the same time, it is easy to understand that any art reprint is 'begging' to be checked out in it's original. The problem that I 'visualize' is that no one else sees that neither the cover art nor the interior prints survive the publication process in the 100% similarity, yet their development and changes tell their own story about the written material. They are all clipped, buffed, abridged, dusted, and even flopped. I tried to make it clear that I could see this in 'Broecker' at the same time I was working to make sure the art matched that of the printed publication. If you do enough, you can spot variations, some quite different, but they 'scream' their connection. I am pretty realistic in my perceptions, but I know many variations in cover art have themselves been approved, yet by the 'interior art' rule none could be done. Aging creates variances in many. My point is I am sitting here realizing that 100% identical is never attainable. I had this same argument about text in some writing, and the reply is that 'small changes' do not matter. In my simplistic hive mind that becomes a rule of thumb. Yet, I am meeting a wall of decision that is completely the opposite for the visual art, but the problem is simple, you can not ask people to scan it in, you can not show 'fair use' for visual art, nor for the reprint. Therefore you can not ever make 100% identical connectivity. If we agree here, then we come back to the 'why' we document at all. It is to show the history and connections of the written word, but while we recognize 'cover art' to the point, where we basically can not verify a 'coverless' book. We do NOT recognize the real importance of the illustration interior art. I am advocating that interior art was as important to the pre-tv and early tv eras as the television/monitor is today. Writing/art can not be easily separated in it's impact. Believe me the interior art would not ever be there if it did not create publication revenues. It therefore shaped the stories and was shaped to the stories. If you read the letters to editors, you can catch that the reader often comments on the art. To the reader the relationship is clear. Now, the reality is that in the books and longer stories the relationship is not to the individual but to the entirety of the set of art. I will stretch that to the original interior art which was often 'editor' varied to fit and any art that was made, but not included. These are in themselves 'treasures' of both the creation process and history. Thus I am advocating that any reprint connects to the original printed series. Okay, yes I know some editors have taken the trouble to separate the art to give it a 'unique' relevance in the paging sequence. They are stating what I am saying, "art has value in the stories". I think their concensus is that if I could make a one for one match, they would not object to the variant connection to their 'exact' image title record. I will gladly ship my copy to anyone who wishes to do that, but will that solve the problem that the images need to be connected to show their relevance to the story and each other. IMO that is too much to ask, since we do what we do voluntarily. Thus the Problem. In my mind the 'variant link' makes the unreasonable goal obtainable for comparison. It would certainly drive some advocates to obtain both the original and the variant, but what I also see is that it starts making the 'demand' for reprint or access or hopefully at some point, fully accessible detailed reproductions by computer. So, What seems impossible for me to convey to the few others interested in this is that the 'variant link' process enhances the original. Since, due to copyright and publication copyrights, and the limitation of 'fair use' we can not show the individual images, but we CAN link them and therefore stir the process and frankly start the process of creating that availability. I do Not understand why other than my inability to say this understandably, there is objection. The Opposition does not understand why I can not hold out for the ultimate connectivity. I say take 'action', and they say 'wait', yet the bulk of the material in question, both the written material and the art, is not getting any easier to get. Worse yet, the more it is 'worked' out the greater the connection to the author and as time goes by we see the greater the connection between the author and artist. Make no mistake, the process of checking and verifying the art should be there has created an avarice in myself for the original, but I will admit that when you get the 'older' pieces they often are of poor quality and they rapidly go into the classification of 'protective' storage to not be enjoyed. I think the 'variant linking' as a process that in itself states that the items are not 100% identical, yet for art we have a standard that denies this. Involvement in the process of checking art for being, there leads to finding material that is missing in this DB. Therefore, it has some value, but if 'IT' can not be connected to it's parent subject matter then it is both orphaned and the effort of doing anything minamalised. Editors who have parsed to the individual art level to a story, would object if that they were denied to show connectivity to the story in the publication, yet the only valid reason for not allowing those same items to be connected to form the 'set' of the work is that someone may find that some is not by that artist or for that written material. On that level, if the art is reprinted, and it meets some degree of verification, such as time/author/story it should get 'equal' treatment. Once the "why bother" stagnates a process, then that process is doomed. Again, too much, Obviously I have thought it out more than others or at least differently than others, it was not my intention to do so, but once a 'process' starts then it 'opens out' to the obvious to some, the oblivious to others. Does my work 'insist' that it be validated. NO, but it does insist that clarity in instructions be attained and like all NO instructions in processes it cuts of the tentacles of exploration to get the greater deed done. NO longer will I or others look at some piece of art and check to see if it's written subject has been included in the DB. The NO will result in the process curtailing exploration for new relevant written material as well as artistic. Is that important, only if you feel the project has merit in the first place. If this project is to be just a listing project, then it has exceeded it's parameters already. If it is to be a research site for it's subject matter, then the NO will cut at least one and probably other tentacles of research to recreate the written material. Sorry again to Mike, others, I never intended to try to breach this wall of contention, but once I submitted my variant connections, I did/do not want them washed away without a fair hearing. The hatred of the 'cancellation' when used on these variations will reflect on other's efforts I foresee. I apologize for 'sneaking' some past, but to me they are no different, and even now when I review the repudiation rule that is being applied to them, I do not see it as I am told it is. I see that rules and process have to grow and evolve to meet the needs, but others see rules and process as protecting the whole from the minor. Such is life, I ascribe to growth and reproduction, others to containment and inoculation. I too have my peeves, why some game books and not others? Will you read and find clarification here, probably not, most people are bored unless their reading as action events, therefore, the sword of justice was raised to cut the variant links of chain binding the project to one of course of action or another or No action at all. Either you can see the work and the beauty of connectivity for research or you do NOT. Apologies again to all who think I may have suckered them in, I also take it that everyone would have preferred to stop my process earlier, as do I. It is much easier to stop something before it starts, thus the review request of 'Broecker', than to let it continue to wind itself to what is a self-evident conclusion by the person doing it. My credo is if you can not try something, you probably will not value it. Thus I have rode on a roller coaster only twice, and both times for the sake of others. Thanks, Harry. --Dragoondelight 13:26, 31 March 2010 (UTC)
As briefly as I can put it: Harry, you're not addressing the original argument. (Some people have confused it as well, as Ron does below.) It has nothing to do with merging or varianting records based on 100% agreement that the two records represent the same art (such as the cover art in the Broecker collection). It's about how the original records were created. It's not simply the art itself. Some records represent ALL the pieces created for one story, while other records represent EACH piece created for the story. The question is: how can you create a variant or merge records when you can't know they're for the same work? If you have the original magazines, and can create records that are variants or merges of the same records for those magazines, I will personally approve EVERY submission that you make. If you're basing the variants on records created by someone else, you have to depend upon them taking the time to tell you if they're for one work or several works. If it's for several works, then a variant or merge can't be created. If it's for one work, you and the original verifier will have to determine which one matches the one in your book. I don't feel that most of the magazine editors/verifiers, based on those who have responded here, believe the results would be worth the effort. Mhhutchins 15:17, 31 March 2010 (UTC)
I think that is the original defense for NOT doing variants, not my argument. Believe me I thoroughly understand the dynamic on one for one identical matching, but it is almost impossible, the easiest course would be to mail my source book to each owner and let him parse it. Though that would work in a 'miserable' manner, since it demands effort from free willed people who may not wish to encumber themselves. It does not really address both the real problem, which is to show connectivity between interior art from other sources, hopefully stimulating 'research' that will bring forth the material into the common weal of those interested (or generate interest in the subject). Since it is almost humanly impossible to do all such, I suggested establishing a criteria of what is needed. The variant link can be thought of as a marker to promote future development in those interested in that facet, or who become so interested in the genesis of stories/art. The generic merge rule though is one thing, we all have done interior art (maps) and some interior art (frontispiece) on this basis, but in itself art suffers as much as written publications from change. My point in the variant link is the art is as is 'reprinted' but there are going to be in many cases alterations for effect by others, so it is with 'cover art' continually, so it must and shall be for interior art, and so it has been will be forever with written material, one can see for instance 60 reprints of 'Stranger in a Strange Land' with no alteration, then comes the manuscript printing as presented and we deem it a variant? Kind of like examining a chicken in the egg before hatching. Still I have little problem with it, because I feel the DB's job is to bring the alteration/change to people's attention and let them develop the process for greater analysis, etc. I think I have shown that 'identical matching' therefore is not practical, therefore we are dealing with variation in shape and form and thus the 'variant link' is being used to point this out, "Merging" however is the function of 100% identity. I'll stop here as most is rehashing of my points everyone else, LOL. I missed stating part of why my reasoning differed not a whit from cover art to 'interior art', I thought back and I remember that each time I started I generated the item as 'interior art' and therefore when I finished and wished to connect a variant link, I saw they differed only by the minor labeling from each other. There were variations between cover art, original cover art and by implication interior art. So my mind sees it to a differing degree and I knew there could be objections, though I did not see a 'steadfast' defence in place. I had already merged interior art (frontispiece) and (map), so the hard NO is a quasi-occasional, thus I am stating the NO rule does not work, as simplistic we have all done work arounds for effect. Again, if I had known that this kind of dissension would ensue, I would of cowardly kept away, but after creatively chasing down the connections for cover and interior art, I have discovered the real interest in recording this is in the variant link. It is the signboard of check this and here is this. I do not pursue the mistake, because that is a higher function of having all the material, I just am stating through the link elements may be available here and it is stated thusly. I know the original position of NO change, but that basically says look at the bottles on the shelf and guess if they have relationsips, by using the origin year, we are already pointing to similarity and making users guess what the next relationship is questionable. It is much like accepting a LOCUS dating on a book. What really makes them reliable, or do we always guess there is no relationship, believe me I have lived the 'let's check every discrepancy' versus 'accept some and not others' even the most novice person in any field will occasionaly find the dynamic item that saves lives and enhances the effort. Thanks, Harry. --Dragoondelight 12:20, 1 April 2010 (UTC)
The logic for not merging art, is the same reason that serial reprints of novels, or first serializations of novels are made variants. Between the original publication and the reprint, the story may have been rewritten, expanded, or abridged. Likewise with the interior art. A reprinted story may not reprint all of the art in the original, so there may not be a 1:1 correspondence between the two, or the sequence of printing may be altered. The art may have been redrawn for reasons of copyright(?) (British Astounding covers are a good example of this), i.e., the covers look identical from a distance, but are distinguishable at close range.--Rkihara 22:58, 30 March 2010 (UTC)
Apologies, that this disturbs you. I understand the 'negative' position to my appeal. I also read more into it than I presume others do. I know the 'truist' vision is that variant of novels has been used as you describe, but in actuality it has been used in a multitude of variations of what you think it has. Not only that, but when some obvious variations are pointed out in text, they are most frequently relegated to the positions that 'minor editor variations' are not included. My position is that change a word, change a world, but that is not the 'usage' of variant linking. Most variant linking is connected to wording in title and name. The same is true of cover art, no one can see all the variations of colouring and frankly the small changes probably can never be properly addressed unless one obtains both the original cover art and the as used for a printing and then all the variants (which we do not make) of that art. I know that the logic behind not connecting interior art reprints to original is that it can not be shown 100% identical. That is a tough standard for anything/everything. The reasoning for not merging the [2], [3], [4] of interior art in a story is to preserve that order. But, that standard is not required when entering the art, so it is mostly a 'respect' for the effort of the person doing that work. My 'proposed variant linkage' is to the first or only of such series, it is not meant to show 100% identity, but to show the claimed, and documentation obtained through the reprint source. This is odious to most/all, but I can see the stimulation of seeing the connection to the original and it's root written component. It opens the possibilities of comparison and frankly advised users where there is other material on the interior art. I will not pretend that all art gives me some idea to mull, but occasionally a piece of art will spin my imagination to wild heights and yet another in the same set will tone my imagination in another way. One of my favorite short stories, was done as comic and it did little for my past the grand eloquence of the writing, till that last panel and a small blurb of 'evermore, my love, evermore'. Imagine my chagrin in reading the original without art that the artist added the blurb to create a memory forever. Art is in itself an experience and many people never drift or absorb it till such a moment. Thus I see 'variant linking' accomplishes something in certain cases and in others it would not. I do not think it lessens the original in any case, but may with those with 'deep' pockets promote the obtaining and preservation of material that is getting rarer by the day, not to mention altered. I appreciate your input here, and hopefully my long windedness can be slightly forgiven for the obviousness impact the process I used to obtain the connection. That envision others could also do so if it is 'mapped' by the variant linking process apparently is not appreciated. That I see stimulation to aquire and document on a personal basis, may not be apparent to others. I can not insist others share my 'wild ride of imagination', but I am glad you at least tried to see my point. Thanks, Harry. --Dragoondelight 13:50, 31 March 2010 (UTC)
I can see your argument, and agree that ideally all records for similar, near identical, and identical works of art should be connected in the database. Unfortunately, the two methods of entering magazine interior art has caused an inability to do this. Perfect world resolution: demand all interior art be entered the same way and correct all records that are currently in the database. Real world resolution: ask editors to enter as much information as they feel they can, and enter all reprints of interior art with the "(reprint)" appendage. Mhhutchins 15:17, 31 March 2010 (UTC)
I would not have a problem if you are able to verify that two pieces of merged art are "identical," but if you can't, then merging them incurs a high probability of error, and I'm not talking about trivial details. I have magazine runs extensive enough to check both the original publication and the reprints against each other. I've seen all of the situations that I mentioned previously. The British Astounding covers are complete re-paintings of the originals, and not editorial alterations or over-paintings of the originals, done by a completely different artist (always uncredited). I should also mention that reprinted art is often of much lower quality than the first printing, since the "original" art is often no longer available and the reprint is a photocopy. In addition, the first printing was in a magazine that was either "pulp" or "bedsheet" size, and the reprint is reduced to a digest size.--Rkihara 15:48, 31 March 2010 (UTC)
Thanks, for taking the time, I appreciate the effort it must be to express something that is solidly understood by you. Also apologies, for the occasional bother caused when I 'passed' through checking items and occasionally getting them to meet the data I encounter. I know it is exasperating. I do know some of the problems and believe me that has lead me to the conclusion that it will need to preserved to fully appreciate it. Identicality is almost impossible to measure, at least for me, in fact to add to your ideas against, I read a small article in a rather recent magazine that talked of George Barr and three other artists working there. In essence the article states that it is not uncommon for the signed work of an artist to have been altered by the other three as needed. I can not see how this would not be true for many more of the actual 'signed' works we have. My case is thus not based on identical which is a merge action, (possibly?). My 'variant linking' frankly ignores the [2], [3], etc action, for the above reasons. Each would need detailed analysis that requires tools of inspection that I and very few have. My actual point is the first indent the [0] is established as the 'keystone' to the others. My connection is to that for the purpose of showing (not establishing irrefutably) that other 'reprint' art has been 'stated' as in that keystone set. I imagine do to variation and alteration that very few art items really could be match, but when they are it will be to establish the variation and alteration, but for those doing it, it will not matter what medium they use (other than greater clarity). Your analysis of reprint quality is agreed, and I am not looking at for instance "Cosmos/Wildside" rips of original art, I am looking at stated artist work with extra specificity to what, though I have a feeling some could be wrong. The quality of the reprints in question is far better than what was published originally. Frankly, if I understand the methods the original art was produced under I am amazed they could do this with such quality. Just that differentiates it from the original, making identicality questionable. Back to OUR points, All the commentary really add up to changing the "NO" to a rigid rule set. You need this, that, that is always more workable. The NO is being circumvented by people stating identical and merging, but what qualifies us to do so. I have a dozen maps I wish to merge, but I do not have acumen to do so. There could be variance, possibly important, just the reprint process in some books has made identification of some ON/OFF. I would suggest getting a copy of "Virgil Finlay's Far Beyond" but the prices seems to have jumped and apparently the higher price copies spurred reprinting? Do people read this junk? Rather than beating the limited audience here any longer with my viewpoints and defences I will follow this with a summary/conclusion as a 'deadline' was established in the first place. Thanks to you and any others who tried to follow my reasoning, especially if it seems unfathomable, Thanks, Harry. --Dragoondelight 12:55, 1 April 2010 (UTC)

(Unindent/Finis)I really do understand how hard it is to understand my position and do wish it never need be said. But it is and has. Therefore, I offer this as the resolution and the example of the hardness I have in destroying the linkages. I ask that anyone, with any interest in the subject, check out the entry and the linkage to the original art, go there see what is there, and then 'push that reject button, knowing that a variant linkage had no value' 'that it leads to nowhere/nothing'. Think this and continue till you have finished, thus hopefully you will be 'armored' for future incursions by others. I would/will not spend the time researching other connections for 'no purpose' as it would reopen this wound. This is tough duty, but once done you will feel refreshed in your positions. Thanks, Harry. --Dragoondelight 13:03, 1 April 2010 (UTC)

Is the consensus that variant title processing of interior art is not a viable option at this time? If so, I will reject the submissions in a couple of days.--swfritter 15:02, 2 April 2010 (UTC)
I concur. ~Bill, --Bluesman 15:29, 2 April 2010 (UTC)
I concur. Perhaps we should set the wiki to disallow merges of artwork until a protocol has been agreed on?--Rkihara 18:06, 2 April 2010 (UTC)
Maybe just for interior art, cover art is pretty easy to compare. ~Bill, --Bluesman 19:16, 2 April 2010 (UTC)
The arguments against allowing VTs seem to be more convincing at the moment. As far as software changes go, it would be easy to modify "Check for Duplicate Titles" to ignore identically spelled Interior Art records. Ahasuerus 19:51, 2 April 2010 (UTC)
Sorry but that was not the method I used, it was a physical check item to item, I was not linking groups, but if you are thinking group type bans. Essays and (excerpts) seem likely candidates also. Thanks, Harry. --Dragoondelight 21:54, 2 April 2010 (UTC)
I will reject them tomorrow and propose a more rigid change to Help. As always, there will be special cases. Mastheads and repeated column art, like covers are not multiple art. I think RTrace has physically verified the mastheads.--swfritter 16:05, 3 April 2010 (UTC)

Server downtime?

According to our tracking service, which monitors ISFDB availability by checking the main page every 5 minutes (give or take), we had 100% uptime in March. Does it match everyone's experience? I don't recall the server going down, but I didn't spend a whole lot of time editing during the last month. Ahasuerus 22:01, 2 April 2010 (UTC)

I did not notice any problems in March. --MartyD 10:14, 3 April 2010 (UTC)
An occasional temporary burp where I had to do retries which were usually immediately accepted.--swfritter 16:06, 3 April 2010 (UTC)
No problem your end, my provider though is doing the slows, stalls, and stops. Excellent job keeping it up. Thanks, Harry. --Dragoondelight 23:07, 3 April 2010 (UTC)

Galaxy question

Hi -- haven't been here in forever, and there are a million notes on my talk page. If any are important I do finally have my books out of storage again after a year or two in boxes so I can verify anything that's been requested.

The reason I am posting is that I hope someone here can help with a question about Galaxy. I am working on the Wikipedia article and would like to be able to cite the issue where Gold announces the writing contest that would be won by "Edson McCann" for Preferred Risk. It must be mid-1954 or earlier; I've scanned my issues and can't find it. Does anyone have any idea when this was? Or is there a way to search the isfdb for a record that mentions "Contest" in a title that includes "Galaxy"? Thanks for any help. Mike Christie (talk) 00:23, 7 April 2010 (UTC)

It's the March 1953 issue, p. 77, rules on p. 129. Reference found in David L. Rosheim's book "Galaxy Magazine: The Light and Dark Years," verified by checking the issue.--Rkihara 02:43, 7 April 2010 (UTC)
Also referenced in Mike Ashley's "Transformations," p.119.--Rkihara 02:48, 7 April 2010 (UTC)
Doesn't look like our entry has it, so the search wouldn't have found it. But you can use "%" as a wildcard (match any arbitrary text) when searching. So, for example, searching All Titles for "Contest%Galaxy" would find any title containing the word "Contest" at some point before the word "Galaxy" (and "Galaxy%Contest" would find titles with the word "Galaxy" at some point before the word "Contest"). Or you could Advanced Search and do TITLE Galaxy AND TITLE Contest to find any title containing both the word Galaxy and the word Contest anywhere. Simple and obvious, no? --MartyD 02:05, 8 April 2010 (UTC)
Thanks for both the location of the contest and the help on searching. Much appreciated. I am not active here any more but this is still one of my top online resources for sf research; thanks to all of you who are doing this work. Mike Christie (talk) 12:13, 8 April 2010 (UTC)

Help on searching

Is there a way to search the database for all stories by author X in magazine Y? I am looking for Varley stories that were published in Galaxy but couldn't figure out a way to do it -- putting in "Varley" for author on the advanced search form, plus "Galaxy' for title, doesn't work, presumably because the title field is for the story title in that context. What I'd need is a "containing pub" field, I think. So is there a way? Mike Christie (talk) 01:00, 10 April 2010 (UTC)

You'd need to get a copy of the database and do it through SQL. The best the current Advanced Search interface lets you do in this case is a title search for Author = Varley and Title Type = SHORTFICTION. Then you'd have to look at each title to see the publication appearances. --MartyD 12:25, 10 April 2010 (UTC)
If you're interested in the SQL way, here's the query:
select t.title_id, t.title_title, t.title_copyright, p.pub_title, p.pub_year
  from titles t, pubs p, pub_content pc,   canonical_author ca, authors a
 where a.author_canonical like '%Varley'
   and p.pub_title like 'Galaxy%' and p.pub_ctype = 'MAGAZINE'
   and ca.author_id = a.author_id and ca.title_id = t.title_id
   and pc.pub_id = p.pub_id and pc.title_id = t.title_id;
and the results of that against the 2010-04-03 database back-up (see ISFDB_Downloads) are:
41067The Phantom of Kansas1976-01-31Galaxy, February 19761976-01-31
51670Bagatelle1976-09-30Galaxy, October 19761976-09-30
51671Gotta Sing, Gotta Dance1976-06-30Galaxy, July 19761976-06-30
51672Overdrawn at the Memory Bank1976-04-30Galaxy, May 19761976-04-30
589401The Ophiuchi Hotline1977-10-31Galaxy, November 19771977-10-31
620205The Persistence of Vision1979-02-28Galaxy, March-April 19791979-02-28
976702Bagatelle [2]1976-09-30Galaxy, October 19761976-09-30
Note that "%Varley" is because we know it's a last name (you'd have to use "%Varley%" if you don't know where in the canonical name that portion of the name occurs) and "Galaxy%" is because we know the magazine titles start with Galaxy (you'd have to use "%Galaxy%" if you don't know where in the publication title that word occurs). --MartyD 12:25, 10 April 2010 (UTC)
I might note that Bagatelle [2] was an input error. Varley was not the artist.--swfritter 13:27, 10 April 2010 (UTC)
Thanks for the details; and also for the list. I know SQL so I will think about downloading the DB at some point in the future. Perhaps this form of query could be added at some point? It's particularly useful when you want to do research on a magazine's history, which is what I'm doing now -- e.g. Brian Aldiss, in Trillion Year Spree, claims that women began to write more sf in the 1950s when Galaxy and F&SF appeared, and he lists some writers such as Henderson, deFord, St. Clair, and a few others. That's not easy to research from the ISFDB as it stands. Mike Christie (talk) 19:33, 10 April 2010 (UTC)
If you're interested in Women in SF, "Partners in Wonder," by Eric Davin is an excellent reference. His book debunks the idea that women are relative newcomers to the field, pointing out that 203 women published almost a thousand stories between 1926 and 1960. An additional 127 women can be added to the list if you count Weird Tales, and another 31 from less well known magazines.--Rkihara 20:45, 10 April 2010 (UTC)
What a coincidence. Just had the book in my hand today but I have to wait a week for the local university to process my alumni library card. There are a lot of contributions from women in the letter columns of the mags in the 50's and earlier. That might have been the editor's choice but it might also have more accurately reflected the readership. Maybe Ron has some ideas as to how we can get the names of the authors mentioned in the Davin book into the database. If the discussions in the book lend themselves to that could we consider some 'book reviews'. We don't have a gender field in the database and it is often difficult to determine the women writers because they often used pseudonyms or initials to disguise their gender.--swfritter 23:30, 10 April 2010 (UTC)
The lists of female authors in Davin's book were mostly taken from US and British magazines, so I think most of the authors are already in the database. There are probably only a handful hidden behind initials or pseudonyms (there is a moderately long list of writers of undetermined gender). One of the points he made was that unlike the other pulp genres, women were openly welcomed as SF writers and didn't have to hide their gender.--Rkihara 05:16, 11 April 2010 (UTC)
In about 1982 Connie Willis wrote an article (probably an editorial) for Asimov's claiming that women had always been a strong presence in sf. I remember my wife was sceptical at the time, so she and I went through my collection and counted the percentage of women who appeared in a given year for each of several magazines, for one year from each decade. I think we went back to the fifties; not sure if we did the forties. We included New Worlds, F&SF, Astounding, and possibly one other magazine; we put the results in a letter to Asimov which was published a few months later. As I recall the numbers showed a very low percentage of women through all the magazines and all different decades. New Worlds was the lowest, I remember; I believe the percentage of women in the 1950s in NW was about 2-3%. I don't remember the other numbers except that there was a reasonably clear upturn towards the end of that period; I think F&SF got over 30% women in an early 1980s issue, for example. SF may well have accepted women, but I think that women writers accepting sf in large numbers took longer. It would be interesting to see the stats done again, more rigorously. Does Davin do that? That is, does he show total number of stories by men at each period, so that percentages can be calculated? Mike Christie (talk) 11:21, 11 April 2010 (UTC)
It would be kind of nice if we had a SQL library, even if just cut & paste. Setting up MySQL is generally pretty easy (don't know about Apple).--swfritter 23:30, 10 April 2010 (UTC)
I dumped some of my initial SQL thoughts here. But yes, a more organised page with scripts for particular tasks could be good, no need for us to each reinvent the wheel. BLongley 17:16, 11 April 2010 (UTC)
Davin identifies and lists 154 female writers active in the period between 1950 and 1960, and 56 between 1961 and 1965. Women who were identified with fantasy publications were also listed, but not counted in the totals. I think Davin calculated some percentages, but I don't recall how large they might have been. Women writers would certainly be a minority, as they are today. Davin's survey is extensive, appearing to cover almost all of the published magazines in the periods studied. The results are summarized in the bibliographies in the back of the book which lists all identified authors by name and all of their published stories. Each chapter is also followed by an extensive body of notes. Connie Willis' article is referenced in the opening pages of the book. Davin spends a considerable amount of time refuting assertions that the field was prejudiced to women, pointing out not only the numbers, but that a significant number of the editors were also women, and that negative response from the readership was nil. More significantly, some of the women writers past and present seem to have bought into the myth, to the extent of revising their personal accounts. Memory is malleable. I'd like to read your letter, what issue of Asimov's was it published in?--Rkihara 16:16, 11 April 2010 (UTC)

(undent) I just dug it out (thanks to the ISFDB for helping me find the Willis essay); it's in the May 1993 issue. Turns out my memory is worse than I thought; Willis's essay was in October 1992. Here's a table (the Asimov's 90s number isn't explicitly given in the letter so I've put a question mark; the letter just says Asimov's maintained their percentage, so it would have been around 23%).

Asimov'sn/an/a25%23%23% ?
New Worlds1.5%1.4%n/an/an/a

-- Mike Christie (talk) 16:41, 11 April 2010 (UTC)

Thanks! I pulled both and read Connie's essay and your letter (and reread a couple stories to boot). I think Connie and Davin are arguing against the perception that there were almost "no" woman SF writers before the late sixties, and not that they were there in a great multitude. We have no way of easily investigating this until we can set a "gender" flag on the author entries. We'll probably need to verify gender entries too, as both men and women wrote under pseudonyms of the opposite sex.--Rkihara 22:29, 11 April 2010 (UTC)
And don't forget those who medically and legally changed both gender and names.--swfritter 00:51, 12 April 2010 (UTC)
I haven't reread her essay, but as I recall she was saying that there was a perception within the sf community that the situation had changed since the 1950s and more women were now writing sf. She said that that was wrong; that women had always written sf. Our letter was basically arguing that she was half right -- women did not begin to contribute more in the 1960s and 1970s, but that it the level was so low that her argument was misleading. As I recall, Willis was not a strong supporter of feminism and some of its goals, and we felt her article was a bit unfair in the way it characterized the history. But I haven't reread it and you have, so I probably should keep quiet. I wonder what the gender ratio is now?
Re a gender flag: yes, that would be a nice feature! Mike Christie (talk) 01:43, 12 April 2010 (UTC)
The magazine was still out, so I read it again to make sure I had it right (one the first things to go . . .). The essay starts "The current version of women in science fiction before the sixties goes like this: . . . There weren't any." Your summary of your letter is as I interpreted it. As I see it you and Connie are on the same side of the argument, i.e., that there has been no radical change in the number of woman writers. Connie arguing for continuity, and your statistics demonstrating that.
I would think that there are too few transgendered to worry about, only Jessica Amanda Salmonson comes to mind.--Rkihara 02:48, 12 April 2010 (UTC)
There is also Hank Stine aka Jean Marie Stine. BTW, good to "see" you, Mike, glad you are still working on Wikipedia -- that takes more patience than I have! :)
Re: the original question, there were some female writers active in the pulps before the Golden Age, e.g. A. R. Long, but not a lot. The percentage of "femfans" was also very low, perhaps even lower than the number of female writers. Except for C. L. Moore, they produced few memorable works and are mostly forgotten now.
Leigh Brackett stood out in the 1940s, but many fans didn't realize that she was female, something that Kuttner took advantage of in one of his "recursive SF" stories. Wilmar H. Shiras made a splash with "In Hiding" and its sequels in the late 1940s and Margaret St. Clair was also reasonably well known after the war, but a vast majority of SF writers were male during the Golden Age.
As of 2010, things are very different in some sub-genres, e.g. urban fantasy and paranormal romance, but take a look at the ratio of male-vs-female writers on the Locus best SF and Fantasy Novel lists for 2009. On the gripping hand, J. K. Rowling probably sold more copies in the last decade than all male writers on the Locus, Hugo and Nebula lists combined :) Ahasuerus 03:40, 12 April 2010 (UTC)
From my experience with magazines Leigh Brackett was known to be a woman from the start. Ray Palmer introduced her in his intro to her first story in Amazing, published in January 1941, and she was featured with a photo and a biography in the same. Other editors likewise introduced her, and her photo and biography appeared in several magazines in the forties.
Many woman authors were well known and popular in their day, but forgotten, as were many of the men. Who remembers Nat Schachner, Richard Shaver, or Nelson Bond?--Rkihara 15:48, 12 April 2010 (UTC)
It's certainly true that most pre-Golden Age writers (including Schachner and Bond) didn't adapt well during the Campbell era and have been mostly forgotten, although Bond had a mini-revival of sorts a few years back. That's pretty much par for the course in any and all genres, e.g. see E. Phillips Oppenheim, one of the most popular and prolific writers of his time, now almost entirely forgotten. However, even during the first dozen years of pulp SF, before the boom of 1938-1941, female writers were not at the top of the pyramid, which isn't surprising given their low numbers. Ahasuerus 17:06, 12 April 2010 (UTC)
It occurred to that you might have meant C. L. Moore instead of Leigh Brackett in connection with Kuttner. She was also known from the thirties to be a woman, shown by reader correspondence, and the fact that she signed her full name when responding to the reader's columns.--Rkihara 16:14, 12 April 2010 (UTC)
Oh no, it was Brackett all right. The story in question was "Reader, I Hate You!". Although Brackett's sex was not a secret, not everyone was paying attention, a common occurrence. At one point circa 1990 a voracious SF reader and an amateur writer (although not a fan in the traditional sense of the word) told me that C. J. Cherryh was her favorite writer. She was stunned when I mentioned that Cherryh was a woman :-) Ahasuerus 17:06, 12 April 2010 (UTC)
I think I got Andre Norton wrong at first in my pre-teen years, but the only surprise since has been Julian May. Both read as male first names to me. BLongley 19:19, 12 April 2010 (UTC)

WorldCat links

Links from pub records to WorldCat are no longer valid. Was it us or them that changed the way pubs are linked? Mhhutchins 23:41, 14 April 2010 (UTC)

It's on their side -- even the examples of ISBN linking that they provide on their Help pages are broken now. We'll see if it's a temporary problem with their software or a new direction that they are taking. Ahasuerus 00:54, 15 April 2010 (UTC)
You can still search using an ISBN but you have to do it through the Advanced Search option. ~Bill, --Bluesman 16:01, 15 April 2010 (UTC)
That's the advanced search on their website, not the ISFDB's. I didn't realize how much I used the linking from our pub records to the WorldCat records. Mhhutchins 19:15, 15 April 2010 (UTC)
Just checked again, and the linking is back. Hooray! Mhhutchins 19:19, 15 April 2010 (UTC)

Complete Stories of Theodore Sturgeon

Does anyone know why volumes 8 (Bright Segment, 2002) and 10 (The Man Who Lost the Sea, 2005) of this series are impossible to find at a fairly reasonable price? Did North Atlantic severely cut the print run for these volumes? Volume 7 is also OOP, but used copies can be found at about $50. Strangely, copies of Volumes 1-6 are cheaper than later volumes. Even more strange, Volume 9 (And Now the News..., 2003) is still in print and available from Amazon at a 30% discount! Mhhutchins 02:26, 15 April 2010 (UTC)

Just got an email from Karen Windham of North Atlantic Books:
"In conjunction with the last volume coming out in September (Case & the Dreamer) we will be reprinting all of the other volumes that are out of stock, in Hardcover format. There are no definite ETA's for these reprints, they will be staggered though. I believe Bright Segment is at the top of the list."
They won't be first editions, but at least you won't have to pay $200+ for one volume just to have a complete set. (The reprintings may also lower the value of the first printings.) Mhhutchins 18:06, 4 May 2010 (UTC)

Our Home page

I know we haven't sorted language preferences out, but should Mas muerto que nunca really be on our home page at the moment? Can anyone here support that language? BLongley 21:00, 15 April 2010 (UTC)

I don't think there is a way to prevent non-English pubs from appearing on the home page until we add the ability to assign languages to pubs/titles :( Ahasuerus 00:40, 16 April 2010 (UTC)

The blurry line between Essay and Shortfiction

For background, see User_talk:MartyD#Introduction_.28The_Berserker_Wars.29. Saberhagen's 1981 collection of Berserker stories includes an introduction that is (a) written in the first-person voice of a fictional character and (b) set in the future (at a time well beyond the Berserker Wars). It's in the form of a cover letter from an archivist, providing some setting and historical background for what it calls the "Historical Documents" -- a term referring to the stories in the collection. It does this in a very stylized way, speaking of other fictional characters' opinions and so on. It is clearly a work of fiction. At the same time, it is clearly a factual introduction to the collection, rather than another story or a prologue or epilogue.

Willem H. is inclined to treat this as SHORTFICTION, while I am inclined to treat this as ESSAY. Neither of us has a particularly strong opinion or emotional attachment to his preferred treatment, but I thought since it came up we could see if there's a preference among the community at large (or if anyone is aware of existing precedents that could guide us). Opinions and other feedback solicited. Thanks. --MartyD 10:20, 17 April 2010 (UTC)

Ancient issue. My own current inclination would be to classify this as SHORTFICTION and leave the length as SHORTFICTION.--swfritter 14:42, 17 April 2010 (UTC)
As Swfritter said, there is no easy way of classifying in-universe essays, which is the main reason why we have an outstanding request to create a separate sub-type for them. For what it's worth, in this case it sounds more like Shortfiction. Ahasuerus 17:15, 17 April 2010 (UTC)
I prefer fiction to remain as SHORTFICTION even when written in Essay style. Leaving the length blank is a bit of a clue that it's not a short-story, though not conclusive. BLongley 18:16, 17 April 2010 (UTC)

Slight change to Publication display

Series numbers are now displayed in the Contents section of the Publication display page, e.g. see this omnibus pub. It also works for pseudonymous titles, e.g. see Worlds of The Golden Queen. Ahasuerus 23:59, 17 April 2010 (UTC)

Tales of the Talisman

While processing Fixer's submissions, I stumbled on the Tales of the Talisman archive. It occurred to me that we may want to capture their issues' tables of contents while they are still online. Ahasuerus 04:49, 18 April 2010 (UTC)

Captured volumes 1, 2 and 5 - 3 and 4 aren't on that site. There's some obvious typos - e.g. Rick Yennick or Rick Yennik? - but I'm sure we can sort those out later. Someone may want to double-check, do the Wiki page etc: and there's page numbers on the Locus site for some issues that could be added. BLongley 11:34, 27 April 2010 (UTC)

Department of classic excerpts

From Ralph Milne Farley's The Golden City (first published 1933): He (the villain) stepped down from his throne and began to draw his sword. "Just a minute." hastily interposed Mayhew. "My sandal is untied."--swfritter 15:32, 24 April 2010 (UTC)

ISBNs display corrections

Thanks to Marty's indefatigable efforts, ISBNs starting with 6, 7, 8 and 9 should be displayed correctly from now on. (Ahasuerus 02:24, 27 April 2010 (UTC)

If anyone notices ISBNs with dashes in the wrong places (groups of numbers not the right size), let me know. I didn't have a good way to ascertain how well the rules from the ISBN site correspond to actual usage. --MartyD 02:46, 27 April 2010 (UTC)
And thus having cursed myself, I found that the penultimate hyphen is being placed one digit too far to the left in some cases (making publisher group one too short and the title group one too long). Fix coming. --MartyD 10:19, 27 April 2010 (UTC)
This affected only the multi-digit group codes (ISBN-10s starting with 6, 8, or 9; ISBN-13s starting 978-6, 978-8, or 978-9). The publisher range comparison was not being done appropriately, resulting in all ISBNs from the same country getting the same hyphenation pattern, regardless of publisher. A fix has been checked in. --MartyD 11:27, 27 April 2010 (UTC)
And what about this? Should an ISBN-10 actually be generated? The ISFDB uses the ISBN-10 for the Barnes & Noble lookup but only the ISBN-13 is valid for that purpose.--swfritter 14:54, 27 April 2010 (UTC)
I'm pretty sure that's not a valid ISBN. Only 978 and 979 are valid on ISBN-13s (see this FAQ entry). So we should probably flag it as invalid. That would then prevent both the conversion and the formatting. The "valid ISBN-13" check does not test the first three digits, which is why this slips through. I don't know what the ramifications of flagging an ISBN as invalid are in the displays and other features of the site, though (e.g., in the B&N look-up). --MartyD 23:36, 27 April 2010 (UTC)

(unindent)The fix has been deployed. I am not 100% sure that every single Zimbabwean ISBN will display correctly now (the changed section of the code is over 2,000 lines long), but at least nothing got broken (that I could find.) Ahasuerus 03:40, 28 April 2010 (UTC)

ISBN's would seem to be a subset of EAN's and appear essentially to be barcodes. The 978 and and 979 would seem to be indicators that the product is a book. So it would seem that the prefix that Barnes & Noble uses also is some sort of a product designator.--swfritter 03:52, 28 April 2010 (UTC)
The ISBNs for this publisher are not being displayed correctly, e.g. 0-9650188-5-7 should be 0-965-01885-7. Thanks. Mhhutchins 05:30, 2 May 2010 (UTC)
The rule for 0-965... says the publisher group size should be seven digits. There are other exceptions (by the rules, 0-765-x should be 0-765x and 0-812-x should be 0-812x, but we had -- and still have -- exceptions for those in place because that is how the publishers (Tor and others) use them; you'll notice all 765 Tor pub ISBNs start 0-765-3xxxx and their 812 pub ISBNs start 0-812-5xxxx because these group numbers assigned to Tor are 4 digits). Googling "0-9650188" finds plenty of hits, though, so I'm not so sure about this case. I tried Googling "0-965-01880" through 01889, and I found only two hits -- a random site for one pub and an Ebay auction for the pubication you cite above (there's a third site that had a hit, but it had many variations of one ISBN, moving the hyphen around). Do you have example(s) printed on the book(s)? The Ebay hit makes me think just that one pub may have it wrong.... --MartyD 10:51, 2 May 2010 (UTC)
p.s. The "rules". --MartyD 10:53, 2 May 2010 (UTC)
I was the person who made the inquiry about the false publisher identification for that Ebay auction. When the auctioner responded with the corrected publisher information, I created the ISFDB record based upon his description including the way the ISBN is presented. I don't think he would have arbitrarily recorded hyphens that weren't present in the book. I could be wrong. I see a couple of sellers on are using the seven-digit publisher identifier, but that doesn't prove anything. Sellers can't record ISBNs as stated if they don't match ISBN standards. Look back on the listings for Book-of-the-Month Club in 2001. Would it make sense that they would have four different publisher identifiers for five titles? The 0-965 prefix also appears in 1999 and 2000. Mhhutchins 14:25, 2 May 2010 (UTC)
It's getting more complicated than I first thought. There are many publishers that have books with ISBNs starting with 0-965, so it appears that if BOMC placed the first hyphen after that prefix they were mistaken. (Do an ISBN search for "0965".) I see now why the system chooses to give the publisher identifier as seven digits. Sorry for bringing up this non-issue. Thanks. Mhhutchins 14:45, 2 May 2010 (UTC)
Well, I grant you it's pretty bizarre, as a 7-digit code would get them only 10 ISBNs (it leaves just a 1-digit title block). From what I've seen, though, those rules are accurate, which means that is indeed what they did. So from our software's point of view, it's more a question of how do the publishers represent the ISBN pattern during printing. In researching the validity of the existing 0-765 and 0-812 exceptions, I found that Tor and at least one other publisher sharing the same first three digits both used 3-digit presentation of their 4-digit numbers. So I left that packaging-as-three-digits in place, even though those are 4-digit group numbers and the hyphenation after three digits is technically wrong. We can override the correct hyphenation pattern for any groups or subsets (even individual ISBNs) that seem appropriate. I figure it would be very confusing to have our software use a different pattern than what we see on the books. But I do think we want to go by evidence from the books, as people might hyphenate unhyphenated ISBNs in arbitrary ways. --MartyD 18:24, 2 May 2010 (UTC)
7-digit codes do indeed exist, and are usually bought by individuals. A friend bought such a block and has only used one of them so far. (978-0-9782719-7-8). BLongley 10:58, 3 May 2010 (UTC)

Nominating editor Willem H. for moderator

See Moderator Qualifications#Becoming a moderator for the nomination process.

Willem (talkcontribs) has shown good familiarity with English works and I believe he is also an asset on Dutch and German works too - maybe more, as he claims some familiarity with three other languages (i.e. five more than me). The ability to at least self-moderate would also help him work faster, or with less effort. Willem is willing and I think meets the Moderator Requirements. BLongley 18:32, 29 April 2010 (UTC)


  1. Support, as nominator. BLongley 18:32, 29 April 2010 (UTC)
  2. Support. Willem has 6,467 edits as of today and possesses good working knowledge of the submission creation/review process. He has contributed to numerous discussions over the last few months, demonstrating good communications skills and ability to work with other editors. Ahasuerus 18:41, 29 April 2010 (UTC)
  3. Support. Wholeheartedly! ~Bill, --Bluesman 00:27, 30 April 2010 (UTC)
  4. Support. I see a lot of his submissions (and he makes a lot of submissions) and was having a Bill Longley moment just the other day, thinking to myself "Why isn't Willem a moderator already?". --MartyD 02:10, 30 April 2010 (UTC)
  5. Support. Most certainly.--swfritter 13:16, 30 April 2010 (UTC)
  6. Support. Agree with all of the above. He'll make a great asset to the team. Mhhutchins 20:50, 30 April 2010 (UTC)
  7. Support. Very careful, most complete. Harry, --Dragoondelight 11:50, 1 May 2010 (UTC)
  8. Support. No problems with any of the submissions I've approved, should be ok now.Kraang 02:54, 2 May 2010 (UTC)


Neutral/Comments I believe the five days are up? Or are we waiting to have this nomination translated? BLongley 18:57, 4 May 2010 (UTC)

Did you notice that you posted exactly 5 days and 25 minutes after the nomination? :) Ahasuerus 21:23, 4 May 2010 (UTC)
Yes. With Fixer's submissions still clogging the queue after a few days, I thought "the more mods looking at them the better". Consider the 25 minutes an allowable margin of error in case we haven't got clocks synchronised. BLongley 20:25, 5 May 2010 (UTC)
Definitely time for the hot coals!! ~Bill, --Bluesman 19:10, 4 May 2010 (UTC)


Nomination passes. Congratulations, Willem, you are now a moderator! Ahasuerus 21:23, 4 May 2010 (UTC)

Thank you all for the heartwarming support. I'll try not to abuse all this power. --Willem H. 08:56, 5 May 2010 (UTC)

Content import/export changes

The export/import screen has been slightly changed. Uneditable fields are now grayed out to match the behavior of other display-only fields in edit screens. Ahasuerus 00:16, 2 May 2010 (UTC)

Publishers linked on the Title page

In the "Publications" section of the Title page, publisher names are now links to the associated publisher pages. The separator between the pub author(s) and the publisher link has been changed to the semicolon to help distinguish between authors and publishers. You can see what the result looks like when multiple publishers have published the same book.

Although the additional functionality appears to be useful, I am somewhat concerned about the impact on readability. Does it make the "Publications" section too busy? Ahasuerus 00:16, 2 May 2010 (UTC)

Looks useful. Will know better after I've used it for awhile.Kraang 02:51, 2 May 2010 (UTC)
I imagine that it will come in useful at times, but the display is, as you anticipated with your question, "too busy". The juxtaposition of author and publisher links might cause some users to think it's one link. Mhhutchins 05:01, 2 May 2010 (UTC)
Good point. I am trying to think of a way to make the distinction clear, but nothing comes to mind at the moment... Ahasuerus 10:59, 2 May 2010 (UTC)
Is the author link necessary if it's linked in the header of the title page? Mhhutchins 05:01, 2 May 2010 (UTC)
The name displayed in the "Publications" section is of the person responsible for the Publication, which is not always the same as the author of the Title. It's not very useful when displaying pubs for novels and container Titles, but it's informative when displaying short fiction, essays, etc, e.g., see the way Omnilingual appears. Ahasuerus 10:59, 2 May 2010 (UTC)
Yes, I see your point. I'd forgotten that content titles are displayed the same as publication titles. Hope we can come up with a solution. Thanks. Mhhutchins 14:48, 2 May 2010 (UTC)
I wouldn't miss the publication author links if they're the same as the title author link. In fact, I don't use the publication author links much at all, unless it saves me scrolling back to the top. BLongley 16:44, 2 May 2010 (UTC)
As to having publishers as links - I like it, so long as it encourages regularisation of publisher names in the KISS (Keep It Simple, Stupid!) way. If it goes the other way and things like "Pan, an imprint of Pan Macmillan UK" become the norm then I'm against it until/unless we separate imprint and publisher. BLongley 16:44, 2 May 2010 (UTC)

(unindent) [insert an emoticon for scratching one's head here]

How about adding "publ." so that the lines listing pubs would look something like this:

  1. Raft, (Jul 1991, Stephen Baxter, publ. Grafton, 0-246-13706-1, £14.99, 264pp, hc) Cover: Chris Moore
  2. Astounding Science Fiction, February 1957, (Feb 1957, John W. Campbell, Jr., publ. Street & Smith Publications, Inc., $0.35, 164pp, digest, magazine) Cover: Freas

where the italicized parts are links? Ahasuerus 18:57, 2 May 2010 (UTC)

Could the Publisher link just be a different color? The separator 'publ.' works, too. ~Bill, --Bluesman 16:53, 3 May 2010 (UTC)
I am not a good judge of colors, so I have added "publ." for now. Let's see how it works. Ahasuerus 05:08, 8 May 2010 (UTC)

Policy on deletion of talk page history

Is there one? See here. Mhhutchins 14:56, 2 May 2010 (UTC)

The talk page history has not been deleted. It is still available via the page history. Instead of a seperate archive page, the same thing can be accomplished via adding a static link to the history (example). However, the user has also blanked their user page as well. It may be a case that the user no longer wishes to participate and, if so, that should be respected. --JLaTondre 16:23, 2 May 2010 (UTC)
I was incorrect to say the history has been deleted. Unfortunately, the history can't be edited, comments can't be responded to, questions can't be answered. Without archiving, it remains inaccessible and unavailable as a tool for communication. I wholeheartedly respect anyone's wishes not to participate, but these user talk pages don't belong to the user. They belong to the project. Imagine anyone deleting a Wikipedia page because they no longer feel the need to participate. Mhhutchins 17:31, 2 May 2010 (UTC)
Wikipedia user & user talk pages get blanked all the time. If a user wants to leave, they can have their page deleted so even the history is gone. --JLaTondre 17:40, 2 May 2010 (UTC)
I was referring to Wikipedia articles and their talk pages, not to the user pages. Sorry. Mhhutchins 17:43, 2 May 2010 (UTC)
Unfortunately, our disk space is fairly limited, so we have to drop all but the most recent 20-100 versions of Wiki pages every year or two. If we stopped using Dave Sorgen's Talk page, it wouldn't make a difference since the last 20+ versions would still be available, but we need to figure out what to do with his page going forward. Wiki pages are pretty much the only way for editors to communicate, so if editors stop using the Wiki they way Scott Latham and Dave Sorgen have, the whole paradigm is compromised. How do we know that a record is still accurate if it has been modified by other hands and the original verifier hasn't confirmed that it still matches his copy? (Which reminds me that I need to clear the backlog on my Talk sub-page...) Ahasuerus 17:10, 2 May 2010 (UTC)
Well, that makes me think of several things:
  1. It would be a nice feature to see the history of changes to an ISFDB record. I recognize that's probably a lot of development, but being able to see who actual made what change has benefits.
  2. Whether the talk page content stays or not doesn't help with that.
  3. Hopefully, if someone is making a substantial edit to a verified publication, it is because they have a copy of the publication & will also verify it. There is nothing special to being the original verifier in terms of accuracy. You're already placing trust in them to have gotten it right. I don't see much issue with extending that trust to follow-up editors.
--JLaTondre 17:28, 2 May 2010 (UTC)

  1. Yes that would be nice, but the database and the wiki work differently. I don't write software, but I can't imagine a database that records changes.
  2. You're correct. But it does help with many other issues. I believe you've been around long enough to appreciate the value of the talk page.
  3. The protocol for primary verification is explained here. Trust is equally extended to other primary verifiers. It's not always a matter of trust. Lately it's become a matter of availability. When editors leave (or have little respect for the "paradigm") the second verification becomes very important. Mhhutchins 17:43, 2 May 2010 (UTC)
At one point Al started working on an "edit history" page, but it turned out to be quite a bit of work. It would also make the backup file much larger since it would contain all versions of Notes. Some day, perhaps. Ahasuerus 17:36, 2 May 2010 (UTC)
As far as the second issue goes, it's not uncommon for two verifiers to have slightly different versions of the same book (book club editions, unmarked reprints, etc) and a discussion is usually required to figure out whether a new record needs to be created or the current one needs to be modified. Ahasuerus 17:36, 2 May 2010 (UTC)
Dave has told me in the past "I just don't like the wiki paradigm and tend to ignore it" - but that doesn't mean he's stopped editing ISFDB itself. It might be useful to leave the talk page with a note not to bother leaving messages there - but you might want to ask him via email if he really dislikes the Wiki that much. BLongley 16:32, 2 May 2010 (UTC)
Consider this attitude "paradigm": "I will continue to edit the database, but don't bother asking me any questions about my edits or any of the records that I've marked as primarily verified. Such questions waste my time and yours because I'm not going to answer them." What a lovely attitude! I wonder how long this project would last if that were the prevailing belief? Lately, I don't like the typing "paradigm". Any suggestions for replacing that? Mhhutchins 17:31, 2 May 2010 (UTC)
I think it should have been archived, then if he wished no further intrusions then he could state it or possibly the edit function could be blocked. Still, I wrote many notes and frankly I can not remember a reply, I will bet most notes ended up like mine, "I did this, please object". Still it was/is a good lesson in politeness. In most cases the secondary verifier will be the person who did any new work on the entry. I do think his portion of the wiki should be blocked with an advisory, I know it looks bad to new editors, especially when it takes so much work to get them to write a quick note, but better a note than have them come back to 'blanked out' pages. I do consider the wiki contents the DB bases property, not mine for history purposes. Thanks, Harry. --Dragoondelight 19:57, 2 May 2010 (UTC)

Canonical names

I know many folks are suffering from rules discussion fatigue, but please weigh in on Rules_and_standards_discussions#Eastern_names if you have ideas or opinions about the handling of names whose "natural" order is not first name + last name. Thanks. --MartyD 11:35, 6 May 2010 (UTC)

Magazine grids going live

I am about to install a patch which will let us view magazines as grids. It's a little tricky, so if you see anything unusual, please post your findings here. I will explain the new functionality after I am done with the install. Thanks! Ahasuerus 04:26, 8 May 2010 (UTC)

The patch is now live. You can now view all issues of any magazine as a grid. The details are as follows:
  • The EDITOR records have to be in a series, otherwise this functionality is not available
  • If an issue has no month associated with it, then it appears in the "No Month" column on the far right, e.g. see what the grid for Authentic Science Fiction looks like at the moment. In most cases this can be fixed by correcting the month/date of publication. Some cases, e.g. quarterlies, may require a Rules/Standards discussion if we want them to display correctly.
  • The grid page is linked from each Magazine and Fanzine pub page as well as from the EDITOR Series page
  • You can change the sorting order by clicking on "Show earliest/last year first" at the top of the page
  • Unverified issues are yellow, secondary verifications are blue - since I am color-disabled, let me know if you would like me to change the color scheme
The yellow color is hard on my eyes, maybe a softer color like khaki (#F0E68C) would be better? I've been using gold (#FFD700) for manual highlighting for example Thrilling Wonder Stories.--Rkihara 07:24, 8 May 2010 (UTC)
I agree. The blue is pretty bright and somewhat obscures text, too -- it could be lighter. In another life, I have found this set of paler colors to be useful, FWIW:
--MartyD 11:13, 8 May 2010 (UTC)
Thanks, will do! Ahasuerus 15:52, 8 May 2010 (UTC)
  • Multiple issues per month are allowed, e.g. see how mid-December issues are displayed on Asimov's page
  • Nested series are supported, but there is a catch. If the parent super-series, e.g. Astounding/Analog contains only sub-series and no EDITOR records, all sub-series are linked to their respective grid pages, but there is no link next to the super-series name. (It's possible to do, but it requires a fair amount of work and my ISFDB time is still very limited.) The workaround is to display of the sub-series, which will say "Sub-series of Astounding/Analog" at the top of the page, and link to the grid for the super-series.
  • There are some minor wording/grammar inconsistencies which I will be happy to fix once I get some sleep.
Please report any bugs/issues here. Thanks! Ahasuerus 04:51, 8 May 2010 (UTC)
Another problem is with pseudonyms. See Future and the Future grid. As far as the grid month; we don't need a new standard. Help allows for the use of secondary sources for dates and the other major mag bibliographers almost always assign grid months. It would actually be nice to eventually go directly to the grid.--swfritter 15:24, 8 May 2010 (UTC)
I can't find any differences vis a vis other grids. What an I missing? Ahasuerus 15:52, 8 May 2010 (UTC)
Ah, I see what's going on. Yup, it's a bug with pseudonyms, will be fixed tonight, thanks! Ahasuerus 16:54, 8 May 2010 (UTC)
I started on the Perry Rhodan series, and so far the original series looks nice, If I insert a comma after "Perry Rhodan" for each issue, it will look even better. The Ace series however is not very accessible. Is this only because there were multiple printings or am I missing something? --Willem H. 16:30, 8 May 2010 (UTC)
A few years like 1976 and 1977 look OK if you don't mind the long titles. The rest are messy, in part because multiple printings are not really supported for magazines and there is no easy way of ignoring reprints in the grid. Another problem with the Ace series is that we are apparently missing months of publication for the first few years. Finally, 1978 and 1979 are unusual in that they were apparently publishing 4-6 per month and then nothing for long periods of time. That's not a schedule that can be readily accommodated in a year/month grid. I think the Ace reprints may be usefully reclassify as Anthologies. Ahasuerus 17:08, 8 May 2010 (UTC)
Perhaps this can be avoided if we allow separate editor series for different printings (e.g. "Perry Rhodan Ace 1st", "Perry Rhodan Ace 2nd" etc). Then each printing would have it's own grid. The resulting series can then be added as sub-series to one parent series (Perry Rhodan Magazines?) --Willem H. 18:38, 8 May 2010 (UTC)
Something else I find unsightly is UK issues of Astounding merged with American issues. Same with If. That doesn't make a lot of sense to me but apparently it does to others. Use of the auto-grid actually makes things a bit less confusing.--swfritter 17:11, 8 May 2010 (UTC)

(unindent)A patch to fix various issues has been deployed:

  • Colors have been changed to gold and light blue as per the codes suggested above.
  • A link from the Grid view to the Series view has been added.
  • Minor wording changes.
  • Pseudonymous editors, e.g. Future as a series vs. Future as a grid, should be OK now.

Ahasuerus 23:31, 8 May 2010 (UTC)

Great enhancement! I've already discovered several issues of magazines that had the wrong or missing month in the date field. The grids made it easy to spot them. Also I saw several issues that I entered and forgot to verify, easily spotted in the grid. I did discover an anomaly. Look at the grid for Worlds of Fantasy. The links for the September 1968 and September 1970 issues are missing...BUT the fields are blank where months without an issue have the dash. Mhhutchins 05:31, 9 May 2010 (UTC)
That's an interesting point. The problem was that the two "missing" issues had, according to our rules, incorrectly formed titles. The way the code determines what to display in the grid for each issue is as follows:
  • Retrieve the full title of the pub
  • If there is a comma in the title, strip everything up to the first comma
  • If the last 4 characters of the title are the same as the pub's year, delete them
  • Remove leading and trailing commas and periods
The offending "Worlds of Fantasy" issues had titles like "Worlds of Fantasy #2, 1970", so steps 2 and 3 deleted everything from the title. The resulting empty string was not a valid value for an HTML link, which is why the issues didn't appear in the grid. I have changed the titles to "Worlds of Fantasy, #2, 1970" etc and they looks much better now. I suppose I need to make the code smart enough so that if the stripping process results in an empty string, it will use the original title instead.
Also, we have a single magazine, The American Museum of Science, Literature, and the Arts, which contains commas in the EDITOR title, which in turn slightly messes up the grid view. It's possible to have it fixed by making the algorithm more sophisticated, but I am not sure it's worth introducing extra complexity as long as only one issue is affected. Ahasuerus 06:01, 9 May 2010 (UTC)
Checking the backups, I see that we have another 90 issues that have the same problem that "Worlds of Fantasy" had. No big deal, I can can change them tomorrow. Ahasuerus 06:22, 9 May 2010 (UTC)
Spectacular! I really like the effect. It really breaks down the publications visually. Foresee researchers loving this. Thanks, Harry. --Dragoondelight 11:21, 9 May 2010 (UTC)
Appreciate the kind words! I have changed the code to restore the original title if the stripping results in a blank title -- see, e.g. Wonder Story Annual. Ahasuerus 16:36, 9 May 2010 (UTC)
I figured the missing links for the issues of Worlds of Fantasy had something to do with the punctuation. I discovered a couple more missing issues of other titles that had a period instead of a comma. Again the grid helped to locate those missing links and it was easy to correct the titles. And, voila!, the missing issues were now linked to the grid. Again, love this feature, not only for linking issues to one source but it's ability to find errors. Mhhutchins 16:53, 9 May 2010 (UTC)

(unindent)It would appear that the major issues have been addressed. Let's give it a few days and if everything looks OK, we can start migrating the grids from the Wiki to the database. Once Series Notes have been added, we should be able to move the rest of the Wiki-based magazine pages to the database, although we may need better support for Notes to handle some of the more extensive/hyperlinked pages. Ahasuerus 19:22, 9 May 2010 (UTC)

A search for nice girls found 0 matches

Genuine ISFDB search result, caused by Fixer submissions. Or the reason I'm still editing here rather than coping with real life? :-/ BLongley 23:20, 8 May 2010 (UTC)

That's because you ran a "Fiction Titles" search -- try a more comprehensive search and you'll see! Besides, you don't need "nice girls", try "nice girl" and the search results will be even better! :-) Ahasuerus 23:34, 8 May 2010 (UTC)
A search for "bad girls" will produce more results. Others may want to search for "bad boys" which will also bring results.--swfritter 14:56, 9 May 2010 (UTC)

Opinions on the Perry Rhodan series wanted

I started with the first 50 issues of the original German Perry Rhodan series, see the grid (very nice feature). Number 50 marks the start of the second cycle of the series, see here. I would like some opinions on how to proceed, or, is this the way we want the series entered. My main questions are:

1. The grid will probably look very nice, even after the first 50 years are entered, but will look like hell if the 2nd to 5th (or 6th) printings are entered in the same magazine series. My suggestion would be to give each printing its own series. Does that pose any problems?
2. I'm turning all titles into novellas, and delete the (complete novel) from the titles entered as serials (compare old with new. Does anyone have a problem with this?
3. Is the setup of the series here ok? I use the German titles for the cycles. Eventually the parent series (Perry Rhodan Universe) will have several children (at least "Perry Rhodan Weekly", "Atlan" and "Perry Rhodan Novels", all had titles translated in English), and it will be a very very long list (over 2500 titles in the main series alone). Has there ever been a feature request to be able to fold and unfold sub-series (like the tree structure in Windows Explorer)?
4. Does anyone have a problem with treating the British series the same as the Ace series (turn the publications into a magazine, and the titles to novella's). So far I've left the Orbit/Futura editions as they were, but they don't look right.
5. I could use some help with the series wiki page. Are there rules, how do you set up such a page, etc.

Opinions and help would be appreciated. --Willem H. 21:26, 10 May 2010 (UTC)

I'm not a reader of PR and likely never will be, so my opinions are from a purely bibliographic POV.
  1. I don't like the idea of putting reprints into another series, just from a slippery slope perspective.
  2. Turning the titles into novellas is fine, if it's determined that they're less than 40K words, BUT you should still follow the current ISFDB standard when a novel is published in a magazine. I'm assuming here that you maintain that these are magazines and not books.
  3. There have been discussions about the display of series, but I think your plan (which I really like) would take someone with heavy software writing skills. Anyone up to the task?
  4. I think someone awhile back was talking about splitting the British and Ace series. The British series was simply the novellas/short novels, while the Ace series was a magazine with essays, columns, etc. Maybe that's why they're treated differently? (And probably should be.)
  5. There really are no rules about what can be added to a series page. Just add anything that is important, but can't be easily added to the database records. Discussions about the page should be handled on the page's talk page. The talk page can also be used to discuss the series. Your last line: if you're asking how to create a series page, it's simple. Just type into your browser's address window followed by the name you want to give the series. If one already exists with that name you're led to it. If not, you're asked whether you want to create the page. Mhhutchins 22:55, 10 May 2010 (UTC)
  1. Our software doesn't handle magazine reprints well, so we have to be careful there.
  2. As we discussed a few weeks ago, all early PR "novels" were actually novellas, so it's fine to enter them that way. I am not sure if they got much longer in the 80s/90s, but some those may be "(Complete Novels)"
  3. There are at least three problems with very long series. First, generating a complete list of PR issues may put a significant load on the server, making the user wait for the data to appear and possibly affecting performance for other users. The second problem is how long it will take for the browser to render the data on the screen, especially when there are nested tables (when there are multiple issues per month), which tend to slow down even modern browsers. The third problem is navigating around a page with 2,500 entries, which is likely to be painful. Adding a "Windows Explorer"-like solution would only help with the third problem and even then we would have to use Javascript, something that we have been reluctant to do in the past except for editing. A non-Javascript solution would be to add an option to "collapse sub-series" and, conversely, "expand sub-series" at the top of the page. We could also make the code smart enough to automatically go in the "collapse" mode if there are more than 10 sub-series. Something like Forgotten Realms Universe, with its 60 sub-series, could really benefit from it. It wouldn't be that hard to implement.
  4. The British series is a bit tricky. It started out as an omnibus series, then switched to the regular novel format, but some volumes added a story or two. Messy. Ahasuerus 01:45, 11 May 2010 (UTC)

Comments by Terry Kemp on Database CD

Reply from Terry Kemp regarding our questions about the database CD. Posted in entirety.--Rkihara 23:32, 11 May 2010 (UTC)

This is a fairly long note, so responses should be posted in a new section.--Rkihara 23:53, 11 May 2010 (UTC)

Hi Ron,

Thank you for the note, and the links.

To begin, I'm deeply impressed with the work that all of you involved in ISFDB are doing. The number of experts and variety of expertise, all in one place, is indeed impressive. After working alone on my project for so long, I'm delighted to receive such varied feedback.

First, let me thank each of you who have purchased a copy. That being said, I hope you all have read the instructions that came with the CD. Among other things, the instructions explain the time delay on close, as the database compacts after use. I have tried to be very careful and very clear in regards to using this application. The CD contains what are essentially two versions of the database. One is very limited, but can be used as a read-only application directly from the CD. The other needs to be "installed."

Many of you have touched directly on my immediate concern regarding this application, the fact that MS Access is constantly being updated. I wanted to present this application to those that might be interested before yet another (costly) MS Access update occurred.

Let me take a moment to address some concerns and make some informal comments. I apologize in advance if my comments appear either jejune or trivial.

(Willem H.) It is my understanding that my MS Access application does require having MS Access. It is, sadly, not self-running. This does limit the potential user base...for now. I have been looking into other formats, such as FileMaker for the Apple/Mac computer platform.

(Ahasuerus) A short while ago, someone suggested making the database available in OpenOffice. I do not know much about OpenOffice, but will be looking into it now to see if it can read MS Access without modification, or if a distinct version is required. I suspect that a distinct version will be required.

(Dragoondelight) Thank you for your insight. This is exactly the type of input I was hoping to receive. I hope that some of my comments will help make the database more useful.

In order to accommodate as many different potential users as possible I did a few basic things.

I included a read-only version for use directly off the CD. This was meant for quick searches and easy viewing. No download times involved.

For the more flexible, installed version, I locked down the application in every basic way that is offered by the program. For instance, I turned off the right-click mouse button, I also turned off everything in the background, and locked the forms as well. My intent was to limit access to the raw data in order to prevent inadvertent deletion or corruption. MS Access is unrelenting, changes in data even by an accidental keystroke are immediate and permanent.

All this "security" can be easily defeated, and the data opened to more complex uses. Open a blank database, and import all the tables into the new database. This will allow a "spreadsheet" type view of the data, where not only all the contents of a single issue are revealed, but also all the contents of a series title. In fact, all the data is open to use and to view at a glance.

At this point, needing more feedback, I wasn't certain that the few forms and few reports I created would continue to be useful. My thinking was that the user, capturing the raw data, would create forms and reports to suit their own needs, no longer needing my generic ones.

My idea behind the database was not to replace either Miller/Contento or ISFDB, but to create a different and equally useful tool. Once you see the raw data, you can create unique relationships between the fields. Using the data in its table form, I think, is as useful as either Miller/Contento or ISFDB, because in this table form it appears as though in a traditional written page format.

A relational database allows a different perspective of the essential data. For instance, lists/reports can be made containing all of the work by Jack Williamson and Rog Phillips. This is something that static HTML does not allow. Accurate counts of all issues under a single series title can be done as well. This is of some small note when dealing with a series title which has developed an unusual internal numbering system.

Finally, as a relational database, in table format, the user can view all the data available. There are no download times, no going from one HTML link to another. It is all there, visible simultaneously. And the database can be added to with ease. New fields can be created to contain any data desired. The possibilities are endless. Unique comments can be added to each record. Links to other data tables can be created as I did linking various stories to their first specialty publication. A unique field can be added in order to describe the contents more precisely, like as "science fiction" or even "about Mars." Checklists, price idea that I've had recently is to add pay rates for contemporary magazines.

From the viewpoint of an indexer, the data is missing some detail. When I began this project, I was initially only interested in listing the fiction. This is still my primary focus. It has been a topic of some concern. Listing letters (without content), just as "letter by author" has some merit, but like listing "editorial by unknown" it is limited. (So the topic of the editorial and the author are both really unknown. This information blurs into the hundreds of other "editorial by unknown" from various other magazines. What does this really tell the user about a magazine? That it contains an unknown editorial by an unknown author?) Again, I was not trying to duplicate the work that your group is doing, but to create a different tool. That being said, I've begun to capture more of this incidental, albeit equally important, information, and adding it to my database.

But there will always remain a gap between listing "letters by unknown as a column" and "letters by author" and the usefulness of either is in the eyes of the user.

(swfritter) Regarding canonical names: Gosh, what a blunder on my part. I went through a lot of trouble with this feature, and still made a gaff. My database does in fact associate all canonical names. For instance a report on Henry Kuttner does list all stories under his real name, and pseudonyms. My mistake is in the display of the generic report. Looking at the raw data, I see all the listings, what my report does not (yet) do is display those alternate works along with the pseudonym. So, the report does contain "The Mad Virus" by Paul Edmonds (Science Fiction) and "We Shall Come Back" by C. H. Liddell (Science Fiction Quarterly). What it does not do is display this alternate identity (list the pseudonym) when it lists the story (which it does) under Henry Kuttner. Thanks for spotting this easy fix. The problem is not in the data, but in the report display.

Again, thanks to all of you for your interest, and for the opportunity to respond.

Well, I sure have gone on, so I'll stop here. I hope I have addressed the points raised and made the database more useful. Cheers, Earl Terry Kemp

P.S. Ron, don't hesitate to delete any or all of the above if you do post my comments.

Terry Kemp CD

Thanks Terry. I might note that although the ISFDB superficially appears to be just an HTML database, the data is actually dynamically generated from a relational database. This page has links to the weekly MySQL backups. And this page has instructions on how to set up MySQL. I generally use MySQL Query browser to run the SQL scripts but also wrote a clumsy program in Borland Delphi for generating personal reading lists. As far as level of detail: a lot of techies here so we probably do get a little carried away. Our detailed coverage of major prozines is one of our strengths; one of our weaknesses is coverage of contemporary and minor periodicals which appears to be one of your strengths. As with any software, especially new software, it takes a little while for the user to figure how get the devil out the details. Thanks for making your data available.--swfritter 00:46, 12 May 2010 (UTC)

Amazon images on pre-publication records.

I've discovered lately that many records that were generated from pre-publication records on Amazon have broken links for images. It seems that once the book is published the image is replaced by a new "Look Inside" generated image. The previous image files are deleted from Amazon, breaking the ISFDB's links. Any ideas about how to solve this problem? Mhhutchins 22:15, 12 May 2010 (UTC)

Samples: here, here, here, here, here, here, here, here and these are just the ones I've run across today. Mhhutchins 00:20, 13 May 2010 (UTC)
It's a good question. The only way to find replaced/upgraded Amazon images (that I can't think of) is to do the following:
  • Periodically compile a list of all pubs which use Amazon images
  • For each pub, request the current image URL from Amazon
  • Compare the URL that we have on file with the URL returned by Amazon
It's not hard to implement and should only take a few of hours to run at night (given the way Amazon throttles requests and other obscure technical issues), but then we have to decide what we will do about the discrepancies. Should Fixer create an automated submission updating the image URL whenever Amazon reports a different one (and let moderators sort it out)? It may be viable, but Amazon has been known to surprise us. Perhaps we should run this process for the last 6 months worth of books and see where it gets us... Ahasuerus 04:22, 13 May 2010 (UTC)
Sounds like a plan. I personally place images low on the priority list, but hate to see those broken links. Another thing I discovered is that Fixer and/or Dissembler was picking up the ISBN-10 for pubs published in the past few years instead of the ISBN-13 which actually appears in the books. No big problem as far as pub records go since both ISBN-10 and ISBN-13 are displayed, but searches for ISBN and publishers give the wrong ISBN. I've noticed the last few months that the bots are picking up the ISBN-13, but is there a way to find those pubs that were entered with ISBN-10 and make automatic changes without going through the submission process? Thanks. Mhhutchins 04:37, 13 May 2010 (UTC)
The ISBN-10-to-ISBN-13 transition happened in 2007, so I suppose we could use the following simple logic:
  • Find all unverified books published in 2008-2010
  • If the pub has a valid ISBN-10, then replace it with its 13 digit cousin.
I am not sure how many IBSN-10s appeared in 2007, so it may be safer to leave them alone.
Going forward, Fixer uses ISBN-13s for all 2008+ books, so we should be in good shape. I seem to recall that Dissembler, which is currently inactive, was upgraded in 2008, some months after the transition, which may account for some of the discrepancies. Ahasuerus 04:56, 13 May 2010 (UTC)

Fixer findings

One of the things that Fixer does in his plentiful spare time is calling an OCLC service to look for linked SF ISBNs (reprints, other editions, etc) and then check which ones we are missing. Examining the roughly 40,000 ISBNs correlated so far, we find:

  • 15,500 ISBNs which we have on file (including 158 foreign language ones)
  • 24,000 ISBNs which we do not have, including:
    • 9,500 English language non-audio ISBNs
    • 5,200 English language audio/digital ISBNs
    • 9,200 foreign language ISBNs

Since this OCLC service is free, it doesn't provide much information about the missing ISBNs, but now that we know what we are missing, we can use other sources to get publication level data. We'll see where we are once Fixer finishes his analysis... Ahasuerus 04:10, 13 May 2010 (UTC)

Some 10 days later:

  • Already in ISFDB: 30,210 including 5,497 foreign ISBNs
  • Not in ISFDB: 74,242 including:
    • English non-audio: 13.932
    • Foreign language: 45,291
    • Audio/digital: 15,019

Since we have 127,000 ISBNs on file, we are almost 25% there. Once the rest of our ISBNs have correlated in another 6-ish weeks, we can start querying other sources (like Amazon and LOC) and creating submissions. Ahasuerus 05:09, 24 May 2010 (UTC)

Bibliographic Comments

I noticed that the link to the Bibliographic Comments page on the pub records page has been changed to "Edit existing Publication comment". I feel this is incorrect and should be changed to "View existing Publication comment" as the link leads to the bibliographic comment page. After that the user can choose to edit the page or not. Anyone feel the same or differently? Mhhutchins 19:16, 13 May 2010 (UTC)

I agree completely, and the same for "Edit existing Series comment". --Willem H. 19:42, 13 May 2010 (UTC)
Makes sense, I'll add it to the list. Ahasuerus 00:46, 14 May 2010 (UTC)
How about "View/Edit" since once you follow the link, you can both view and edit the data? Ahasuerus 19:09, 14 May 2010 (UTC)
It's been changed to "View/Edit", let's see if it looks OK. Casca is a good series example and we have any number of pubs with bibliographic comments. Ahasuerus 03:16, 15 May 2010 (UTC)
Looks better this way. Quick action! Thanks, --Willem H. 18:31, 15 May 2010 (UTC)

Unnumbered titles in a series

Now that the software has been stable for some time, I wonder if the way we currently display Titles in a Series is optimal. At the moment, all unnumbered titles are displayed first, followed by the numbered titles, e.g. see how Miles Vorkosigan appears. The decision to go with this display order was based on very limited user feedback in 2006, before we even started beta testing. I wonder if it would make more sense to display the numbered titles first instead? Ahasuerus 01:05, 17 May 2010 (UTC)

It seems the most desirable presentation is date ordering with numbered items as an uninterrupted, contiguous block. What about using the date of the unnumbered title relative to the date of the last numbered item? If earlier, put the title before the numbered items, if later, put it after them. This would probably capture the most common cases of short works written before numbered novels and after-the-fact collections of numbered stories. --MartyD 01:34, 17 May 2010 (UTC)
When I applied this approach to Miles Vorkosigan, I got the following peculiar mix:
   * The Borders of Infinity (1987) [SF] by Lois McMaster Bujold
   * The Mountains of Mourning (1989) [SF] by Lois McMaster Bujold
   * Labyrinth (1989) [SF] by Lois McMaster Bujold
   * Borders of Infinity (1989) [SF] by Lois McMaster Bujold
   * Weatherman (1990) [SF] by Lois McMaster Bujold
   * Vorkosigan's Game (1990) [O/2,5] by Lois McMaster Bujold
   * Memory (Excerpt) (1996) [SF] by Lois McMaster Bujold
   * Young Miles (1997) [O/1,2] by Lois McMaster Bujold
   * Miles, Mystery & Mayhem (2001) [O/7,3] by Lois McMaster Bujold
   * Miles Errant (2002) [O/4,6] by Lois McMaster Bujold
   * 1 The Warrior's Apprentice (1986) by Lois McMaster Bujold
   * 2 The Vor Game (1990) by Lois McMaster Bujold
   * 3 Ethan of Athos (1986) by Lois McMaster Bujold
   * 4 Brothers in Arms (1989) by Lois McMaster Bujold
   * 5 Borders of Infinity (1989) [C] by Lois McMaster Bujold
   * 6 Mirror Dance (1994) by Lois McMaster Bujold
   * 7 Cetaganda (1996) by Lois McMaster Bujold also appeared as:
         o Magazine Appearances:
         o Cetaganda (Part 1 of 4) (1995)
         o Cetaganda (Part 2 of 4) (1995)
         o Cetaganda (Part 3 of 4) (1995)
         o Cetaganda (Part 4 of 4) (1995) 
   * 8 Memory (1996) by Lois McMaster Bujold
   * 9 Komarr (1998) by Lois McMaster Bujold
   * 10 A Civil Campaign (1999) by Lois McMaster Bujold
   * 11 Diplomatic Immunity (2002) by Lois McMaster Bujold 
   * Winterfair Gifts (2004) [SF] by Lois McMaster Bujold
   * Miles in Love (2008) [O/9,10] by Lois McMaster Bujold
Part of the problem appears to be that omnibuses are probably best displayed at the end regardless of when they were published. I also suspect that minor short fiction published in the middle of the series would look better at the end of the list rather than at the beginning. Ahasuerus 02:17, 17 May 2010 (UTC)
Compare against the first numbered item instead of against the last, then. --MartyD 10:18, 17 May 2010 (UTC)
Sounds like the best solution I think. It would certainly make series like Discworld, Wild Cards and Amber look better. --Willem H. 12:18, 17 May 2010 (UTC)
We can certainly give it a try and see how it goes. Ahasuerus 18:43, 17 May 2010 (UTC)
However, it leads to the next question of course. Can sub-series be sorted on the date of the first item in that sub-series? --Willem H. 12:18, 17 May 2010 (UTC)
It's possible (although not trivial), but the solution that was discussed last year and that I think would be more comprehensive is to add a new field, "Series order", to the Series record. That way we could make prequels etc appear any which way we want. It will also be a good proof of concept for adding new fields to existing tables, something that we haven't done since Al became inactive about 2 years ago. Ahasuerus 18:43, 17 May 2010 (UTC)
I'm game, if you're looking for volunteers. It shouldn't be particularly difficult. --MartyD 00:33, 18 May 2010 (UTC)
Excellent! :) As you said, it shouldn't be particularly difficult, but you may want to take a look at the database modification scripts that we used in the past just to make sure that we are more or less consistent. Ahasuerus 00:55, 18 May 2010 (UTC)
A couple more thoughts:
  • We'll need to update the Web API to handle the new data element
  • The Edit Series logic is more complicated than it may look at first glance since it has to process various permutation of "Add Superseries", "Remove from Superseries", "Change Series name", "Move to another Superseries", etc.
If Edit Series becomes too complex, we could always split it into a number of separate Web pages. Ahasuerus 16:54, 18 May 2010 (UTC)

Author Directory

Is the author directory a new thing or something I had not noticed before? That's a cool user interface. I don't think I've seen a "directory" formatted like that before though it makes perfect sense.

Al added it about 3 years ago, but I don't think it has been used much. Ahasuerus 12:24, 26 May 2010 (UTC)
I see what happened. It's only in the left navigation bar when you are on the home page and not part of the standard "Other Pages." I'm almost never on the home page. I'd need to think about if we want this in the "Other Pages" for all pages. It would help increase usage. It also turns out the display is hard to read in Internet Explorer (at least in IE version 6, see bug report) plus added a feature request (see below) that could increase usage. --Marc Kupper|talk 21:54, 26 May 2010 (UTC)
Bug confirmed under IE 8.0. Ahasuerus 05:29, 27 May 2010 (UTC)
Bug fixed. Ahasuerus 05:52, 27 May 2010 (UTC)

I happened to find it as a Google search included Something that would be useful on a search results page is a link back to the main directory. --Marc Kupper|talk 04:12, 26 May 2010 (UTC)

Just a generic link from any and all search results pages? Ahasuerus 12:24, 26 May 2010 (UTC)
Yes, back to I submitted a feature request in case you are bored. :-) --Marc Kupper|talk 21:54, 26 May 2010 (UTC)
Oh, you mean add links between different directory pages! Yes, I can see how it would be useful, but I am hesitant to add explicit support for single letters. We have over 6,000 authors whose names start with a "B" or an "S", which can affect other users' performance. Also, we may want to add the same "Pseudonym"/"Has pseud. titles" logic that we use in regular Author searches. Ahasuerus 05:52, 27 May 2010 (UTC)

Cover Scans

This seems to be a good place to pick up cover scans if you don't do your own: Just do a search for the title and when the page for the specific book is found, click the 2nd link under the image that says covers. It will bring up all covers (if any) that were scanned by members of this site. Underneath the members scans will be the Amazon scans. --Astromath 18:27, 26 May 2010 (UTC)

Just checked it out. The ones I saw were somewhat small for our purposes. Or is there a way to get a larger image? Thanks. Mhhutchins 19:02, 26 May 2010 (UTC)
I asked a while back about getting larger images and the answer was "no." A bit annoying as they tease you with things like "high quality 626x855" stamped on a cover.
I had wanted a particular cover scan of a next-to-impossible to locate publication and struggled for a while to see if LibraryThing would serve up something better than a thumbnail. It turns out if you figure out which user uploaded a cover (not easy but it can be done) you can look at their image library and zoom in on covers. Even there, they only show thumbnailish covers despite the fact that the user uploaded a high-res scan.
I also signed up for and tried the developer API but that does not work either. Hence, I gave up on using that site. --Marc Kupper|talk 22:25, 26 May 2010 (UTC)

Should transient verification count as "verified?"

At present when someone uses "Primary (Transient)" verification this is not displayed as "[Verified]" in title displays. Help:How to verify data says

Primary (Transient) Verification
This is used just like Primary Verification, and should meet the same standards. The difference is that a Primary Verifier is generally assumed to have possession of or ready access to the publication in question, and is prepared to respond to queries about it. Primary (Transient) Verification is for situations where the verifier is likely to be unable to respond to queries -- for example if the book is borrowed or is scheduled to be sold or discarded after the verification is complete.

Per "should meet the same standards" it seems we'd be safe in having these publications show as "[Verified]."

I'd like to put this in as a feature request but first thought I'd ask as it's possible editors are using a lower standard for transient verification. --Marc Kupper|talk 22:06, 26 May 2010 (UTC)

Before "Primary2," etc., came along some of us used it to indicate a second verification. I'm still using it that way when I recheck some of my verifications. I've also used it instead of Primary when I've had a publication that was missing pages.--Rkihara 02:18, 27 May 2010 (UTC)
I think it could usefully be displayed as Verified on that display. On the other hand, the number of Verifications shown to Moderators when approving edits is probably too long, and I find Secondary-source verifications distracting when all I'm really checking is whether Primary Verifiers have been notified. On the third-hand, we have no easy way of showing when a publication has been verified against a review, which is something I'd quite like as we don't link reviews to publications, only titles. It would be a quick way of indicating a fairly good reason for a pub record to be here, without having to leave notes in the pubs or setting up every Magazine as a new Verification source. BLongley 23:01, 27 May 2010 (UTC)
Sounds like a feature request to me! :) Ahasuerus 00:08, 28 May 2010 (UTC)
Well, most of that could be accomplished by just adding "Review" as a Verification Source, but it might be nice to change the code for that source so that an unverified but reviewed pub showed as "reviewed" rather than "verified". But if we do that we might want to separate what Primary and Secondary Verifications should show as. Primary trumps Secondary I guess, but are there some Secondaries that are better than others? BLongley 12:06, 28 May 2010 (UTC)

Title page changes

A couple of minor changes to the Title page:

  • 0000-00-00 publications now appear at the end of the pub list rather than at the top
  • when displaying anthologies, magazines and fanzines, "ed." is displayed before the editor's name so that users would not think that editors are responsible for others' works

Also, do we want to display the "Reviews" section before the pub list? Books like The Hobbit have a lot of pubs, pushing reviews so far down that they become almost invisible. Ahasuerus 04:55, 28 May 2010 (UTC)

Wonderful. I've been hoping for sometime that the undated pubs be placed at the end. And the "ed." is also a good clarification of the role. About moving the reviews to the top, in most cases that would work well, just as the awards and variants are now listed above the publications. I wonder how it would look if there were a large number of reviews. Sometimes when there are a large number of awards (especially in the case of the Prometheus Awards when a title has been nominated several times) it can be distracting. Mhhutchins 14:16, 28 May 2010 (UTC)
Could there be an algorithm be developed that basically puts the reviews and awards at the top only if the number of reviews and awards is less than the number of pubs in the listing, otherwise put them at the end? --Astromath 15:49, 28 May 2010 (UTC)
It's certainly possible, but it would introduce a display inconsistency that may be difficult for users to follow. Ahasuerus 13:48, 29 May 2010 (UTC)
I second the "wonderful". As for reviews, I personally am not interested in them (or awards) and am happy to have them at the bottom. That said, I'm sure other people are very interested in them. Perhaps a good compromise would be to move them to the top AND provide a preference to hide them. --MartyD 16:41, 28 May 2010 (UTC)
A new User Preference to suppress reviews wouldn't be hard to implement. Feature Request 3008957 created. Ahasuerus 13:48, 29 May 2010 (UTC)
I'll third the wonderful. I like both changes. I'd leave reviews where they are, but don't have a strong preference. --JLaTondre 17:53, 28 May 2010 (UTC)

Reviews are now sorted chronologically

Reviews are now sorted chronologically on the title page. Reviews with "unknown" dates now appear at the bottom of the list, e.g. note the location of Ken MacLeod's review of Paul J. McAuley 's Whole Wide World.

A couple of related issues:

1. It turns out that the Title page currently displays the first (and only the first) appearance of each review. In most cases it doesn't matter since reviews are rarely reprinted, but some reviews see multiple publications, e.g. Damon Knight's In Search of Wonder. Do we want to show all publications of each review on the reviewed Title's page? I am not sure it's necessary since you can get this information from the Review's Title page, but it's worth considering.

2. If you are not a developer, you can stop reading here. If you are and were missing the auto-vivification feature found in Perl and elsewhere, please note that I have added a class to auto-vivify nested Python dictionaries. The "AutoVivification" class is available in common/, our main library, which is used by almost all modules, so you should be able to use it directly. See biblio/ and biblio/ for examples. I find that it makes nested dictionaries much more viable.

Ahasuerus 16:12, 29 May 2010 (UTC)

I see that the undated review is still at the top of the review section on Ken MacLeod's page, is that going to be changed? BLongley 19:04, 29 May 2010 (UTC)
That's a separate problem with the Summary Bibliography page as opposed to the Title page. The Summary page lists all 0000-00-00 titles first within each respective title type, e.g. see John F. Rossmann's bibliography [edit: no longer a good example since the biblio has been cleaned up]. If we decide to change this behavior, we will presumably want to change it for all title types to keep things consistent. Ahasuerus 19:46, 29 May 2010 (UTC)
Probably ought to be consistent. Does Unknown come after Unpublished? I can't think of an example display page to check. BLongley 19:58, 29 May 2010 (UTC)
Sure, consistency is a good thing, it's just that the Summary page is co convoluted that changing it is much more difficult. Given all the other problems that we have with it, it may be due for a rewrite anyway -- I'll poke around later today. Ahasuerus 13:53, 31 May 2010 (UTC)
Incidentally, while linking a mass of Vector reviews, I came across the Hemingway Hoax Novella and Novel. I think a lot of the reviews have auto-attached themselves to the wrong one. As has the Novel award entry. I appreciate that the Awards Editor is a long way off, but perhaps this is a sign that auto-linking of reviews is not working well? I find I'm fixing a lot of auto-links to short stories that have the same name as a collection as well. BLongley 19:08, 29 May 2010 (UTC)
A resounding "Yes!" to the question asked in #1 above. I'd like to see the reprints of reviews also listed. None of the titles which are reviewed in this book are currently linked back to this collection of reviews. Mhhutchins 19:26, 29 May 2010 (UTC)
I have made the change on the development server, but I am not entirely sure the result is what we want -- see the way Damon Knight's review of War With the the Newts and Michael Bishop's review of The Fifth Head of Cerberus appear. Is there really value to showing multiple printings of the same review? Ahasuerus 13:53, 31 May 2010 (UTC)
I see what you mean about reprints of the same publication. That wouldn't matter if the review appears in a magazine, because they're rarely reprinted. I see no value in listing reprints (which is not to say all appearances.) Is there a way to list only one printing of a publication? Mhhutchins 22:17, 31 May 2010 (UTC)
We could identify the "main" Title in each pub that the review appeared in and only display the first pub for each title. It's not all that hard to do, but are we sure that this is what we want to happen? If a review first appeared in a Magazine and was later reprinted in a Collection or a Non-fiction book, we would be showing it twice without any indication that it is actually the same review. Wouldn't it be somewhat misleading? Ahasuerus 04:39, 1 June 2010 (UTC)
I assume that reviews of a single title by the same reviewer published in different pubs are essentially (but not necessarily) the same text. But I've learned that's not always the case. I've seen several reviewers who "farm" their review of the same title to different pubs, without knowing how or even if they differ. And I've seen entirely different reviews by the same reviewer. I see nothing misleading about listing all printings of a review. I also can't see how you could differentiate between different reviews by the same reviewer of the same title. There should be no assumption that in a listing of reviews all of them are different. On the same hand, one can't assume reviews by the same author are different just because they appeared in different pubs. It's just a listing of sources for reviews of the specified title. Without listing all of the sources you're limiting the user's access to information. Maybe he doesn't have a copy of Future, June 1954 in which Damon Knight reviewed The Heads of Cerberus, but he's got access to a library copy of In Search of Wonder. Looking on the title record for the novel, you'd never have known that. If it's been verified that the reviews are the same text, they would have already been merged and the review record will show all printings of the review. Mhhutchins 23:36, 1 June 2010 (UTC)

(unindent)How about this approach: If a review has appeared only once, display it the way it currently appears. If a review has been reprinted, add "also reprinted in:" to the end of the line and then display all other pubs with the same review on the following line, properly indented and chronologically sorted? Ahasuerus 00:01, 2 June 2010 (UTC)

I like it (not being a programmer, it never occurred to me that the display could be thus manipulated.) Of course, this would only happen when the review records have been merged. The display would remain the same if it's not known if the two reviews are identical (as Paul Di Filippo's reviews here.) Thanks. Mhhutchins 00:29, 2 June 2010 (UTC)
OK, FR 3010193 created! Ahasuerus 02:23, 2 June 2010 (UTC)
I made the changes on the development server, but then I found another issue. Take a look at In Search of Wonder, which was reviewed almost 20 times between 1956 and 1998. See how the displayed year goes from 1956 to 1957 to 1956 to 1958? And how the 1952 "William Atheling, Jr." review appears between 1969 and 1979? That's because at the moment the displayed date is the date of the Review title, but the lines are sorted by first publication date. Whenever there is a discrepancy between these two dates -- either because we have an error in our data or because we are missing the pub where the review first appeared -- the sorting will be out of whack.
The underlying problem is that we are mixing and matching Titles and Pubs in this section of the page, and that's always recipe for trouble. To quote Christopher Anvil, there is got to be a better way... Ahasuerus 04:54, 3 June 2010 (UTC)
Another question is why are the reviews not dated the same as the publication in which they were published? Mhhutchins 05:31, 3 June 2010 (UTC)
The code currently deployed on the production server uses Title dates for sorting and display purposes, so as long as the Title date and the first publication date are not the same, we will have a discrepancy.
Granted, we can change the code to use the first publication date throughout, but if the Title date of the "Atheling" review is indeed 1952 and the first publication that we have on file is from 1974, then one could argue that we should be showing something like:
  • <link>Review</link> by James Blish (as by William Atheling, Jr), 1952, in:
    • The Issue at Hand: Studies in Contemporary Magazine Science Fiction (Feb 1974)
    • The Issue at Hand: Studies in Contemporary Magazine Science Fiction (Jul 1974)
For regular single-shot reviews the display will change accordingly, e.g.:
  • <link>Review</link> by Thomas A. Easton (as by Tom Easton), 1997, in Analog Science Fiction and Fact, September 1997
Also, I just noticed that we say "book review" in the Content section of the Publication page. We may want to change the wording to simply "review" since many of them review short fiction. Ahasuerus 16:21, 3 June 2010 (UTC)

(unindent) A hybrid approach to reviews has been implemented -- see the way In Search of Wonder now appears. The underlying review record is now linked directly and the sorting is by review date first and publication date second, with reprints appearing when warranted. Also, "book review" has been changed to "review" in the Content section of the Publication page. Pseudonymous reviews still need to be addressed.

I am somewhat concerned that the new approach makes the same review date appear multiple times for the most common "review date = publication date" case. We may want to suppress one of them when they are identical. Ahasuerus 01:26, 6 June 2010 (UTC)

Looks good, and I have the same concern about the dates. I would suggest that the publication date be suppressed if it matches the review date. In most cases, the only time the publication date is displayed would be for reprints in a later year. Mhhutchins 05:34, 6 June 2010 (UTC)
Done! Ahasuerus 13:43, 6 June 2010 (UTC)
Looks even better. Hope the nitpicking on my part doesn't add any extra burden to the programming. But, if you're like me, you probably enjoy the challenge. Thanks for the effort. Mhhutchins 15:38, 6 June 2010 (UTC)
Most bibliographers are closet perfectionists (why else would we be here?), so I am always torn between making the screen that I am working on as good as possible and moving on to the next bug/FR which is crying for help. To re-quote Voltaire, "The perfect is the enemy of the good" and striking that balance can be tricky, especially given our limited programming resources. OTOH, if everything goes according to the plan, I should have a lot more time to work on ISFDB in another 7-8 months or so... Ahasuerus 18:43, 6 June 2010 (UTC)

Author directory changes

  • Note that "Has pseud. titles" is almost always indicative of missing Variant Titles except in some ghost writing cases like Andrew Neiderman. Directory pages are a good way to look for them as well as for unidentified pseudonyms.
  • Single character lists, which in the past could be accessed by manipulating the URL, have been blocked since they put a significant load on the server.
  • There is now a link from each page back to the main author directory.
  • The "horizontal" links that Marc requested the other day (from "Ab" to "Aa" as well as to "Ac") were not implemented. I don't know enough about the bitmap tricks that Al originally used in this script to do it quickly and efficiently. Do we have volunteers? :) Ahasuerus 20:01, 29 May 2010 (UTC)
Very nice! I'll take a look at the bitmapping. --Marc Kupper|talk 04:33, 30 May 2010 (UTC)
The bitmap only works for the first two letters and only works with ASCII A-Z. It does not support other characters, digits, Unicode, etc. The logic to add/remove authors also updates the bitmap table. The table has 26 rows with one row per letter from A to Z. The directory_mask field is a 26-bit value with 1 for A, 2 for B, 4 for C, 8 for D, etc. where A-Z is the second letter. Thus we can use the bitmap to go sideways when we are at two letters. "Down" to three or more letters would be done by creating a bitmap as we are spooling through the select results to display a page. "Sideways" when at 3 or more letters involves doing a select for N-1 letters and making a bitmask from the results. I believe sideways can be handled with two select statements but will need to experiment with that. I'll think about how to best code this but need to shut down for the night. --Marc Kupper|talk 05:58, 30 May 2010 (UTC)
Personal tools