User:ErsatzCulture/Talk2021H1

Archived talk

Castles Made of Sand

Fixer got a US price as usual. Yours truly tracked it down to the publisher site and grabbed a price (and date verification at the time) from there - and was too lazy to also find the UK ASIN - so Amazon UK was not listed as a source; the publisher site was. ;) As I keep reminding you - what Fixer finds and what is getting recorded are very very different things ;) Annie 13:09, 6 January 2021 (EST)

PS: Happy New year! Keep up the good work you are doing with the UK books :) Annie 13:37, 6 January 2021 (EST)

You too - I got a bit unnerved when you seemed to be very quiet in December, especially as I'd read on forums elsewhere that Arizona - which is your part of the world right - was particularly bad for COVID. (Although it's hard to interpret the context from the other side of the Atlantic; I'm sure some of the reportage of the UK/London looks equally alarming to the rest of the world.)

Anyway, I've still got plenty of backlog to submit; whilst I've got the Lin Carter/Gateway ebooks down to roughly a dozen yet to enter, I did a trawl of the Adam Blade ebooks earlier today, and now I have ~30 of those in my queue, and I'm sure there are more that I hadn't gotten to... ErsatzCulture 15:09, 6 January 2021 (EST)

Work early on (December can be weird) and then needed a break from internet for a bit. :) Arizona is... Arizona. We were the poster child for what not to do and on top of everyone's list during the summer and fall but now we seem to be doing relatively well... compared to the rest of the States anyway - if you compare to the sane countries, well... :)

Better you than me on Blade -- a lot of his older ones also need conversion to chapbooks (which I am slowly working on especially when Fixer throws at me a box set) :) Annie 15:19, 6 January 2021 (EST)

Changing types

Just a reminder - when changing the type of a book, you cannot just change the title. When you are creating a book, you only add the name once but we create 2 records - publication AND title. Once created, a change in the type, author or title of one of the two should always be matched with a change in the other as well. And with title with a single publication, the fastest and cleanest way is to update the publication, not the title :) Annie 15:41, 11 January 2021 (EST)

Sorry, is this related to the poetry collection that was tagged as a novel? I've been a bit distracted with real world annoyances the past 36 hours, and I must confess I'd completely forgotten about it, although if I'm honest the two records issue hadn't entered my mind in this particular case... ErsatzCulture 16:16, 11 January 2021 (EST)

Yep - I finished the conversion :) It was just a reminder - the more one does the changes, the more likely is not to forget ;) Hope everything is ok with you. Annie 16:22, 11 January 2021 (EST)

Thanks. Will hopefully be able to get back into looking at the imminent UK books, in the hope that it'll distract from the annoying domestic issues which have been stressing me out. (Neighbour stupidity, which hasn't been resolved, but is currently in a state of uneasy armistice. Sigh....) ErsatzCulture 11:18, 12 January 2021 (EST)

Ah. My upstairs neighbor flooded my apartment a week ago so... you have my sympathy (no books were harmed partially because it hit my kitchen mainly and partially because I was around so I moved the ones that were on the way; on the other hand now I seem to need a new toaster - but that was long time coming anyway - the poor thing was on its last legs). Hopefully things get resolved quickly... Annie 11:39, 12 January 2021 (EST)

I had a similar problem with mine, but then I fell to the idea of reading this book to him - and it brightened him up in a way that he was around with me for almost one more year. So if you can get hold of the story that might work (or then it might not, there's a faint possibility that other factors were involved). Christian Stonecreek 12:46, 12 January 2021 (EST)

Well, it got drenched in water and it is not very happy when plugged in so... that won't help I would think ;) But then the poor thing was one of my first purchases when I moved 10 years ago and was bought because it was cheap (with the idea to replace soon) so 10 years of life and work is a lot more than expected anyway... :) Made me smile with this note though. :) Annie 13:34, 12 January 2021 (EST)

I hoped so: in these times we all are in need of some smiling, I'd think. Christian Stonecreek 13:55, 12 January 2021 (EST)

Indeed. I had the water-from-above problem ~8 years ago (fortunately not much, and nowhere near anything important), but my current problems are new neighbours in our block who think it is OK to have long noisy parties with many partygoers, despite all the restrictions currently in place in the UK, and who think that I am being unreasonable for complaining about it :-( ErsatzCulture 14:34, 12 January 2021 (EST)

Oh dear... :( Some days I am not entirely sure how humanity managed to survive as long as we had... Well, hopefully they will get better. Not much you can do with unreasonable excuses for human beings. Annie 14:43, 12 January 2021 (EST)

Heaf of Zeus (Apollo)

Our good friends at HoZ had resurfaced one of their old imprints Apollo. Heads up when you adding their books - I just added one. Annie 19:37, 17 January 2021 (EST)

Thanks. I thought I vaguely recognized this one, and it looks like I've got scraped data for an earlier tp (already in the db) and an ebook (currently missing). The ebook is an annoying one in that the Amazon listing has just HoZ, but when you check the preview, it's Ad Astra :-( I think because it didn't strike me as obviously speculative I didn't bother going any further with it, but now that I see there are other pubs of it in the db, I'll submit the ebook some time this week.

This did prompt me to check their site, as I hadn't noticed much new from them since the Ings Robot anthology in December. Annoyingly it looks like they've stopped issuing half-yearly upcoming release catalogues, which I've found useful in checking for anything that slipped through the cracks :-( ErsatzCulture 11:27, 18 January 2021 (EST)

When I process HoZ, I always check their site to see which imprint it is - their loading into amazon is indeed atrocious. And the catalog may be back - 2020 was a weird year. :) Annie 15:55, 18 January 2021 (EST)

Master of Sorrows

Did a bit of an update on the book you just added here - the Amazon book is a newer version that is still to come this May (possibly with a new ISBN? If not - we can reconcile later and add more notes but at this point I am not comfortable with using the Amazon 2021 record in conjunction with the Kobo/Publisher 2019 record) and August was a long time ago so fixed the "as of" date. Annie 13:59, 20 January 2021 (EST)

Sorry, I think this was me screwing up. Short version, I mixed up the details of the first volume published 2019 and the sequel due later this year.

Long version: I normally only submit when I have data from 3 sources, and this is much more visible in my tools when the sources can all be reconciled to the same pub. This is easy when you have ISBNs, but of course Amazon ebooks don't, so it falls back to author and title matches. However, the title match is very paranoid, to avoid the risk of matching (for example) "Dune" and "Dune: Messiah". If the titles don't match, then my tools show these as separate titles, and it's down to a fallible human (i.e. me) to manually go through and match things up.

In this case, the Amazon listing is "Master of Sorrows: The Silent Gods Book 1", which doesn't match up with the proper title as listed on Kobo and publisher site. There's a similar mismatch on the sequel, and when the aforementioned fallible-human was copy pasting from the two different sources, they (I) mixed up the Kobo & publisher details for the original and the Amazon details for the sequel.

(All details for both volumes were scraped on Aug 25th. If I've only got old data from before a book was published, I usually grab updates before submitting, but because this was a 2019 pub, it didn't seem necessary. In retrospect, I should have realized something was amiss, because I had different prices from Amazon and Kobo, which happens for some publishers, but not in my experience Gollancz.)

I'll go back and check I didn't mess anything else up on this - anything to get away from doing the boring variant title stuff for that Mexican author's different entries ;-) ErsatzCulture 14:23, 20 January 2021 (EST)

No worries at all - I click on the sources any time I approve and this one just did not add up - so I changed it and came over to let you know that something was up. Plus - even if you scrap in August, posting with an August date in January is saying to me "this was in August, now cannot be checked/vaidated - who knows what happened" so either change your wording somewhat or always use current date after a check. Or both - if it is validateable on the day of submission, it should be validated and the proper date entered. :) Annie 14:39, 20 January 2021 (EST)

PS: See here - feel free to reword it anyway you want but unless the source is not valid anymore, "as of" should be submission/validation date.

I made the 3 variants (1 to existing, 2 making new parents) - took all of 10 seconds - it probably took a lot longer to grumble about the need to do them ;) Go fix the author pages :) Annie 14:45, 20 January 2021 (EST)

OK, I feel appropriately shamed ;-) I've just submitted an edit for the legalname of the primary author record, but I'm not sure what if anything else might need doing? ErsatzCulture 15:33, 20 January 2021 (EST)

The alternative name should not have anything but a language and a directory name :) Annie 15:37, 20 January 2021 (EST)

One note - the directory name for an author is based on the NAME not based on the legal name. So it remains Zarate (the alt record will have the full name. Annie 15:41, 20 January 2021 (EST)

Incomplete

Just a kind reminder to use the {{incomplete}} template when you're adding collections or anthologies without (or incomplete) contents, like this one (I've added the template). When the contents gets added later, that template can then be removed again. Cheers! :) MagicUnk 15:23, 29 January 2021 (EST)

No need to add it for EMPTY ones - we have a separate report for them. Once a single story is added, it needs to be added though. Annie 15:28, 29 January 2021 (EST)

Thanks for jumping in and saving me from having to dig through my old talk to find when someone - probably you - said that { { incomplete } } wasn't needed if no contents were entered.

Generally I've tried to always submit contents for anthologies or collections I add, but this one is a bit awkward in that although the 18 stories in it aren't too onerous to add (compared to say the 100-story Robot anthology I did just before Xmas):

- Roughly 75% of the stories are non-genre, and I'd rather mark those appropriately, or omit them

- For some of the genre ones, the fact that they are genre is a twist/spoiler and/or not documented in the Wikipedia synopses

So whilst I'll try to add the contents - at least the genre ones - at some point, it's not something I'm rushing into :-) ErsatzCulture 15:50, 29 January 2021 (EST)

I was around. :) Most of the anthologies I add do not have contents either because I add them too early for it to be easily found or because I just don't have the bandwidth to also add contents. I do occasionally on review but depends. :) Annie 16:37, 29 January 2021 (EST)

Empire in Black and Gold

Well, this one appears to be 18 cm so pb but the other 2 are definitely tp so fixed.

And yes - there is a historical problem in the DB - 15-20 years ago, A format was a lot more often seen in UK. Then they mostly switched to a B format as the cheaper option and C as the premium but some US based editors kinda sorta missed it so in a lot of people's minds, A was still the norm (US still has a lot more mmpb than UK and the C format (or anything close to it) is rarely seen state-side but even here 5x8 and 6x9 inches books are the norm with self publishing and small publishers). So in a lot of cases editors defaulted on pb automatically just as a matter of fact. Thus the big amount of "pb" in UK books are that actually "tp". I do occasional cleaning sprees on these, a publisher/year at a time but... there are a lot of them and there is only one me and my list of projects is big. So anything from the last 15 years (20?) is "pb" and UK probably needs a once over to verify the format. Annie 12:11, 5 February 2021 (EST)

Thanks - I dropped out of reading fiction for ~2 decades, so I wasn't sure when the UK started doing tps as the main format.

It looks like that series is being re-issued one per month, so I'll check the existing pubs for pb vs tp issues when I add the new ones. (Given that my tools pull any dimension info in the scraped data to determine tp vs pb, and they reconcile to any existing pub record, it might be fairly easy to report if the dimensions don't match the stored format in any existing record - will have a look at it over the weekend...)

I had wondered whether it was worth asking/suggesting if we could have a new pub type for C-format/Demy/Royal/XTPB, but I suspected it wouldn't get much traction, and they're rare in the US, I definitely doubt that I'd get much buy in :-( ErsatzCulture 12:58, 5 February 2021 (EST)

They started transitioning around the time you dropped off I think :) I started reading in English ~2 decades ago (now I feel old...) so my "new" cheaper UK books were mostly B formats while the US cheaper ones were mmpb. But once I started looking at older books, the A formats in UK started showing a lot more... One day someone will write a fascinating dissertation on this topic... :)

Don't open that can of worms please... :) We had had a lot of bitter discussions (you have C format, Eastern European formats do not match anything else for example and so on - most of the smaller Bulgarian books are too wide for "pb" for example). What I wish we had was a way to just be able to add dimensions and/or free hand field to add format notes that is not the general notes field. You can try of course but expect the conversation to be hijacked by the German editors wanting their own formats... and to get nowhere as usual. So after a few attempts, I just gave up and use the notes. One of those days I will get on it again probably but... people already cannot grasp the difference between pb and tp, adding more formats will make this even more problematic (thus the "use one format for a softcover and allow the distinction to be in another field when known". But oh well... :) Annie 13:18, 5 February 2021 (EST)

(Only just noticed you replied - for some reason the Wiki didn't give me the yellow/orange notification bar.)

One possibility I did ponder, that seemed less contentious, would be to have some (optional) new templates for the notes field e.g. :

* { { Dimensions|130x193 } }

* { { Size|C-format } }

* { { Description|Slipcased limited edition } }

which isn't ideal - compared to proper data fields/columns - but (in theory) doesn't change much, and these template values could be something that could be programmatically extracted from the notes field. (The Description template would be free-text, but with a fairly short limit e.g. 30-40 characters.)

Or would this be just going over old ground again? ;-)

Something unrelated to the above, but from me going through this series, this pub looks like another one that never came out. I skipped it in my first pass, because it shows as a Tor (US) pub, but the cover style is the same as the 2016 Tor UK ones, and the ISBN is in a similar range and shows as assigned to Pan Books. If I search for the ISBN (9781447295082) on both Amazon UK and US, I see the page for that title, but that 2016 pub doesn't show, and it's not listed on the panmacmillan.com site. Safe to mark as unpublished, do you think? ErsatzCulture 17:30, 5 February 2021 (EST)

You probably reloaded the page after I posted and did not see so the system thought you knew of it - so no yellow notification :)

I will try to find some of the old threads - but yeah... pretty much. Templates may work but good luck getting people that are not you and me to use them -- I still need to chase people to use Tr and Narrator and incomplete... Still may be a way to collect the data - I know I collect it when I add Russian/Bulgarian books... Let me think on it a bit.

For the book - OCLC(you need to toggle to see the library list for only this edition) shows only one library cataloging it (a US one)... which almost sounds like an old entry that never cleaned it up. And Abe (plus the Amazon UK, com and CA checks) has no copies which almost never happen with regular modern books and the ISBN is definitely on the UK side (Tor is owned by Pan Macmillan in UK which is why it shows so weirdly in some listings). So I'd say that it most likely did not make it out. Annie 17:43, 5 February 2021 (EST)

Duh, just plugged that ISBN into Waterstones, and they report it as "Publication abandoned" (which they did for one of the earlier ones, but not the others), which seems good enough - will submit an update in a few minutes. ErsatzCulture 18:01, 5 February 2021 (EST)

Yeah - even without Waterstones I would have said so - the thing is just obviously not there :) Annie 18:31, 5 February 2021 (EST)

And we have a template for OCLC links :) Annie 18:34, 5 February 2021 (EST)

Adam Blade books

You know that if you CLONE instead of ADD, the stories will come with you, right? So you do not need to import them later... Annie 17:27, 8 February 2021 (EST)

Yes - brain fade on my part. I think I was so pleased to find one lacking an ebook that was a CHAPBOOK rather than a NOVEL (needing CHAPBOOK conversion, if I'm trying to do these properly) that I forgot to do a clone rather than my habitual AddPub. Will submit an import in a second... ErsatzCulture 17:32, 8 February 2021 (EST)

The converted ones are the ones that had had an edition coming to me basically - I am converting as and when I get a book to add... I probably should just hit the whole series one evening and convert them all... Annie 17:46, 8 February 2021 (EST)

Well, I have a long list of ones lacking a UK ebook, which I'm trying to slowly plod through doing 1 or 2 every week. I'm trying to pick off the easiest ones first (i.e. the ones which are already novellas) to get the list down to a more manageable size, but I'll happily do the novel->novella conversion as part of that process once I get to it. (I think I'm finally getting close to the end of the Lin Carter/Gateway ebooks, which will free up my attention a bit.) ErsatzCulture 17:56, 8 February 2021 (EST)

Talking about that, is this publisher on your radar (I just added these 4 after the paperback showed up)? They are mostly literary but they have some of ours... Annie 18:09, 8 February 2021 (EST)

I've added a couple of W&N books - they did some Daniel Keyes tp reissues late last year - but I think those were more a case of stumbling across them by accident; IIRC I saw that the ebooks were missing, and when I went to scrape the data from Amazon, it also flagged up those new tps. Unless Amazon decides to recommend me one of their products, or they have a book show up somewhere like The Guardian's monthly (?) SF&F roundup, I'm likely to miss them.

BTW, there are a few titles/pubs in the Locus recommended reading that aren't in the database. I've added the missing ones from the SF category (other than the James Bradley, which isn't actually out yet in the ISBN they link to - so much for a "2020" list ;-) but there are 3 in the fantasy list and a couple in horror that I haven't gotten round to. I did look into the fantasy ones, and they didn't actually seem to have any speculative content (going by the Amazon blurbs and Goodreads shelves), other than maybe the Elizabeth Bear, which is the 3rd or 4th in a series, where the preceding volumes don't appear to have any speculative content. The Bear isn't available in the UK except as a grey import on Amazon, so I haven't gathered enough data (as yet) to make a submission. LMK if you want more details; I'm sure you don't have anything else to fill your time ;-) Some of them appear to be unknown to Fixer, although that might be down to using a slightly out-of-date Fixer dump. ErsatzCulture 18:48, 8 February 2021 (EST)

Bear is above threshold so even if she writes a cooking book, we can add it. :) Some of the horror ones may not be eligible (remember that we ONLY do speculative horror - and the horror awards and lists are not always that specific - I know that I kicked a few of the list even though Fixer gave them to me based on pre-publishing information - but I am planning to take a second look and read them and then decide on them). I have the Locus list on my "todo" for one of these weekends... And then there is the "slightly off and into magical realism" books which are hard to judge...:) Annie 18:55, 8 February 2021 (EST)

The Harpy

39K in Kobo may mean 40,001 from what I had seen so too close to call. Audible is 4 hours 18 minutes which again is too close to call. I will do some more digging... Annie 17:35, 8 February 2021 (EST)

I was mildly inclined to add a note to the title similar to this one, but similarly I'd want to do a bit more investigation. On the odd occasions I've extracted the XHTML from a non-DRMed epub and converted to plaintext, the word count (via the 'wc' utility) has always matched what Kobo report, but that might just indicate that we're both calculating it the same way, and I certainly wouldn't care for that to be used in something like award eligibility decision-making. ErsatzCulture 17:48, 8 February 2021 (EST)

The problem is that Kobo calls 39,999 39K. So if the text has a few typos that connected two words together (happens...), it will report 39K on a 40,002 book for example - which on the next edition may get fixed thus sending it into 40K. That's what I meant more than actually having an issue with the counting on Kobo. The over 4 hours audiobook is also pointing to something possibly very very close to 40K (again there is some wiggle room...). If I do not find anything else, I will just add a note explaining what the different editions show - and leave it like that for now until someone does a proper count or something... I hate these borderline cases.Annie 17:57, 8 February 2021 (EST)

Pattern Recognition

Hi, as we already have two Penguin 2004 editions of this book, and ISBN-13, price and cover art all suggest a post-2007 publication, can you please alter the publication date to 0000-00-00?--Dirk P Broer 08:48, 10 February 2021 (EST)

Um, although I'm the PV on this one, the entry as it currently stands is the work of another mod, per this discussion.

I've got no objection to making the change you request, and in fact was in the middle of doing the submission, when I was tweaking the pub note to record the 2004 date that was being lost from the pub date field, when I spotted this comment that the other mod added, specifically the final sentence:

> "Book only has copyright year; publisher's website lists a publication date of 2004-06-24. Since a 2004 date and an ISBN13 are unlikely, it is possible that the ISBN was added in an updated version of the ebook. However, recording as is listed in book per standards of documenting per book."

which reads to me as entering a 2004 date is following the standards (even if, I agree, it seems highly unlikely for this cover - although it seems plausible likely that a new ebook file was issued post-2007 with the newer cover, but no updated interior content)?

It seems pretty messy on the part of the publisher, and maybe the rules on how to cope with such inconsistencies and unlikelinesses need tidying up? I know there's at least one missing Gibson ebook that I've avoided submitting 'cos I wasn't relishing going through that hassle again, especially as in that case I don't have the book myself that I can PV. ErsatzCulture 09:50, 10 February 2021 (EST)

Alexandria

That "ebook 2 days before the tp in UK" usually means that there is a US edition out there coming that week. UK's big day is Thu, US is Tue. So I went looking for the US edition - and discovered something different. Apparently the ebook was published 2 days early so it comes out with their other ebooks (or something) but it turned out that the US edition was published back in 2020-10. And on its copyright page it actually says that the first edition was Faber and Faber in 2020 - this one that got pushed into 2021. I left a note to that effect here - this is bound to throw a LOT of people into a research frenzy in a few years. I thought you may at least smile at that - publishing is...interesting. Annie 11:44, 16 February 2021 (EST)

TBH, I should have maybe spotted/realized that, as I had scraped data from Blackwells for that 2020 US pub. However, I didn't have info for it from any other sources, which usually means it's an import into the UK, and they get put to the back of my mind/queue, so I forgot about it when I wrote that mod note.

I imagine there are also missing US pubs for the middle book in that trilogy, as Blackwells lists a 2017 Graywolf ebook, although at least that's after the one UK pub I've submitted for it so far (I've got Faber hc & tps still to submit for it).

Re. the US=Tue/UK=Thu thing, a few days ago I had a tweet from a Canadian author RTed into my timeline about that affecting their upcoming book, and blaming it on Brexit - I had to use all of my self control not to reply with a "Well, actually..." ;-)

If you want another one to look into, another author tweet RTed into my timeline was about this nominally UK pub coming out today. I'd submitted the AddPub for it, but I hadn't clicked it was a Tuesday release. I know you added US ASIN and price to the corresponding ebook, but I suspect the Tuesday UK pub might indicate it's an international tp? (Although I think there's also a 2020 Australian tp not in the DB that was the original physical release, but I never delved into it properly.) ErsatzCulture 12:12, 16 February 2021 (EST)

No worries on not adding - I did not expect you to.I just thought you may appreciate the story about the copyright note. :) And yes - there are more missing ones in that author - I put it on my list of things to look at... which may mean that I will get to it either tomorrow or in 5 years...

Yeah - the Tue/Thu is not really valid for small presses but the big ones? 95+% of theirs follow pattern on the US side, a bit less on the UK side -- the ones that do not are usually matching the US editions or being international. Brexit has as much to do with it as the phases of the moon. So even if Thu is usually the UK day, some do make it on Tue even when they are local. I will look into this one shortly. Annie 12:31, 16 February 2021 (EST)

"Ghost Species": It seems like there is no US publisher for that one (none projected...) - which makes Hodder's edition probably legally exportable and available in the States so either they claim international or they just published Tue to hit the US market. See my note above on UK books which come on Tuesdays sometimes out of the blue... And with Bradley being Australian, having a first edition there is kinda expected. I will add the missing AU publications :) Annie 12:38, 16 February 2021 (EST)

One more thing: careful with Hachette links: vip6.hachette.co.uk sends you into login screens if you are not into WP already (I guess... or not in UK or something); you need the www.hachette.co.uk ones. I fixed it on both Bradley editions, will check if more need repair. Annie 12:51, 16 February 2021 (EST)

Ah, that's weird about the vip hachette links - I actually get the canonical URL defined in an HTML <meta> tag for the URL field, on the presumption that this is likely to be cleaner and/or more permanent than the scraped page's actual URL (which might be something like /search?isbn=9780123456789). Now that I've actually clicked on one directly, I see that I get the login page as well, so apologies for submitting bad data for god knows how many pubs :-( I'll put in a hack to change the vipN. prefix to www. for future submissions. ErsatzCulture 13:08, 16 February 2021 (EST)

Only 2 more still standing - I remember fixing a few more in the past on approval and it just clicked that I should better stop by and tell you when I saw this one. I will fix these. Meanwhile, unless it had a publication in another market, I think I tracked all editions. I will check NLA, OCLC and Goodreads later as well just in case. Annie 13:12, 16 February 2021 (EST)

And fixed - they were these two. I approved one of them and missed it as well so... sometimes it works like that. I tend to check link in a clean browser in anonymous mode (or whatever) the first time I see a new site) - but sometimes they slip. The rest of the links and notes look clean - searching for hachette.co.uk. Annie 13:15, 16 February 2021 (EST)

Yeah, I ran a query against my copy of the DB (as of Saturday's backup) and got the same. I'm a bit confused, as I thought I'd seen more than that - which might be explained by you fixing my submissions - but I don't see any more cases in my scraped pages, other than non-genre pubs I picked up whilst idly browsing. (I also checked for any URLs containing "vip", because Hachette UK seems to have loads of different domain names, all of which I suspect are hosted on the same server(s) ErsatzCulture 13:20, 16 February 2021 (EST)

May have been a temporary change that you got caught into. Who knows. I tend to fix links silently when I know that the editor is usually careful unless it is a pattern (mistakes happen) so... :) No worries at all. Annie 13:25, 16 February 2021 (EST)

(unindent) As it turned out, this needed moving 2 days early to accommodate the ebook. I cannot find an ISBN for it though - see if you can (and I will check if Fixer has it). :) Annie 14:14, 16 February 2021 (EST)

Hey, are you taking away all my UK pubs? How am I going to climb the ranks of the top contributor lists if you do that? ;-)

Anyway, I've submitted an edit for that ISBN - both Kobo and Blackwells had it. ErsatzCulture 14:46, 16 February 2021 (EST)

I need coffee apparently - I looked at both (US Kobo has only the US book)... Oh well - NOW I see it...

PS: Sorry, I was adding the 5 US ones and saw that it needed a day moved to add the ebook and while looking for the ISBN spotted the paperback (with price and everything) and another clone made sense. In case these last two did not make it clear, when I get on a specific title and decide to fix it, I cover all the bases I can get my hands on. But don't worry - there are a LOT more missing ones. ;) Annie 14:51, 16 February 2021 (EST)

The London Reader, Volume 17

If you are going to submit more magazines, here is the corrections that were needed this time:

The Magazine issue should have its number and/or date as part of the title
Non-Genre magazines are not credited to their proper editor but to "Editors of Magazine". In case the editor is someone we know, they can be added as a second editor.
Magazines get yearly records so links about issues and any comments for specific issues go in the publication notes and web pages list, not the title ones. The title level should contain comments for the complete year (if there is something special there)
The series number is never used for magazines, the number of the issue goes into the title
Magazines always require a title series so we can build the grid. :)
Magazines are dated based on the dates printed in them, not Amazon. Autumn 2020 means 2020-00-00 unless you find another statement to lock the month. Amazon date can go in the notes if you want.
As this is not a genre magazine, it needs to be marked as non-genre.
As it is a non-genre magazine, adding the genre contents that made it eligible is kinda recommended. :)

The result is here. Welcome to the magazines part of ISFDB - it is a whole different ballgame here. And you thought the juvenile chapbooks are annoying... :) Annie 18:35, 25 February 2021 (EST)

Thanks, IIRC the only time I've touched mags before is either (a) when an issue had an award-nominated non-fiction, or (b) a couple of issues of Vector that were stopping having a complete set here, and in both cases, there were plenty of existing issues that I could follow the pattern of. (And I think in at least one case I ended up running a copy of the full system here and working out what needed to be submitted to match the existing records.)

What's also a bit of a puzzle is that there's another pub from this organization, which I don't think is a regular issue: https://www.amazon.co.uk/cyberpunkNOW-Dystopian-Moment-London-Reader-ebook/dp/B08NQMFNMF/ref=sr_1_2?dchild=1&keywords=london+reader+cyberpunk&qid=1614293342&sr=8-2 It's listed as "Volume One", but has roughly the same pub date as the "Volume 17" I just submitted?!? I'm more than happy to let you work out what's going on there, if you want ;-) ErsatzCulture 18:55, 25 February 2021 (EST)

That is why I said welcome :) The magazines have their own rules... No covers for non-genre magazines if they are not ours - this one is obviously ours so I left it but if you are adding more... Annie 19:07, 25 February 2021 (EST)

EDIT: close inspection of the copyright page of that "Volume 1" indicates it's originally from 2016, but seemingly only recently been made available on Amazon? ErsatzCulture 19:03, 25 February 2021 (EST)

Yep - VERY common with Amazon e-versions of Magazines. Date as described above, note in the Publication on Amazon date. If you want to treat it as a reprint as opposed to a magazine issue, then it will become an anthology. If you want to - add it. If not, I will see what I can do later. Annie 19:07, 25 February 2021 (EST)

Drogan the Jungle Menace

I finished all other actions needed on this one. When converting look at the list of publications. In this case it is only 1 so this can be fixed in only 4 edits, 3 steps (need to wait for approval between steps - 2a and 2b can go in parallel):

Step 1. Edit Pub: Change the Pub to Chapbook, the TITLE record to short fiction and add a chapbook
Step 2a. Edit Title: Mark the new chapbook as juvenile
Step 2b. Edit Title: Convert the parent of the story to a short story (this can also go as 1b if you prefer) :)
Step 1. Make Variant: Create a parent for the chapbook.

You went the long way round:

Step 1. Edit Title: Change the novel to short story
Step 1b. Edit Title: Convert the parent of the story to a short story (can also be 2b or 3b)
Step 2. Edit Pub: Change the Pub to Chapbook and add a chapbook
Step 3. Edit Title: Mark the new chapbook as juvenile
Step 4. Make Variant: Create a parent for the chapbook.

Not that the second way is bad in anyway but it requires one more step that needs to be waited on before you can proceed. If there were multiple Pubs, you had to go the long way but with just one pub, fixing the pub directly is the fastest. Thanks for working on these! :) Annie 16:11, 5 March 2021 (EST)

Aargh, it was actually your instructions I was trying to follow! :-) Specifically:

> When converting a novel to a chapbook, the old novel record becomes the story, not the chapbook as you did here

I guess in the context of that earlier talk, it was obvious whether that was referring to title or pub, but not today when I was revisiting things several months later :-( Ah well, maybe next time - God knows, there are enough Adam Blade "novel" pubs that need this fixing that I'll got plenty of chance to practice... ErsatzCulture 16:53, 5 March 2021 (EST)

That is why I wrote it down this way so you have it spelled out now. You do a few hundred of these and you will get used to them. What you did was NOT wrong (well, once you made it a story anyway) - there is just an EASIER way in this specific usecase that saves you one waiting time. :) Annie 18:04, 5 March 2021 (EST)

Larry Correia

When we disagree with a major source (Wiki for birthdays), add notes explaining why as part of the real notes and not just moderator ones (edit history is NOT visible if you are not logged in so someone WILL try to fix it back based on Wiki unless we have the explanation of why we disagree). I converted your note. Feel free to change it further. Annie 16:12, 7 March 2021 (EST)

That did cross my mind, but given various real-world things that those sources have been involved in, I chickened out of linking to them directly, in case it embroiled ISFDB in any Puppy/culture war type battles. I've got no objection to you having your name on the change ;-) ErsatzCulture 16:28, 7 March 2021 (EST)

Well, using author-run pages and author posts as sources of the author's biographical data is unlikely to cause issues. I think the closest we came to real controversy was approximately 10 years ago when an editor linked an author record to a third party social media post accusing the author of some kind of misbehavior. Luckily, we caught it quickly and deleted the link, which wasn't eligible as per Template:AuthorFields:WebPage. Ahasuerus 21:59, 7 March 2021 (EST)

Although weirdly, if I go to that author page, I don't see an Edit History link in the top right , even though the sidebar says I'm logged in. (I do get those links if I go to one of his titles or pubs, but not if I go to a different author's page.) Do you see the same, or is that a bug? AFAIK edit history should show on pretty much all pages now? ErsatzCulture 16:28, 7 March 2021 (EST)

Ah yes. Author pages - it is moderator only because of our rule to remove personal information if asked by the author. Not much of a point if a free account makes it visible. Annie 16:34, 7 March 2021 (EST)

Plus I assume all the submitted content ever is visible via Recent Edits, albeit you might have to page through 4 million submissions to find out what people entered? Or does that have a limit, or can records of edits be removed or hidden? ErsatzCulture 16:40, 7 March 2021 (EST)

They are there - but you need to find the previous one - not the one that changed it (because of how the screen works). If someone wants to do that, they can find it elsewhere as well - so we do the best effort and do no make it easy. And the public backups do not contain the submissions data. (Or did not anyway - had not checked lately)

Our publicly available backup files do not contain the submission table. If they did, they would be MUCH bigger :) Ahasuerus 18:07, 7 March 2021 (EST)

P.S. I am also wondering about his place of birth. Wikipedia doesn't give one and links to his "About Me" page, which says that he "was raised in El Nido, California". Goodreads says that he was born there, but "raised" doesn't necessarily mean "born", a distinction which Goodreads occasionally fails to make. Since Goodreads also claims that he was born in 1977, its data presumably doesn't come from the author. Unless we find a better source, we may be better off moving "El Nido, California" to the Notes field with an explanation of what various sources claim. I would also expand the current Note to make the year of birth (1977) that the other sites -- Wikipedia, SFE3 and Goodreads -- give explicit. Ahasuerus 18:47, 7 March 2021 (EST)

Feel free to edit as I said way up high. I just made the moderator note into a formatted note so we have the source - it can be improved in a lot of ways (as pretty much any note we have anywhere). :) Annie 18:58, 7 March 2021 (EST)

Done! His date of birth was also off due to conflicting sources. I have fixed it and let SFE3 know. Ahasuerus 19:38, 7 March 2021 (EST)

I've left a note to the editor and the moderator who approved the change (we used to have the correct date) just in case they had seen another source that may need to be added in the notes (or that changes something) - the submission had not moderator note... I think we are all done here :) Annie 21:36, 7 March 2021 (EST)

(unindent) Thanks for finding this one! Annie 21:36, 7 March 2021 (EST)

Kyron, Lord of Fire

The book, the chapbook title and the short story goes to Adam Blade (as credited), then you variant to unknown (until we find who actually wrote them - then these unknowns change). All fixed here. Annie 11:37, 9 March 2021 (EST)

Thanks - is that way just because (now I've thought about it) it's probably one step less than the way I was doing it? I looked at one of the other titles/pubs, and because the "unknown" record was the parent, and the "Adam Blade" record the child/variant, doing the parent first naively seemed a reasonable approach.

It will be possible to start with a parent publication IF there was a book with the parent record as a PUBLICATION author - so it works for translations and books published under both the names but not for pseudonyms like this one where the books are always under the child record. Just use whatever is on the title page (or cover if the title page is not available), then we will sort out. Using anything else will mean that the pub author is now wrong :) Annie 12:25, 9 March 2021 (EST)

One more thing on the Adam Blades - does the "unknown" author parent variant need to be done on both CHAPBOOK and SHORTFICTION records? Looking at this and this, that seems to be the case, but I dunno if that happens automatically for both titles if you variant one of them? ErsatzCulture 05:07, 10 March 2021 (EST)

Yes, you need to make both manually - the system does not know that they are connected technically so it can guess it (and you want to make sure the child is set as "juvenile" before you create the parents - so the setting carries up. Otherwise you will need to set it on each of then. This has nothing to do with these going to unknown technically - each title from a pseudonym needs to have a parent regardless of its type. On a separate note - want to rejoin us on the live server - these links look like your local version (? :) Annie 10:19, 10 March 2021 (EST)

Ha, the local links are due to a (debatable) bug that caused me to run up a local copy of the system:

If you go to the live Adam Blade page, the 2 upcoming titles/pubs I've added appear in the lower section of the page. However, on the same page on my local server, using data from Saturday that doesn't have those titles/pubs, I have the "Alternate Name. See ...{lots of names}... (or view all titles published using this alternate name)" lower section that you'd normally see on the live system for Adam Blade - and I wanted the latter link to be able to easily find some recent Adam Blade titles, rather than randomly clicking on all the real author names until I found one.

Not sure if/how that could be solved if there are titles/pubs that are attributed to house/gestalt authors rather than "unknown" or actual authors - any ideas? ErsatzCulture 10:30, 10 March 2021 (EST)

Not a bug, the link is there :) If there are titles, the link is only on the left (Menu: "Show All Titles" opening this). When it is empty, it is both on the left AND under the names. Annie 10:56, 10 March 2021 (EST)

BTW, completely different subject: you wouldn't happen to know anything about 9781597809474, which is supposedly a Night Shade US hc of a Jodi Taylor book, also listed on Amazon UK, but with both showing as currently unavailable, which makes me wonder if it never came out? (I don't recall exactly when Night Shade had financial/ownership issues, and whether that meant some pubs never came out, possibly including this one.)

It is a 2016 book and Google knows about it. If I am not mistaken, it was a pretty limited run that sold out very fast (I know that when I looked for it in 2017, it was nowhere to be found anymore) - but I can look around and see what I can find out. Annie 12:25, 9 March 2021 (EST)

Fingers crossed, I'm getting down to the last one or two Jodi Taylor UK pubs to add - excluding non-genre pubs, and seemingly another *4* new titles that are due by the end of the year - and this Night Shade one seems like it will prevent me having a "clean sheet". I just searched for that ISBN on B&N, and they didn't seem to know it? ErsatzCulture 12:06, 9 March 2021 (EST)

At least 4, yes. I read this author and as soon as you are done with her, I am going to also hit her with all Fixer known ISBNs/ASINs as well and then trawl GR and OCLC and so on :) Annie 12:25, 9 March 2021 (EST)

The First Omega

Did a little surgery on your note for this one. If it is not on Kobo (they have only the US one), we better not have it as a source - and then the next paragraph got a bit moved around to accommodate that, keeping the note about Kobo lacking it (As opposed to not having been checked). Feel free to edit more :) If you have any concerns, let me know. Annie 17:51, 11 March 2021 (EST)

Sigh, I left the note on your talk page re. this pub to try to avoid anyone having to revisit this one :-(

I note that the linked Hachette UK page has links to 5 ebook stores, only Google Books knows about the UK ISBN; Amazon, Kobo and eBooks.com failed to find anything, and Apple Books wants me to install an app, which I suspect isn't available for my Linux machine...

Another one for User:ErsatzCulture/TODO_Reminders... BTW, on the vaguely related subject of checking stuff after it comes out, do you have any understanding of when/why ebook stores might not have previews of books that have been published? Band of Gypsys/Gypsies came out over 2 weeks ago, but neither Amazon UK nor Kobo have a preview to allow me to verify which spelling is used on the title page. ErsatzCulture 04:54, 12 March 2021 (EST)

Well, these are not automatic most likely so probably the publisher did not send one or Authorize one or who knows. Get sample should work in Amazon - and that usually contains more of the text than the look inside. Annie 09:22, 12 March 2021 (EST)

Ha, that never crossed my mind, as I've never used that functionality before. (Having a sample wouldn't make me any more or less likely to buy an ebook, as 99% of the ones I buy are only when they've been reduced in a sale to the point that I don't care if I've spent 99p or £1.99 on something I never get round to reading...)

Anyway, it confirms that the cover image ("Gypsies") doesn't match the title page ("Gypsys") - will update the pub note. Fortunately that's all that needs doing, no title varianting for the alternative spelling :-) ErsatzCulture 09:30, 12 March 2021 (EST)

I almost never buy a book I get a sample of but the samples are very useful for here - they almost always give you the complete copyright page and the complete table of contents and quite often the complete or partial introductions and forewords and the like so their authors and titles can be seen. And you can read these on your computer - you do not need to open it on a kindle - so even easier to look at them. ;) Annie 10:18, 12 March 2021 (EST)

Silexa the Stone Cat

So.. remember how I mentioned that you should have the child record's flag set to juvenile and THEN create the variant? As you forgot to do that, there are now 4 more edits that need to be done here - marking the 2 chapbooks and 2 stories as juvenile. :) For next time:

When submitting the initial chapbook, this submission, click on the Juvenile. This will take care of the chapbook.
Once approved, on the same update on which you are adding the series (this one), mark the juvenile again. This takes care of the story.

Now the two "Make Variant" will set these in the parents on creation as opposed to needing 4 more edits to deal with that. Yes - Juvenile chapbooks are extremely annoying... :) Annie 20:11, 11 March 2021 (EST)

Sigh #2 - obv. I knew that this should have been juvenile, but too many things I was trying to keep track of on these crowded that out of my tiny brain :-(

Besides all the faffing around to make these match the ISFDB data model, to get to that point I also have a bunch of copypasting from different files on my side, because the publisher managed to get both of these upcoming Adam Blade books listed under 5 different titles on all the stores: "Beast Quest: Silexa the Stone Cat", "Beast Quest: Silexa the Stone Cat: Series 26 Book 3", "Silexa the Stone Cat", "Silexa the Stone Cat: Series 26 Book 3", "Silexa the Stone Cat: Series 26 Book 3 - Beast Quest". My code does strip out some of the more obvious title crud - SEO stuff like "The most unputdownable thriller of the year" - but doesn't try to handle those, and so I have to manually "merge" them from 5 different local pages :-(

Maybe I should just forget about these, and let you and Fixer deal with them... ErsatzCulture 05:03, 12 March 2021 (EST)

Then you need better code :p. Now back to being serious - if you don’t want to deal with them, don’t. They are annoying but there is a system in them and when you do enough, they become almost automatic. With Fixer, the very first one is not set as juvenile so I need to go the long way around as well.:). And the publishers seem to be hiring drunk penguins for their data entry this days. Annie 08:22, 12 March 2021 (EST)

> if you don’t want to deal with them, don’t. They are annoying but there is a system in them and when you do enough, they become almost automatic.

Well, that's why I'm doing them, hopefully I'll get them right eventually. However, if it's more effort for you/the mods to fix with the mess I've been making of them, compared to you/Fixer adding them in the first place, just say, and I'll leave them.

> Then you need better code :p.

The code for scraping the vendor and publisher sites, munging it all together, and comparing it against a copy of the DB, is fairly stable now. There are lots of bugs, but I know what those bugs are, and can generally deal with them, even when it's painful like these Adam Blade books. (And >90% of the submissions are fairly easy copypaste jobs, the bad ones are the exception.) The code is pretty gnarly now though, so I'm a bit reluctant to touch it, unless I'm prepared to spend a lot of time doing a "version 2.0" to fix all the bad assumptions I made in my original naive understanding.

Instead the main thing I've been working on lately is a tool to view all the pubs for a given author in tabular form. This has a few uses (e.g. seeing which titles are most out-of-print), but in the context of ISFDB submissions, it can be useful to see where the gaps are. e.g. here's Jodi Taylor:

[File:

]

If you know that Headline reissued all her prior titles in ebook in 2019, then it's easy to see which 4 pubs are currently missing from the DB :-) This ASCII rendering is just a starting implementation; I plan to do something with HTML and/or SVG to highlight what the individual pubs are (US vs UK; hc vs tp vs ebook etc), rather than just the simple number-of-pubs-that-year value.

Let me be clear about something - IF an editor is willing to work and learn and improve, I am more than happy to spend even 10 times more time and effort than it will take me to fix it myself. I will write notes, I will explain how to improve, I will approve and fix as many submissions as it takes. And then I will do it again and again until it sticks. So I do not care if it takes us 20 iterations or if I could have done them faster. I really don't. If that makes you independent on these, I am fine with that.

And in case it is not clear, I was joking about the code. :) You had seen the initial submissions of my dancing partner - half of the time I need to change every single field but the ISBN. It still is useful but I fully understand how complex that thing is - and Fixer is pulling only from 2 stores for new books... Annie 13:21, 12 March 2021 (EST)

(unindent) OK - the stories are now juvenile. Now the two chapbooks. Or it will pop up on the report tonight :) Annie 14:57, 12 March 2021 (EST)

And this can use a language assignment. Annie 14:59, 12 March 2021 (EST)

Yes, I know there were still 2 juveniles to go, but I was struggling to avoid submitting 2 identical changes on those ones, that I wanted to wait until they were in, so it would be clearer which were the 2 that still needed doing. (I should have stated that in the mod note though.)

Re. the language thing - that's something that's confused me before now. I've seen Stonecreek do a bunch of them in the "Recent Edits" page, but because the UI defaults to having English showing as the current value for "Working Language", it's not obvious that that hasn't been selected by default. (Without looking into the actual edit content, I assumed Stonecreek's edits were to set author records to a non-English language?) Perhaps that field should behave more like the format field for pubs, which explicitly defaults to "unknown"? ErsatzCulture 16:11, 12 March 2021 (EST)

What the UI defaults for the ones without language is whatever is your default in your preferences. Just at the moment mine defaults to Bulgarian because I swapped to add a few books and then left it there so it is more obvious when I need to submit it for titles and authors :) We have 59729 authors with no language assigned. If it does not show as assigned when you look at the page, you need to submit it so it is assigned. :) Annie 16:21, 12 March 2021 (EST)

But that's a bit of a "gotcha" UI - you have to know and be paying close attention to spot that there's no language value. If it explicitly showed something like "Language: undefined" (perhaps only to people who are logged in who can actually change it) then it would be more obvious that something was amiss. As it stands, with having both the regular author page and the author edit page not explicitly showing that something is missing, it's hard for a novice editor like me to know that this needs doing. Perhaps it should be a yellow warning - although I guess that's not really feasible when an author record is automatically created as a side-effect of adding a title? ErsatzCulture 16:26, 12 March 2021 (EST)

The regular author page is showing it just fine - there is no language on it. Look at the one I linked. Now click on the parent. Do you see how one has language, the other does not. Putting a line saying undefined will just repeat that... :) ALL authors are created automatically technically - you cannot create an author - they are always side effect of creating a title or publication. You know who you need to talk to about the edit page and if it can/should be changed.

Once upon a time the DB did not support languages. Then it did. Then came the big titles cleanup where we assigned languages on all titles - one of my big projects from before I became a moderator. And we are making some progress with the authors... :) Annie 16:38, 12 March 2021 (EST)

> Do you see how one has language, the other does not. Putting a line saying undefined will just repeat that...

I think this is going to be an irreconcilable difference of opinion; from my POV, the lack of anything indicating missing information means that anyone not super familiar with the system will not realize there's anything missing, compounded by the fact that if/when you do edit an author record, the language field defaulting to the user's preferred value looks identical to a correctly set language. I appreciate that if your default language doesn't match the author's working language, then that error is more obvious - but on a site where (I imagine) the majority of editors and author records are Anglophone, it's much harder to spot. Sure there's a nightly report for these, but that's "Beware of the Leopard" syndrome again :-(

Let me put it another way: what's the logical reason why the format field when editing a pub defaults to "unknown", but the working language field on the author edit page defaults to showing a value, even if one isn't actually defined?

Anyway, there's an edit in the queue to fix the language for that author, and the juvenile flags on the Adam Blades. ErsatzCulture 17:09, 12 March 2021 (EST)

And you are free to complain about it to the boss or over in CS. I cannot do anything to solve that - all I can do is to help you see the difference so you know when you need to edit (which I did)... :) Annie 17:24, 12 March 2021 (EST)

(unindent #2) Sorry, I appreciate the design/implementation isn't yours. I may well raise this as an issue, but let me just pick your brains on an alternative way to solve this: if authors get created when a novel (or chapbook, etc) is added, and the edit screen has a language drop-down (which I imagine people will set correctly 99% of the time), would it not be reasonable for the created author record to default to the language of the title/pub that it was created from? I'm sure there are cases where either:

an author might be multilingual, and the title that causes their author record to be created isn't the primary language, or
the title record that causes the author creation is a translation, and the original language title isn't in the database

but I'm guessing these are a tiny minority of cases?

Plus, it's not clear to me what function the author language field actually provides, and what the consequences are if it's not set (beyond the nightly report), or if it's set to the wrong value? Schema:authors doesn't give any hints, and it's too late in the evening here for me to feel inclined to work it out by looking at how it's used in the code. ErsatzCulture 17:47, 12 March 2021 (EST)

Actually not even close to a tiny minority. Every Bulgarian book WILL create a new author record (unless we already have a book in Bulgarian by this author) because it uses a different alphabet. Same for Russian and Japanese and all other non-Latin alphabet based languages (plus the funny Latin-alphabet ones like Polish). Every time you add translations of Bulgarian, Japanese, Russian and most Polish books (to name a few language) into English, you are creating a new author that will be marked as English when they are not. So these will be messed up from the onset. Add to that a new French title that we do not have the original author of (happens a lot with small authors for example - we get a translation before we get the first original)... And we are seeing more and more international editors. Making the DB even less user friendly for non-English editors and speakers (especially the ones that are not using the Latin alphabet) than it already is, is not a direction I want to ever see the DB go. Even if it will simplify the life of the rest of the editors. :)

What the language provides is the ability to discover where there are titles which are NOT in the correct language for this author and have no parents - so they need their original to be found - we are slowly working on these via a report (nowhere near being able to handle all types yet although the longer ones are mostly cleared. Any other proposal how to discover if any of our million+ titles may need an original to be attached to it? :) The languages mean less for artists and we had debated their usefulness there but as we do not have separation, we just add languages to everyone. :)

I know that you are working mainly in English so for you translations may be irrelevant and possibly annoying but... we are an international DB. And a huge amount of records I handle daily are NOT English - even from Fixer - which under your schema will always mess up the authors when they do not match the form we have). :) Annie 18:11, 12 March 2021 (EST)

The original reason to add a "Working Language" field to author records was to support a more intelligent display of languages on Summary pages. Let's pull up Philip K. Dick's Summary page and scroll down to the "Collections" section. Most non-English collections are VTs of Dick's US/UK collections. However, there are over 30 collections of translated stories with no parent titles -- the compiler/editor presumably picking and choosing whatever stories seemed appropriate given the collection's budget, audience, thematic relevance, etc. Without a single "Working Language" field in the author record, the software would have no way of knowing when to display each title's language and when to leave it blank. Anthologies, omnibuses and chapbooks commonly have the same problem.

Re: defaulting new authors' "working language" value to the language of the first entered title, the problem is that we commonly enter translations first. For example, the way Japanese light novels are handled, the vast majority are currently identified and entered when their English translations appear. Then and only then do we enter their parent titles/authors. Defaulting new authors' language to English would require constantly going back and changing it to Japanese, increasing the likelihood of error . Ahasuerus 13:47, 13 March 2021 (EST)

Thanks - I did run a query on the database comparing titles and child variants where the language differed, to get a rough feel for how valid the idea that the author language could default to the creating title record. (Using the logic that title_ids are sequential, so if a child variant with a different language had a lower title_id, then that would mean a new author would have been created with the wrong language.)

Annoyingly mysql seems to have lost that query from its history, but my recollection is that whilst the majority of cases, the parent preceded the foreign-language variant, there was a big enough proportion that that wasn't the case, that making that presumption would indeed be invalid.

Vaguely related: I had a variant of that query (which mysql has also lost from my history, grr) that generated some stats about which from/to language pairs were the most popular for titles. I wonder if that might be something worth turning into a stats/top lists report, maybe formatted as a table with "from language" on one axis, and "to language" on the other? ErsatzCulture 19:02, 13 March 2021 (EST)

Sure, more statistical data is always good to have! The nightly logic supports both HTML tables and SVG (like this page), so either one would be fine as long as the query doesn't kill performance. Ahasuerus 19:43, 13 March 2021 (EST)

Just as a note, even if the child title is after a parent, if the child uses a new form of the author name, the problem is still there (most books we add in Russian/Bulgarian/Japanese are translations and even if the parent is here, the new translation adds a new form of the name) so you need to add these to the list of the ones where the title ID is just earlier on the child. :) Annie 20:59, 13 March 2021 (EST)

Hold Up the Sky

See the note I added here - this is how I handle this situation. That effectively assigns a date on the contents and ensures that it is clear that the actual book is not checked anywhere.

I would usually NOT add contents unless I am absolutely sure but even then I may add the note if the contents is based on the publisher site for example and I had not seen an actual Contents page - collections with no contents show up on a report. But as a lot of people working the report seem to just import from other copies, it does not make much of a difference anyway... :) Annie 13:32, 12 March 2021 (EST)

OK, will bear that in mind for future submissions. I definitely wouldn't presume that older anthologies and collections have the same contents (which is why the final dregs of the Gateway ebooks of Lin Carter and Tanith Lee are anthologies and collections that I've been too lazy to double check whilst there were easier novels to do first ;-) but I thought it was a reasonably safe assumption for cases like this where it's just a different format from the same publisher within the space of a few months. But adding a note proviso seems reasonable and easy enough.

Sometimes a paperback will have an extra story - a way for the publisher to get people to buy the second version for example. Or just being nice and paying for an extra story so they can give the readers more content (yeah... probably the sales are what drives it but who knows). Ebooks are even worse - permissions for paper and ebook are still sold separately (And for audio so I have the same headaches there) so sometimes the publisher needs to add/remove a story in either direction. And sometimes longer stories just do not make sense on paper economically. Add also the "bonus stories" -- on ebooks and later reprints and later formats and what's not... and it can get weird sometimes And sometimes they just do it to mess with bibliographers. :) Annie 14:26, 12 March 2021 (EST)

And one note about the note here - Amazon puts Fantasy on both our type of fantasy and on a ton of Romance novels (think of what the word means in English)... and "Historical Fantasy" get slapped on half the historical novels out there. In this case "In a magical ancient Britain" in the description makes it obviously ours. But careful with overrelying on the identifiers from Amazon. Annie 13:49, 12 March 2021 (EST)

Yeah, if it was just the Amazon keywords, I wouldn't have bothered - I just added that as one additional piece of corroborating evidence. (I actually tweeted a few weeks ago about how daft Amazon keywords were, so I would never rely solely on them for a submission here.) ErsatzCulture 14:10, 12 March 2021 (EST)

Oh, I know. But because of the notes, I decided to stop by and share something the Fixer crew had learned the hard way ;) Some of the minor SF related nodes end up in what I call non-genre massacres - most die on sorting as obviously non-genre, then I kill most of the survivors after I take a closer look. Horror is what everyone thinks of when they think speculative or not (because we include only one of them) - but when you are scrapping stores, these differences in usage also come into play. So just sharing experience. Annie 14:26, 12 March 2021 (EST)

Introduction (Arthur C. Clarke Golden Age Masterworks)

Add a note to this explaining why it is the same in all the books and so on. :) And no worries on the dates, 00-00 are excluded when we check for before/after :) However, if that first one was not January, I would have used 00-00 instead of xx-00 on the title records - the pubs from 00-00 are either 01 or later so the 01-00 is a good date. But 03-00 won't be as 00-00 may be before that. We do not really enforce that - thus the "I would" and not "we do this". :) Annie 15:00, 16 March 2021 (EDT)

Uh, I'm going to play dumb: isn't it evident that it's the same intro in all those pubs, by the fact that the same ESSAY has been added to all of them, not to mention that I gave it a more generic title alluding to the fact it's used in multiple titles/pubs? It was the same situation here, with prior discussion on how best to deal with that here, where there was no mention of needing to explain things in a note.

If you know of another example of a shared intro with explanatory note along the lines of what you want, I'll happily copypaste and tweak if necessary... ErsatzCulture 15:46, 16 March 2021 (EDT)

Yes and no. It is evident now. 3 years down the road, someone not knowing the series may decide to "fix" it because they believe we made a mistake and they are different introductions. I added a quick note - feel free to edit. What is obvious to someone who had seen the book and what is obvious to John Doe are two different things. And a note never hurts. :) Annie 15:58, 16 March 2021 (EDT)

Ironically, I'm going to edit your edit because it's slightly misleading IMHO - although it's not immediately obvious when you read the first one, the intro specifically covers the 3 titles it's currently attached to, and wouldn't be relevant to any other Clarke stories that they might add to that pub series in the future, so "Non-book specific introduction" is a bit misleading. I probably should have just done what you told me to in the first place, rather than argue the toss :-P ErsatzCulture 16:39, 16 March 2021 (EDT)

:) Go ahead. :) I skimmed one of these quickly (as we had merged them, the other 2 are the same) and it looked generic enough. Which is why we need someone who had seen these to write the note when possible - I went through a few variants before settling on that one. ;) Then we won't end up with either a different one merged here (if Baxter writes another generic one for a set of books again) or these split. Thanks for fixing it!

Funny how often that happens. :) If I ask for something it tends to be because I had seen things going weirdly - I don't come up with more work for people just to mess with them. Annie 16:49, 16 March 2021 (EDT)

Contents and publication/title

Contents is tied to the publication, not to the reference title - so changing/merging/splitting the reference title does not budge it :) Annie 17:11, 17 March 2021 (EDT)

I know contents are associated with pubs at the database level, but I don't trust the web UI not to take any opportunity to punish the slightest mistake with making the editor do extra work, rather than just doing what I want ;-) Anyway, with the title varianting I've just submitted, I think this one should be finished. ErsatzCulture 04:03, 18 March 2021 (EDT)

FR 1305 Display a 'new editor' indicator on submission pages

Sorry about the extreme delays with FR 1305! Earlier today I implemented the requested functionality since it turned out to require a straightforward change within a single module. I would suggest taking the work that you did in 2019 (see the linked FR on SourceForge) and moving it to a new "Support Request" (SR). We use SRs as development vehicles for software modifications which affect the underlying code but do not change the way users interact with the system. How does it sound? Ahasuerus 17:31, 18 March 2021 (EDT)

Hi! First off, absolutely don't apologize to me to being slow on this (or anything else) - I'm acutely aware that I've been responsible for creating more than one bug ticket and/or volunteering to look into/fix something, and then disappeared into the ether when it actually came round to delivering anything tangible :-(

I had a look at the fix - a lot simpler than what I was proposing :-) I think the flaw in my thinking was that most of the stuff I've worked on in the past few years tried to very clearly separate out argument handling (whether command line args, CGI args, or environment variables) from database access and those from page rendering - so the idea that the code that renders the page would be able to get info about the user itself (rather than have it passed through as a function arg) didn't occur to me. (And also, if I'd read the original request on the Wiki, I would have realized that the user display stuff I was looking at (the 2 copypasted lines at the bottom of ~30 files) wasn't actually the part of the mod approval page that needed changing - oh well... ErsatzCulture 07:30, 19 March 2021 (EDT)

Yes, the "separation of concerns" principle. To the extent that the ISFDB software implements it, it's done by moving all common (emphasis on "common") database functions to SQLparsing.py. Certain other things would also presumably benefit from being centralized, e.g. it would be cleaner and faster to load commonly used user-specific information into a global variable like "USER" up front. Ahasuerus 10:35, 19 March 2021 (EDT)

The way that sort of thing is done in most (all?) of the frameworks I've used in the past few years is to pass around "request" and "response" objects - the latter includes all the stuff from the client/browser, and derived stuff like user details, and the latter is used for streaming the output, setting response headers etc. I don't know that there'd be any useful benefit from trying to do that within the ISFDB code as it stands, but I haven't really looked. ErsatzCulture 14:49, 20 March 2021 (EDT)

I expect that two objects or even one object along the lines of:

session.request
- session.request.user
  - session.request.user.moderator
  - session.request.user.all_translations
- session.request.cookies
session.response
- session.response.html_version
- session.response.headers
- session.response.navbar
etc

would be viable and much better than what we currently have. Certainly putting all output data in a Python object and letting a single set of methods convert it to HTML would be very beneficial, especially when upgrading to the next HTML version or when upgrading Python from 2.x to 3.x.

At the same time, I can see two issues with it. The first one is performance. The ISFDB code is currently optimized to query the database for the data that it needs and nothing else. It was a major concern back when our hardware was much less robust than it is now and it may not be an issue any more.

The second issue is simply the number of man-hours that it would take to do. Over the last year I have been very busy getting Fixer to the point where it would be stable and have a comprehensible specs. Once it is stable, we can start designing a new, Python/MySQL/HTML-based, OS-independent, editor-accessible version of Fixer. I expect that this project will continue to absorb most of my ISFDB time for the foreseeable future. Ahasuerus 14:53, 25 March 2021 (EDT)

Re. the testing stuff, I'm happy to create an SR, but I don't want to waste any of our time if it's not something there's at least a moderate amount of interest in doing. (I've been in work environments with testing zealots on one side, and people who didn't see the benefit on the other - with me being somewhere in the middle - and seen a vast amount of energy wasted and tempers raised, ultimately to no-one's benefit, so don't want to go through all that again for a "hobby" project.)

If it's any relevance, here are some tests I wrote in the past week or so, for my own code that uses ISFDB data. These helped isolate edge cases regarding authors and variant authors, perhaps in large part caused my lack of understanding of the subtleties of the ISFDB data model when I first starting putting together that particular code. Having those tests means that I'm now comfortable handling those edge cases correctly, and so the code that actually grabs the data (which I think is suboptimal, e.g. doing 2 DB queries when it probably only needs one) can be rewritten, and I can still easily run the tests and be confident that the same values are returned.

To what extent that sort of testing might be relevant or beneficial to the ISFDB site code is perhaps debatable - and the nature of my command line oriented tools/reports makes them easier to test than stuff that takes in CGI args and spits out HTML - but you'd be a better judge of that than me. ErsatzCulture 07:30, 19 March 2021 (EDT)

Unfortunately, my exposure to automated testing has been very limited over the decades. I agree that it would be good to have, but my knowledge is sadly lacking in this area and learning new things gets harder as you get older :-(

That said, my thinking has been that the ISFDB code would benefit more from automated browser testing than from automated command line testing. What do you think? Ahasuerus 10:46, 19 March 2021 (EDT)

Browser testing has it's uses, but IMHO it's costly in developer time for the benefit it provides. With ISFDB being a site with relatively Javascript and in-browser interactivity, I don't see personally that it'd be something to look at in the short term - but I don't know what any of the other users/editors/moderators with SW dev experience might think? ErsatzCulture 14:49, 20 March 2021 (EDT)

Let me make sure that we are talking about the same thing. When I say "automated browser testing", I mean that we could use a tool like iMacros to record a suite of test cases and then replay them every time the software is changed. For example, a test case may look like "Navigate to the Frankenstein title page and check if the generated HTML code includes a Python error line" or "Access the New Anthology Web page, enter a predetermined set of values, click the Post button, click the Approve button, check the list of SQL statements for Python errors, navigate to the Publication page of the new anthology, check that the data is displayed as expected". Is this what you had in mind?

One advantage of this approach would be that it's complete end-to-end testing. It doesn't rely on any particular version of Python, MySQL or any other tool. A significant disadvantage would be that it requires an automated Web framework, which may take time (and possibly money) to configure. Ahasuerus 17:18, 25 March 2021 (EDT)

I'll create some separate items further down this talk page to respond, as this discussion doesn't really have much to do with this item subject (mea culpa) ErsatzCulture 17:27, 30 March 2021 (EDT)

From my POV, it would be more beneficial to have some basic Python tests, that can (a) be easily run in different environments, most notably newer versions of Python 2.x, to give confidence that it'd be safe to upgrade, and (b) to be able to easily set up basic new tests, to provide a safety net of current known working behaviour, before they start trying to change anything, and run into unforeseen gotchas. Going back to the earlier comment about separation of concerns, the code as it stands is a bit harder to test than it might be, as it the output is only HTML. Whilst testing HTML (e.g. by using something like BeautifulSoup to parse it) is fine, it's nicer to be able to test pure Python objects - the bug and patch I just posted to SourceForge is a very simple example of what I mean; the tests could instead have parsed the HTML output of the full page, but that would be a lot more overhead, in the context of the particular change I made. ErsatzCulture 14:49, 20 March 2021 (EDT)

Re: Bug 767, Top forthcoming does not exclude all "generic" authors, IIRC, the decision to have the list of Top Forthcoming Books exclude pubs by "unknown" authors while including pubs by "uncredited" authors was a conscious choice as opposed to a bug. I would suggest asking about it on the Community Portal to see if there is support for changing the logic.

Re: "Various reports also exclude unknown-type authors, although they pick up on different ones", that too was a conscious decision. Different cleanup and statistical reports do different things. Certain author mismatches may be OK in certain situations, but not OK in other situations, so a single SPECIAL_AUTHORS_TO_IGNORE global variable may not work for all reports. It's entirely possible that our hard-coded lists of "ignored" authors could be profitable improved or consolidated, but we'll need to examine the implications first.

Re: SQL cursors, yes, they are safer than what the current code is doing. Eventually, I'd like to migrate everything to cursors, but it would need to be a coordinated effort. Doing it one query at a time would only splinter the code base. Ahasuerus 12:10, 26 March 2021 (EDT)

I'll create some separate items further down this talk page to respond, as this discussion doesn't really have much to do with this item subject (mea culpa) ErsatzCulture 17:27, 30 March 2021 (EDT)

Paper and ebooks

The right for e-books and paper books don't always move together. So it is possible for HoZ to have only the ebook ones while Gollancz retained paper rights. Who knows :) Annie 11:35, 22 March 2021 (EDT)

ISFDB code and HTTPS support

Ahasuerus wrote in an item above>

I expect that two objects or even one object along the lines of:

would be viable and much better than what we currently have. Certainly putting all output data in a Python object and letting a single set of methods convert it to HTML would be very beneficial, especially when upgrading to the next HTML version or when upgrading Python from 2.x to 3.x.

At the same time, I can see two issues with it. The first one is performance. The ISFDB code is currently optimized to query the database for the data that it needs and nothing else. It was a major concern back when our hardware was much less robust than it is now and it may not be an issue any more.

The second issue is simply the number of man-hours that it would take to do. Over the last year I have been very busy getting Fixer to the point where it would be stable and have a comprehensible specs. Once it is stable, we can start designing a new, Python/MySQL/HTML-based, OS-independent, editor-accessible version of Fixer. I expect that this project will continue to absorb most of my ISFDB time for the foreseeable future. Ahasuerus 14:53, 25 March 2021 (EDT)

The request/response stuff was intended just as an example; I wouldn't propose making any serious changes in that regard any time soon. Something I'd rather see in the shorter term is decoupling the core code from the CGI/sys.argv/environment variables paradigm - I recently shuddered in horror when I saw some code buried a long way down the request handling/response generation, that suddenly started checking sys.argv; that sort of thing IMHO should only be done at the __main__ level. Amongst other things, that might potentially open up the option of using WSGI thread pools, rather than creating a new process for each request in the CGI style, which may be more performant. (Although quite possibly not making much difference in the context of ISFDB traffic.)

Apologies if I'm speaking out of turn, but is there a plan or priority list for ISFDB infrastructure changes? I saw the brief discussion(s) about HTTPS a few days ago, and that strikes me personally as something that is going to be an increasing risk as time goes on. Beyond the issues users of some browsers already get with forms on insecure pages, in the past week or so, Google has switched Chrome to using HTTPS by default if no protocol is specified (albeit falling back to HTTP if HTTPS isn't supported). Plus, there's there's continued tinkering with Google Search, and I imagine sooner or later they'll start (more) aggressively punishing vanilla-HTTP sites. ErsatzCulture 17:59, 30 March 2021 (EDT)

Good points. Al was going to look into the HTTPS issue (as well as the Wiki upgrade issue) when we last discussed it a few months ago. I have sent him an email to see if he has had any luck. Ahasuerus 18:19, 1 April 2021 (EDT)

I have received an update from Al. He has been investigating, but it's more complicated than it looks. Ahasuerus 12:53, 5 April 2021 (EDT)

Bug 767: generic authors in Forthcoming Books section

Ahasuerus wrote in an item above:>

Re: Bug 767, Top forthcoming does not exclude all "generic" authors, IIRC, the decision to have the list of Top Forthcoming Books exclude pubs by "unknown" authors while including pubs by "uncredited" authors was a conscious choice as opposed to a bug. I would suggest asking about it on the Community Portal to see if there is support for changing the logic.

Re: "Various reports also exclude unknown-type authors, although they pick up on different ones", that too was a conscious decision. Different cleanup and statistical reports do different things. Certain author mismatches may be OK in certain situations, but not OK in other situations, so a single SPECIAL_AUTHORS_TO_IGNORE global variable may not work for all reports. It's entirely possible that our hard-coded lists of "ignored" authors could be profitable improved or consolidated, but we'll need to examine the implications first.

Re: SQL cursors, yes, they are safer than what the current code is doing. Eventually, I'd like to migrate everything to cursors, but it would need to be a coordinated effort. Doing it one query at a time would only splinter the code base. Ahasuerus 12:10, 26 March 2021 (EDT)

In turn

* Re. forthcoming & authors> I'll create an item on Community Portal

* Re. generic author lists> Yes, there's a comment in the attached version of isfdb.py that acknowledges there may be cases where some pages/reports need to use different author lists, in which case some sort of concatenation of sub-lists could/should be done as appropriate.

* Re. using cursors> Can db.query() do parameterized queries, or - as seems to be the case from the code/docs I've seen - is that only possible when using cursors? The risk of SQL injection attacks makes me shudder if the only way to use db.query() is to manually construct the query using traditional string manipulation - this is the main reason I used SQLAlchemy for my own Python code that uses ISFDB - I don't (with one exception) use any of SQLAlchemy's data model functionality, but I wanted to have a more robust & secure interface to the database than what I'd get doing plain SQL queries. ErsatzCulture 18:11, 30 March 2021 (EDT)

On the Forthcoming - you may also mention "restriction per author" - some authors take over half the 22 spaces when they do big reprints (thankfully the Potters are usually the same book but if they get all 7 one week, they WILL take 7 spots on the list)... Or I would mention it when you open the topic :) Annie 18:23, 30 March 2021 (EDT)

It will probably not be until tomorrow, as I want to find at least one example of the "uncredited" Disney books to link as an example... ErsatzCulture 18:47, 30 March 2021 (EDT)

Ask and you shall get: Raya and the Last Dragon. I am adding all of those - I know where to look for them ;) And there are the uncredited anthologies here for 2021 which we may or may not want on the marquee... Annie 18:57, 30 March 2021 (EDT)

BTW, on the subject of pubs that appear in Forthcoming Books - although I think this one will disappear in ~40 minutes - I'm not convinced this actually came out. If I search Amazon UK for the ISBN, it shows as "Temporarily out of stock", and the page on titanbooks.com for it that I'd scraped in April last year now returns a 404. (If you search that site for all Nina Allen books, it does show both tp and mmpb of The Rift and The Race, implying there's no reason why 2 different pubs of The Silver Wind shouldn't appear, if they do indeed both exist?) ErsatzCulture 19:17, 30 March 2021 (EDT)

It looked perfectly valid on February 1 when I added it with nothing to tell me it won't make it out - and I am sure it was marked pb because I checked a few times - Titan had been a pain so I triple-check theirs on both dates and formats. Some books do not make it no matter how carefully we add - although looking at the date, i came as 0000-00-00 (which is somewhat normal for UK books that early on - don't ask - there is a reason I do not add UK books that much in advance)... Amazon US shifted to 2079 which pretty much means abandoned usually. Titan are on a short list I check within a month of publication; the rest usually get checked a bit later (So January gets checked in April for example). Give it a couple of days to shake out and see if it may get out - and if it does not, 8888 it goes. Annie 19:29, 30 March 2021 (EDT)

Testing

Ahasuerus wrote in an item earlier:

Let me make sure that we are talking about the same thing. When I say "automated browser testing", I mean that we could use a tool like iMacros to record a suite of test cases and then replay them every time the software is changed. For example, a test case may look like "Navigate to the Frankenstein title page and check if the generated HTML code includes a Python error line" or "Access the New Anthology Web page, enter a predetermined set of values, click the Post button, click the Approve button, check the list of SQL statements for Python errors, navigate to the Publication page of the new anthology, check that the data is displayed as expected". Is this what you had in mind?

One advantage of this approach would be that it's complete end-to-end testing. It doesn't rely on any particular version of Python, MySQL or any other tool. A significant disadvantage would be that it requires an automated Web framework, which may take time (and possibly money) to configure. Ahasuerus 17:18, 25 March 2021 (EDT)

The problems I see with browser tests of this nature, based on my (probably out-of-date experience with Selenium):

* Take far longer to create

* Are far more flakey and subject to random failures

* If the test fails, their higher level nature means it's more developer effort to work out what caused the failure

* Take longer to run. (Especially annoying if you have a suite of tests and CI/CD.)

* The plugins they use end up being dependent on particular browsers or versions of browsers, making for a support nightmare. The benefit you might in theory get from being able to do cross-browser/device testing never seems to be easy to achieve.

* Often need you to update the markup you generate in order to make it easier to select particular HTML elements. In theory this is a minor issue, but a lot of the ISFDB code I've seen doesn't have properly validating HTML (e.g. unclosed tags), which makes me wonder if there's more work there than might meet the eye.

From my personal experience - which I wouldn't claim is universally applicable - I'd say the dev cost of browser tests versus unit tests is at least one order of magnitude greater, and I wouldn't be surprised if it's closer to 2 orders of magnitude.

In some cases, that can't be helped e.g. websites that are very JS-heavy, but I don't perceive ISFDB as being so. Given that the ISFDB model is simple HTTP page requests and responses (i.e. not lots of Ajax/XHR/Websockets stuff) and for the most part plain HTML (i.e. not some Javascript SPA), I feel most of ISFDB's functionality can be tested more efficiently using Python testing.

NB: it's possibly my personal experiences/biases are coming into play here - the vast majority of my career has been in *nix environments and old-style editors and tools like emacs, where scripting and command-line stuff is the order of the day, and (to my mind) the easiest way to do things, whereas anything involving GUIs, mouse pointers, clicking etc is a much more long-winded and painful way to achieve things. I suspect colleagues who live in IDEs like Visual Studio may see things differently? ErsatzCulture 18:26, 30 March 2021 (EDT)

My only relevant experience is with iMacros, which Fixer has been using for a number of years. It worked well in the early 2010s, but the free IE-specific version that Fixer uses is no longer supported and I am hesitant to move to a paid product since it may create barriers in the future. The IE version has been experiencing an increased number of technical issues over the last year, so I am considering migrating to another solution like Save Page WE, but, of course, it wouldn't address the larger testing issue. Ahasuerus 18:33, 1 April 2021 (EDT)

An Inch of Ashes

Can you check your scrapper? This is at least the second one with no number of pages (when Amazon UK at least has them) - I fixed the first one but it seems like your scrapper is misfiring so figured I will stop by and let you know this time instead. Annie 12:09, 31 March 2021 (EDT)

Also - the date you have on the book is more than a month off from the date on Amazon UK. As you are using Amazon UK as a source, the difference needs to be in the notes... Annie 12:10, 31 March 2021 (EDT)

Sorry, this one is in part down to inconsistent sources, and part down to me not spotting that. Normally my tools highlight when the sources are inconsistent, but for a messy pub like this one, they get listed as separate titles/pubs, and I have to manually "merge" the data into a single submission, and I screwed up.

The page count omission is that the first 2 sources I have (Blackwells and bookshop.org) didn't have it listed, and I failed to notice when I looked at the data obtained from Amazon and Waterstones.

For the pub date, Blackwells, bookshop.org and Waterstones all have 2017-05-18, vs. 2017-06-27. The publisher's site - or rather, the archive.org copy - doesn't have any info that might help indicate which is correct, so I guess it's just a case of pick one, and explain in the notes.

Apologies for screwing up on this one. The bad news is, that the other Chung Kuo reissues look to be equally messy in terms of what's been scraped, so I'll have to be extra vigilant when I get round to them.

Do you recall which was the title/pub that you fixed? It's probably another case of me doing a bad copypaste into the form, but I'd like to be sure. I had a look at the history on a few of my recent submissions, but I don't see any others that you've had to fix. ErsatzCulture 12:47, 31 March 2021 (EDT)

No worries - you should see what Fixer serves sometimes ;) I fully understand how it happens - I work off multiple sources often as well. :) No need to apologize - I would have just fixed it usually but as I know you are scrapping, I wanted to make sure you are not missing a format change or something...

Honestly, no clue. I will try to track it down later. I kinda remember I fixed pages number somewhere in one of yours in the last days (or price? I think it was pages) - so figured there is a scrapping issue. If it is a one off, I tend to fix and move on :) Annie 12:57, 31 March 2021 (EDT)

Looks like I missed the page count on this one. Doesn't look like I've got any excuse for missing it, as both Amazon and publisher have a page count that I scraped OK - probably a case of me just not paying enough attention 'cos I usually omit the page count on my ebook submissions, and hc/tps are much less common for me. (Blackwells has a slightly different page count, but I've learnt to ignore anything they say on that subject...)

I'd guess I have on average one or two issues with the scrapers per month, but that's mostly down to sites changing their markup, or having weird values that I hadn't even contemplated having to deal with. By the time it comes round to me making a submission, any mistakes or omissions are more likely to be due to human error on my part. ErsatzCulture 14:39, 31 March 2021 (EDT)

It is not a problem :) Fixer is better at finding paper UK books generally so I would usually add some of them before you get to them. Annie 17:04, 31 March 2021 (EDT)

(unindent) Another one with no pages number: The Forest of Stars :) Annie 17:59, 20 April 2021 (EDT)

Erk - sorry (again).

I notice the publisher page I linked to now has a different count - 320 rather than 336 - to Amazon (and Waterstones) from when I originally scraped the publisher page back in October last year. (The Amazon and bookshop.org scrapings were much more recent, although the latter doesn't have page counts.) Wonder if that maybe means that 320 is closer to the actual count, but the publisher hasn't bothered to update the vendor details? ErsatzCulture 06:41, 21 April 2021 (EDT)

The First Omega

Even bigger of a mess than you thought here. I think that the old ASIN was cancelled and a new one issued from Amazon UK but who knows... maybe pre-orders and/or promotional price stayed on the old one. In all cases, both ASINs are working on Amazon.com (with the usual for non-US books that cannot be sold in the States "The Kindle title is not currently available for purchase"). Updated the note. Annie 17:35, 31 March 2021 (EDT)

The GollanczFest@Home eBook Sampler

I tend to ignore the e-book samplers on the US side when Fixer finds them - if a user wants to add them WITH contents, I will approve them but I am not adding them and they do not really add value to the DB... So... either you care about then and you are adding contents, or this goes the way of the dodo. Let me know! :) Annie 19:25, 1 April 2021 (EDT)

Hmm, this feels like a flagrant abuse of moderator powers to me ;-)

* I had an earlier one of these submitted (plus a later edit) a year ago accepted by 2 other mods without issue.

* Are there any rules & standards/discussions that indicate that having contents makes a submission eligible, but not without?

I'm not going to disagree that sampler anthologies like this aren't particularly interesting - even though this one is free, about the only reason I'd ever "purchase"/download it personally is if it turned out that some of the extracts therein never ended up being properly published in their own right, for whatever reason.

If it was just a PDF/EPUB/MOBI that was available for download from the publisher's site, then I wouldn't have bothered submitting - but as it's a "proper" ebook available on multiple stores, with an ISBN and ASIN, it seems worth having a "stub" entry, even if only for having those IDs in the database? (That the ISBNs for this and the prior Gollancz sampler are out of sequence with other, "normal" Gollancz pubs and their pub dates strikes me as interesting and potentially useful, although that's probably just a reflection of how weird I am ;-) ErsatzCulture 08:29, 2 April 2021 (EDT)

Approved. Quite honestly, I consider it empty bytes and not a proper book but if you care enough, it won’t hurt us. It will just now sit in the “publications with no contents” report. As I said - I would have skipped it. :) Annie 08:52, 2 April 2021 (EDT)

Thanks. I wouldn't discount me ever submitting the contents - but I'm more likely to prioritize every "proper" anthology over this one (e.g. the Lavie Tidhar World SF that was published yesterday, several British Library anthologies past and upcoming, etc) ErsatzCulture 13:02, 2 April 2021 (EDT)

Exiles of Valdemar

Hello,

Any plans to add some contents here? :) Thanks! Annie 21:14, 14 April 2021 (EDT)

TBH, I'd forgotten about it - with the queue being so long for much of the past week, I'd slowed down on submissions to just stuff coming out in the next ~3 weeks, or fixes for ad hoc errors/omissions that I noticed by chance, and would likely forget if I didn't submit something there and then.

Will see about submitting something for this pub later today... ErsatzCulture 05:21, 15 April 2021 (EDT)

Well, it took 9 hours to approve... ;) Although I also noticed that you did not have your usual "will import" you use on omnibuses so I suspect you did AddPub instead of ClonePub by mistake and that is why it fell off your radar. Annie 12:54, 16 April 2021 (EDT)

Of Ants and Dinosaurs

Hi Ersatz, you have as verified an ebook edition of Cixin Liu's Of Ants and Dinosaurs as a chapbook, which is something that usually would run to no more than around 100 pages of a print edition if it was a novella. Today I bought the Head of Zeus hardcover and it runs to 249 pages: clearly a novel. I think this needs to be listed as such, not as a short fiction/chapbook publication. As the only current verifier of any edition of this title, what do you think? PeteYoung 15:36, 15 April 2021 (EDT)

Hi Pete, I've got no objections. The entered length was primarily based on User_talk:Sansanfeng#Request_for_help:_Cixin_Liu.27s_Of_Ants_and_Dinosaurs_.3D_.E5.BD.93.E6.81.90.E9.BE.99.E9.81.87.E4.B8.8A.E8.9A.82.E8.9A.81_.3F (item #42), but that was mainly based on the Chinese original(s). I think I had seen the reported page count for the print version, which seemed a lot based on my perception/recollection of the OverDrive ebook I borrowed/read, but obv. if you have a copy in hand, that trumps my vague memory.

Weirdly/annoyingly, I recall that the ebook version never showed up at publication date on Kobo, which meant I was unable to make use of their reported word count. It is there now, but for some reason is failing to show the word count?!?

FWIW, I see this upcoming US pub is reporting 192 pages, and the linked publisher page describes it as "a short novel" (and all but confirms it's the same translation as this UK pub), which supports a novella->novel change. ErsatzCulture 16:57, 15 April 2021 (EDT)

SubPress calling that a "short novel" does not always mean what one thinks it might. However - it is over 6 hours on Audible so it seems to be over 40K words indeed - not by much I suspect though. Pete, can you count words per page in the Zeus book? I have a few weird ones (not Zeus) that are well over 200 pages and still novellas. Fonts are not what they used to be... Annie 12:52, 16 April 2021 (EDT)

Of Ants and Dinosaurs

You verified Of Ants and Dinosaurs as a novella. This pub has now been issued by Subterranean Press in the U.S. under the title The Cretaceous Past, which I have purchased. I counted the length as approximately 55,500 words, or a novel. Given the number of pages, it's hard to imagine this story is just a novella. Since the Subterranean Press edition needs to be varianted to the UK edition, they need to be the same length. I'll change the U.K. edition to a novel, if you agree. Bob 21:07, 22 May 2021 (EDT)

After I saved this page, I saw the earlier note on the same issue. I'll make the changes. Bob 21:11, 22 May 2021 (EDT)

I asked the Subterranean Press publisher. He said he didn't remember the exact number of words, but that it was a bit over 40K. Bob 17:35, 23 May 2021 (EDT)

Clockwork Boys

Couple of issues with this one:

Careful when adding to pseudonymed authors - you added it as by Ursula Vernon. I tend to use ClonePub instead of AddPub exactly to make sure that this does not happen. If you use AddPub, please be careful which record you add to :)
You missed to set the format so it came in as unknown.

All fixed. I also added the US price while I was around (same ASIN). :) Annie 17:15, 11 June 2021 (EDT)

Gah, obviously I should know better than to make stupid errors like those :-( I've been trying to squeeze in as many edits today as possible, in the hope that they get accepted before the weekly database dump is generated tomorrow, so that I can in turn sync as much data up over the weekend as possible. I've probably tried to do too many edits today, and as a result got tired/lazy. Thanks for fixing those errors!

(Apologies also for the "burst" nature of my submissions the past couple of weeks - I'm in the process of moving out of London back up to the north of England, which has meant periods of 3-4 days with no time for ISFDB stuff, or even any online access, and then trying to catch up with my backlog, as-and-when I'm back online.)

BTW, how's the Fixer backlog? I can see you've been processing a lot of submissions, but I don't know the queue compares to "normal". I have picked up a few Titan Books pubs for early July, that I was surprised hadn't already been done, but I *think* they're all UK/rest-of-world releases, so I guess Fixer might not know about them? I'll probably push them through in a couple of days... ErsatzCulture 18:00, 11 June 2021 (EDT)

No worries at all. :) It was just a "hey, this is one of these authors - careful around them" reminder.

Fixer is a bit behind because I ended up in the hospital for a week early last month and then was mostly off for almost 2 more weeks while getting back to normal. So for the next few weeks, things will be fluid around what is added and when by Fixer. All paper June and July records are processed so if you have June/July paper records, Fixer does not know about them yet or they had landed under the wrong month. The rest of the formats are... waiting. However - the current month download is under way so some will show up when I get to processing them in a week or two. Annie 18:24, 11 June 2021 (EDT)

User:ErsatzCulture/Talk2021H1

Castles Made of Sand

Changing types

Heaf of Zeus (Apollo)

Master of Sorrows

Incomplete

Empire in Black and Gold

Adam Blade books

The Harpy

Pattern Recognition

Alexandria

The London Reader, Volume 17

Drogan the Jungle Menace

Larry Correia

Kyron, Lord of Fire

The First Omega

Silexa the Stone Cat

Hold Up the Sky

Introduction (Arthur C. Clarke Golden Age Masterworks)

Contents and publication/title

FR 1305 Display a 'new editor' indicator on submission pages

Paper and ebooks

ISFDB code and HTTPS support

Bug 767: generic authors in Forthcoming Books section

Testing

An Inch of Ashes

The First Omega

The GollanczFest@Home eBook Sampler

Exiles of Valdemar

Of Ants and Dinosaurs

Of Ants and Dinosaurs

Clockwork Boys

Navigation menu

Search