User talk:SnowFire

Beeswaxcandle (talk) 05:00, 7 September 2021 (UTC)

The Apocryphal New Testament
Hi SnowFire,

Thank you for your work on The Apocryphal New Testament!

I've only looked very superficially at the pages, but overall they look very good. And because they do, you should probably use the page status widget (the radio buttons down by the edit summary field) to mark the pages as "Proofread". The basic workflow here is that a first person transcribes the text (correcting OCR errors etc.) and does at least basic formatting, making the pages "Proofread"; and then a second person checks over them and corrects anything the first person missed, making the page "Validated". The software enforces things like not being able to go straight to "Validated" status and that the same person cannot do both steps. The reason for the division is that even the most experienced and eagle-eyed proofreaders will miss a few things, so having two people take a pass over a page gives a much higher quality result. But if you're reasonably confident the text is as correct as you can make it and the basic formatting in place, you shouldn't hesitate to mark it as "Proofread".

On specifics, I notice you've added section markup every time there's a heading in the book. This isn't necessary as the section markup is only used when we need to transclude parts of a Page-namespace wikipage into different mainspace wikipages. All these short letters of Pilate and Herod are generally all going to be transcluded onto the same page (think roughly chapter-sized chunks), so the section markup is mostly going to be just unnecessary complication. Labelled section transclusion is a relatively advanced technique so it's not something we usually suggest new contributors try to do themselves. It does no harm (feel free to experiment if you want), but you can spare yourself that additional complication.

But what you might want to do, is add a template to the "Footer" box when there is a footnote in the page. It's not a big deal if you don't, but it's the established practice and what other contributors are going to expect to see. Xover (talk) 07:15, 23 February 2022 (UTC)


 * Thanks for the information. Didn't know about how the widget worked & using smallrefs.  The section thing, I was just copying from a different segment on Wikisource that did happen to use section titles in its style, so that got caught up along with.  (Also, I see that you created the index & did a bunch of the other transcriptions from the book, so nice work, yourself.)
 * Incidentally, while I'm here: where exactly did the original scan of this book come from, and how did Wikisource get the files? SnowFire (talk) 15:57, 23 February 2022 (UTC)
 * The Internet Archive: Xover (talk) 16:25, 23 February 2022 (UTC)

A few helpful hints
Nice work on the transcription of The Prose Edda (1916)! A couple things to note:


 * Template:di can be used to format the dropping initial at the beginning of a chapter.
 * See my edits on 19 for how you can handle page hyphenation between pages the Wikisource way.

Happy editing! PseudoSkull (talk) 12:59, 28 June 2023 (UTC)


 * Thanks for the comment. I should note that there's already an existing OCR of this work on the Internet which greatly speeded things up; while said OCR is definitely not perfect in parts (fixing it was one of the main goals) it did provide a great starting place.
 * The bit about hyphenation is interesting, didn't know the software combined across page boundaries that way. I was definitely optimizing for the "combined" view cases rather than the page-by-page scans, most notably in the various excerpts of poetry - if a poem stretched across two pages, I'd sometimes just ignore the real page boundaries to avoid creating an accidental, misleading stanza break from the page gap.  But maybe there's a Wikisource-compliant way to do that as well?
 * Actually, on the above note, is the mobile view bugged, or have I made errors somewhere? (Insert both of us tracking down the mobile display bug here and reporting it to the Scriptorium.  That was written two days ago.).  Okay, back to the present.  There's still some alarming issues with the poetry, both combining stanzas despite intentional linebreaks, as well as creating phantom new linebreaks due to a page boundary that was not intended.  It seems that if I have something like:

(Page 1)
 * poem line
 * poem line

(Page 2)
 * more poem line
 * more poem line


 * It works fine in desktop view, but the nop generates a pagebreak in mobile view. That seems bad to be inconsistent.  I could remove nops, but that also seems wrong?  There really is a line break at the end.  I guess "if using indentation don't use nop" might be the moral, but I'd want to check first.  The other problem is that mobile view seems to ignore linebreaks in a single page:

(Page 1)
 * poem line 1
 * poem line 2
 * poem line 3


 * poem line 4
 * poem line 5
 * poem line 6


 * There is a gap between 3 and 4 in the desktop view, if rather too small of one for my tastes. Semantically, these are the same things, dictionary list (dl) tags with dictionary definitions in the middle, but desktop CSS has:

dl { margin-top: 0.2em; margin-bottom: 0.5em; }
 * While mobile only has:

dl { margin-left: 1em; }
 * Is this "intentional" that there's no gap at all between consecutive dl tags? You can see it if you search for "Heroes tread Hel-way" at mobile and The Prose Edda (1916 translation by Arthur Gilchrist Brodeur)/Gylfaginning desktop views.  Mobile, with no margin top / bottom, just smooshes the stanzas together.  Is there a way around this?  (Note that when I first started transcribing, I consider using tags, but after I saw the easy colon indentation in use elsewhere, I figured I'd stick to that...  hopefully the answer isn't "redo everything with a poem tag."  SnowFire (talk) 02:25, 1 July 2023 (UTC)
 * (EDIT): Well, you might have to use the source editor to see what I'm referring to, because looks like the editor wants to glom everything together. There are line breaks in the real source.  SnowFire (talk) 02:28, 1 July 2023 (UTC)

Northern languages and literature
These works are started, if you are looking for something to tackle after the Prose Edda in the same vein:
 * Index:Icelandic Primer (1895).djvu
 * Index:An elementary grammar of the old Norse or Icelandic language.djvu
 * Index:History of the Literature of the Scandinavian North.djvu

--EncycloPetey (talk) 22:09, 3 July 2023 (UTC)
 * Thanks for the offer, but... probably not.  I'm more of a Wikipedia editor (and my focus actually isn't even normally Icelandic / Old Norse stuff), and there's plenty of Wikipedia-side editing & updating to do first just on the existing project.  (Also, part of why I did the Prose Edda was more anger at Amazon selling an already existing if flawed OCR online for a dollar while presenting it as someone else's translation...  wanted to fix it out of spite if anything.  Special circumstances, basically.)  SnowFire (talk) 00:10, 4 July 2023 (UTC)

The Prose Edda
It's better to have stable page locations than to move pages on a whim. This work has multiple issues, including that the subpages are not using the same base page. For example, you placed the prologue at The Prose Edda (1916 translation by Arthur Gilchrist Brodeur)/Prologue.

However, your move today created double redirects from multiple other pages, which beaks internal linking. People will see the redirect target instead of the page. And this matters because our works are not only linked here but also from external sites such as schools and libraries. Disambiguation here allows for the possibility that a work might have been published in more than one edition, and we might host more than one edition of a work. For some works, we actually do this already. --EncycloPetey (talk) 19:40, 7 January 2024 (UTC)


 * My thoughts exactly, but... the work started out at the old title, and it was the earlier bold move with no talk page notice, no explanation, no cited policy, no nothing, and also broke links.  I guess I should have more aggressively reverted originally that move?  Anyway, I have to assume that redirects will handle everything mostly fine regardless, right? SnowFire (talk) 19:45, 7 January 2024 (UTC)


 * Also, to be clear, the subsections were also "misplaced" which was an artifact of me uploading the original PDF on an overly long title (I didn't realize at the time at how Wikisource likes to use the title of the PDF). It's basically harmless IMO.  All I did was change the "introduction" page's location, to my knowledge (as did Martensas's bold move).  SnowFire (talk) 19:47, 7 January 2024 (UTC)
 * As one final point, if the goal truly is to avoid redirects, all of the links I personally placed are to just (1916) (which was the accurate link at the time), the original location of the page.  Martensas's move is what broke those and required redirects, so if your goal is to avoid redirects and honor old links, then the old title is probably best.  That said, if you want to make a case that Martensas's proposed title is better, I'm all ears.  I actually agree about the possibility of multiple editions of the same work, but isn't that an argument to use date instead?  Since if a later edition of the same translation is scanned in, the date would then act as disambiguator.  We only need both if there are two separate 1916 editions by different translators, which isn't true at the moment or to my knowledge.  SnowFire (talk) 19:55, 7 January 2024 (UTC)
 * But the software on Wikisource auto-generates links back to the base page. So the subpages you created lead to an incorrect base page name, which them requires redirects just to clean all the redlinks that would be auto-generated.  We want all the subpages to use the same page name as the main page for their base page.  You did not do this.  To make everything work correctly, the base page would nee to be "The Prose Edda (1916 translation by Arthur Gilchrist Brodeur)".  And no, date alone isn't the "policy".  For example, there were two first editions of the novel Moby-Dick, published in the same year.  So there are non-trivial situations where the date alone is insufficient.  And when a work is a translation, and that translation might exist in multiple editions, we need the author's name to distinguish it from other translations of the same work and the date to distinguish the edition of that translation.  Because of the number and variety of situations of possible needs for disambiguation, we have no policy-mandated hard rules for disambiguation.  And thus, there is no rule requiring the page to be moved now. Though we might need to move the subpages to correct the auto-linking. --EncycloPetey (talk) 20:05, 7 January 2024 (UTC)
 * The auto-linking all works correctly, I assure you. Wouldn't have put it in published until I made sure that all the "next" and "previous" section headers all worked.  They do.
 * If you'd rather keep the base page to include the translator as well, it's fine. I just wanted to see an actual argument, and per above, when I asked on the Discord, I was essentially told "if it was moved without a reason, it can be moved back without a reason."  You've provided a reason, so sure - it's fine.  (Just would be better if there was a policy to consult, since I was essentially told "the policy is currently uploader's whim", so I figured I might as well use it).  SnowFire (talk) 20:13, 7 January 2024 (UTC)
 * (ec) Though we probably ought to have some kind of policy to cover the 99.9% of "normal" titles, and not have no rules just because a marginal .1% of titles share the same author and year, or a plethora of other wackadoodle scenarios. I think the sensible thing to do with "edge" cases is to find some consensus to a workaround on a case-by-case basis, but leave the site with a mostly consistent pattern. And we should try to make the edge cases as close to consistency as possible, even if still not 100% consistent. We should be as stringent about titles and style as Wikipedia or Wiktionary, and even they can't get themselves 100% consistent... But, while WP's data structuring might be 95% consistent, Wiktionary's 98% consistent, we're more like 2% consistent, which is pretty awful... I'll be writing an essay about it soon (partly encouraged by SnowFire themselves) to explain my idea of titling consistency, which will incidentally be far less lengthy than the US tax code. PseudoSkull (talk) 20:21, 7 January 2024 (UTC)
 * It's not the "next" and "previous" links in the header that I mean, it's the auto-generated links that appears at the top (above the header, under the page title) that link to the base page. These only work because of lots of redirects that wouldn't be necessary if all the pages shared the same base page (as they should).  If you believe you can come up with a naming scheme that covers 99.9% of titles, you're welcome to propose it.  There are a larger percentage of unusual cases here than you might expect.  In part because some titles are popular, some titles are frequently reused for different works, different translations, or editions; and we also have to contend with articles within journals and newspapers as well, which can also share titles with books.  Consider for example, that a published book review is usually under the same title as the book that it reviews.  There have been proposals in the past for a general naming scheme, and every such proposal has failed.  There are too many variables, too many factors, to consider even for the standard cases.  Look at a disambiguation page like Poems or To — to see just two titles with very complex usage.  --EncycloPetey (talk) 21:50, 7 January 2024 (UTC)