User:Alien333/Opinions

Opinions

Some random and debatable opinions of mine^[1], if you've got time to lose^[2]. Most of it is closer to a rant than anything else.

Reading

I think that here, the ultimate goal of technological optimization is to proofread without reading.

We read slowly, and we only need to read when human input is needed.

Most of that human input is unnecessary, such as

• OCR correction, for most parts, but blaring problems can be seen without reading the text
• running headers
• start/end's in {{ppoem}}, work-in-progress but I've got that.

As time passes, we make new gadgets and other things, and in the end human input can be reduced to a very small quantity.

It can reasonably just be looking at the page to ensure there are no blatant errors, invoking a template or two and pushing a button.

A lot of the formatting can also be factorized, for example a .wst-rh-center { font-style:italic; } can save a lot of time^[3].

Some may say that it is a bad trend to not really read, and that it's silly to not read what we prepare for others to read^[4].

"Wikisource's ultimate goal is essentially a boiling of the ocean problem"^[5], and I think any opportunity to go faster^[6] should be taken.

Telling

In a nutshell: when someone makes a mistake, tell them.

It takes at least a few years to know every single nook and cranny of this place.

We all sometimes make mistakes.^[7]

When someone makes a mistake, someone more "clueful" in that aspect that is passing by corrects it.

Most of the learning we do is not by reading.

We do read some things in WS/Help namespace, to get the basics.

But most of the details we learn by being told (partly because of that, but it's another story).

People have a tendency, when they see a mistake, to correct it.

This is good, it is how wikis work, but I think those people ought to tell the editor, or I don't see how they can learn from their errors.

The only options left^[8] to new people are to:

• watch everything they've done and to check when someone corrects something they've done
• spam random page and look at what templates people use, hoping it'll be useful at some point
• follow every "see also" link in doc subpages to find related but more obscure templates (such as Help:Templates → {{gap}} → {{phantom}})

If someone by some chance didn't, they'd have to wait for someone to tell them, and that might result in their never knowing.

I am aware that we already do a lot of it, I am only saying that due to our reliance on it we must do it as much as possible.

Possible objections and my replies

• It's a bother: It takes at most a few minutes, to check the page history and leave a hint on the corresponding talk page.
• Inactive editors: There would be the inconvenience of inactive editors, but that can be avoided by notifying only for edits more recent than, say, a month.
• Telling someone twice: There would also be the risk of telling someone something they've already learned, but I'll say any day that better be told twice than none.

Documentation

I think we have a problem with documentation.

We have some help pages, sure, but nearly no one actively takes care of those^[9], and there are many things that should be in there.

The example I like to take is the Glossary.

You might tell me that as of now it's fine, and includes most of the basic items, but I and others have done some work on it.

Here is what it looked like before I touched it, and what it had been for the last four years.

Even entries like "Index", which is quite important, were just left incomplete.

Another related issue is templates.

The /doc subpages are in most case good, but they're hard to find.

We have a great lot of them, a nice help page that doesn't cover a lot of them and a category tree that is at the very least puzzling.

Have you ever tried to differentiate Category:Typography templates, Category:Formatting templates and Category:Content templates?

Granted, they are not the same, and if you look at their contents you can understand what they are supposed to contain, but it's not that easy to locate a specific template.

For example, if you were to try and look for something to change the orientation of content, where would you look?

Experienced users might answer that question^[10], but it's not very newcomer-friendly.^[11]

Quite a lot of templates are also not in any of these categories, or in categories that make them hard to find, such as {{ppoem}} that was until recently^[12] only findable in Category:Experimental templates^[13], which I call counter-intuitive^[14].

Even for templates which are in appropriate categories, such as {{phantom}}, it is quite long to scroll through the 494 pages, checking each one^[15] that by its name looks like it could be it.

There is also a lot of overlap between those categories, and they are very broad, which means that it boils down to scrolling through the list of templates.

I would advocate for more, smaller categories, that really reduce the size of the set to search in, and that would be defined through their effects^[16], such as "Gap templates", "Poetry templates", "Color templates", "Font size templates", and so on.

Anyway, all of that to say that we lack sufficient documentation, especially regarding templates, and that I think we ought to make it better, to reduce our reliance on hearsay.

I am neither the first^[17], nor the second, to think that.

There have apparently been a number of attempts to fix this already, and they've more or less all dwindled to nothing.

Still, I hope to change something and not just be like all those editors that eventually left before finishing^[18].

(sigh) I really should get at it.

Spellcheck

It has been said, in a few places here and there, notably in WS:MC, that spellcheckers should not be used to proofread/validate due to their not knowing old spellings.

I personally disagree with that, at least for spellcheckers that have the "Add to dictionary" button, such as Firefox's.

It takes a while^[19], but progressively you can teach it antiquated orthography, so that it only signals real errors.

If OCR puts the word "Bocoming", which from a distance looks quite like "Becoming", it could be easily mistaken, while if you use a spellchecker it is visible.

To be fair, I do have to admit that there is a bit of an inconvenience with names (of persons), as these words are not part of the language but are not errors, but I think it's a small price to pay for instant signaling of unknown words.

Why

Some have asked me: Why?

They^[20] usually meant "why spend time to contribute to Wikisource", but it always reminds me of another linked question: How do we choose our texts?

Why transcribe this book because of this other one?

Some say that notability is important, that notable works should be done first.

Apart from the whole definition problem this entails, WS:GNG is a redlink, thank God, and it wouldn't help anyways.

Having it on WP seems to have given more debate over its interpretation than anything.

This place, intentionally^[21], is not ruled by notability.

Granted, not deleting works for not being notable doesn't keep us from trying to steer the wiki towards notable works, and I'm not making a witch hunt on notability, just saying that it is, for most of us, not truly our main concern.

As a last-ditch argument for that, look at {{New texts}}.

Most of the time, there will be between one and three arguably notable works on it.

That makes 12.5% to 37.5%, which is not a majority.

Moreover, notability is disputable because relative, as we were reminded by the {{Welcome}} image change discussion recently, in which about a third of us^[22] thought that a picture of G. Eliot should be in that template because she is notable enough to be quickly recognized.

I think that's inexact.

Sure, most of the contributors to this place know her, but we're pretty not representative^[23], and I think that if a poll was made in English-speaking countries, asking people if they would recognise the woman on it, most of them would not^[24].

So, not notability, as far as I'm concerned.

Another, much linked criterion that has been advanced is usefulness.

That is, again, completely relative.

What's more useful: a 1923 novel, a 2024 decision of the U.S. Supreme Court, a 2024 British law, a 1909 poetry collection, a 1893 magazine, or a 1866 poetry collection?^[25]

They are all useful, but to whom?

I advocate for usefulness to the general public, meaning everyone, and everyone is not^[26] simplifiable to a single choice, and therefore all has a usefulness, that we can not assess.

According to me, usefulness cannot be mapped to numbers or anything comparable.

If it can't be compared, then you can't decide on its basis, and thus I say that usefulness ought not to be our primary motivation^[27].

The reason that I would understand most would be choosing the books we would like to read, considering we would do it as we go.

I came with that opinion, though I'm doing it less these days, and I still respect it.

For myself, I do some Pulitzers to help the people focusing on them, I did Anna Karenina because I stumbled on the unfinished index, and I'm doing the Poems^[28], though it might seem obsessive^[29], out of an unreasonable liking of running gags^[30].

My answer is based on another related question, that we are asked mostly by ourselves: When?

Implying: When will we be over?

Most of the time, the answer we give is "never".

I'm not so sure about it.

WP will not be over, but we may, at least since WP:COMPLETE's arguments don't work here.

Firstly, we don't have that whole quality shtick.

On WP, a finished article means featured, that has probably undergone at least a few hundred revisions.

Here, finished means two people are satisfied with it.

Also, we do not need to update texts, as can be seen from our policy of protecting validated texts.

A finished text here is truly finished, and in most cases a definitive version.

We also have a much more reduced scope, as although we take every book, we take only books, not everything like WP.

More books are printed with time, but there are more of us as well, and we get more and more efficient^[31] as time goes, as we make new templates/scripts, update old ones, etc.

I believe that we can keep up, and even go past, the publishing rythm, and that therefore we will be over.

Once we admit that, the question of Why becomes very relative, as we'll do everything eventually, so it doesn't matter what we do first.

All of that to answer why with a classic: why not?

Notes

1. with an absurd quantity of footnotes for not really necessary information. I am usually more of a parentheses user, but refs clutter the text less, even though it's a bit awkward
2. and you probably do, if you landed here
3. it is a bit ridiculous that I am taking time that I could have spent proofreading to write on how precious time is
4. I came here thinking, and was probably not the only one, that I would read along as I went
5. as SnowyCinema puts it
6. without noticeably reducing the quality of the texts
7. This is only about about mistakes due to someone not knowing something, such as misusing a template or putting the wrong characters, not about things like OCR errors.
8. from my experience. At least, these are the three I did
9. and despite my complaining about it most of the time I don't either
10. although I suspect a part of them can't, because they didn't learn through the category tree, they got told, so they never had to fumble in it
11. I once tried to find {{em}}, and was unsuccessful, so I recreated it.
12. before I was so outraged at its absence from Help:Poetry that for once I got up and actually fixed what I was complaining about
13. or of course through hearsay
14. given that it's currently the best option for poetry. I vaguely remember some people saying they don't agree with this a while ago, if you're one of them I would like it if you could explain me your reasons
15. since when you're looking for a template you don't know the exact name
16. or main effect, to avoid overlapping. For example ppoem could be put in "Gap templates" because of :, but the main point is poetry and therefore it should only be in that category
17. and that was in 2012
18. thinking of Zoeannl's monumental Proofreader's guide. So huge, and yet so many red links left.
19. I have 3967 words and counting in my file
20. non-wikimedians
21. or so I like to think, or hope
22. stat from Xover's closure
23. take technological literacy, at any rate. I have a growing feeling that at least half of us, and that's a conservative estimate, have coded themselves a script/gadget/module/thingy and can write in multiple programming languages.
24. although it's true that this is a mere supposition and therefore doesn't have much weight
25. not including the last one as I'm unsure as to how to categorize it.
26. hopefully not yet. I think we're not robots or clones yet.
27. not saying is not because it is for some people, I am just saying that what you consider useful not everyone does, and making a decision on a personal criteria seems suboptimal to me
28. Most of those "Poems" popping up in New texts since february 2024 are mine
29. did 41. aiming for the 100.
30. and also a mild interest in closing the gender gap on Poems, though I've finished that a few months ago
31. at least it looks like edits grow faster than active editors