#dev 2024-04-02
2024-04-02 UTC
# 
[aciccarello] Shouldn't it automatically be deployed?
# 
btrem You'd think. When I try to view it with the netlify.app subdomain, I get 404.
# btrem So maybe the problem is not the deploy but something else.
# btrem sighs
[Jan_Tuomi], saptaks, superkuh, [lcs] and [Scout] joined the channel
# 
Loqi It looks like we don't have a page for "best way to mark up poetry" yet. Would you like to create it? (Or just say "best way to mark up poetry is ____", a sentence describing the term)
# 
[KevinMarks] Interesting, I do iy the other way round. A stanza is more like a paragraph for me
lazcorp joined the channel
# 
lazcorp even when it gets into concrete poetry such as https://www.lazaruscorporation.co.uk/artists/hardeep-sangha/poetry/some-fink-at-a-downtown-dive
# lazcorp MDM also suggests <br> for line breaks and <p> for stanzas/verses: https://developer.mozilla.org/en-US/docs/Web/HTML/Element/br
# [KevinMarks] Why are they using <br /> instead of <br>? We have up on XHTML2 a long time ago
# [KevinMarks] *gave up
# lazcorp Another example from my site (not my poetry): https://www.lazaruscorporation.co.uk/artists/hardeep-sangha/poetry/no-more-protection
superkuh, Guest6 and to2ds joined the channel
ttybitnik joined the channel
# [lcs] Did we really abandon the principle of closing tags? This is outrageous
# [KevinMarks] 'closing' tags in HTML is a weird XMLism
# [KevinMarks] I am annoyed by my HTML editor in VSCode turning `<p>` into `<p></p>` when I type it as that breaks formatting when I do it in the middle of a block of text I'm editing.
# [KevinMarks] lots of HTML elements don't need closing as they auto-close within contexts - `<p> <li> <dt> <dd> <td>` for example
mretka and lazcorp joined the channel
# lazcorp all the examples at https://www.w3.org/TR/2011/WD-html5-author-20110809/the-img-element.html do not use the closing slash at the end
# [Jan_Tuomi] `<p>` and `<li>` without the closing tags are cursed and you cannot change my mind
# lazcorp And as a double-check, https://validator.w3.org/nu reports <img .... /> (img element with a closing slash) as an info line with advice not to use the trailing slash
# [KevinMarks] `<p></p>` is cursed because it can exit the context for the `<p>` and break css formatting based on `p {}`
jacky joined the channel
to2ds joined the channel
# [KevinMarks] if you have something like `<main><h2></h2><p>text<p>more text</main>` then the first `<p>` is closed by the second one, and the second by `</main>` so you can style with `p
{max-width: 70ch}` for example. If the second `<p>` becomes a `<p></p>` then it will fall back to `main`'s formatting.AramZS joined the channel
# redblobgames I find it easier for my pea brain to reason about it when everything is closed explicitly
[campegg] joined the channel
# [campegg] Personally, I'm on team </p>, but if you're consistent with whatever you choose, the exit context thing shouldn't be an issue, right?
# btrem I always explicitly close my elements.
# btrem I don't know this to be true, but I suspect that the reason for self-closing tags is because html elements were originally created as dividers, not elements in the tree sense. So <p> was the equivalent of a pilcrow: a new paragraph begins here. <hr> worked the same way: a new section starts here (html5 makes that explicit).
# btrem And when the sgml folks got involved, they had to make it work in the tree sense. Easiest way was to enable shorttags, so <p> didn't need </p>; and declare <hr> to be void.
# btrem No way to verify. But it's plausible, and explains certain unfortunate "features" of html.
# [campegg] Huh, I did not know that. So was the intent to have something like `text text text<p>text text text` as the representation of two paragraphs?
# btrem I *don't* know that.
# btrem But I suspect that is the case.
# btrem Which would explain why orphaned text is allowed in html. But was not allowed in the failed *x*html.
# btrem xhtml being an attempt to fix the problems with html.
# [campegg] Fair assumption. And my response was poor phrasing on my part.
# btrem Well, this is a touchy subject. I actually think implicitly closing tags was a mistake, as it made learning html much more difficult than it needed to be. If you could search the old ciwa* usenet groups, you'd find many a post asking why a validator claimed a mismatched </p> tag when no such tag was in the source code.
# btrem And implicitly closing tags has delayed the introduction of useful features. I have direct experience of this.
# btrem And for all we know, it prevented features altogether.
# redblobgames I think you needed the tree view of things for CSS to work
# btrem Exactly so.
# redblobgames I end up writing my pages and blog in xhtml, and converting to html with a post-process step.
# btrem I think there's a conflation of the misguided notion of sgml purity with the practical need for an easily understood parsing model.
# btrem So as I said, this is a touchy subject here, to the point that it can become hostile environment, but I think there /were/ problems with html as it was originally created. The shorttags "feature" being one. And I sympathize with the desire to fix them, even if efforts to that effect were misguided at best and often much worse.
# btrem And looking above, I mistakenly said "...the reason for self-closing tags is because..." when I meant "the reason for implicitly closing elements is because...."
jeremycherfas joined the channel
# btrem [campegg]: didn't mean to jump down your throat. Just wanted to make it clear that I have no way of knowing why <p> was created as it was. Just a hypothesis. Because these things have a way of taking on a life on their own.
[Paul_Robert_Ll] joined the channel
# [Paul_Robert_Ll] I wonder if MDN is using Prettier to lint their examples, I think that enforces the closing slash
# [Paul_Robert_Ll] https://jakearchibald.com/2023/against-self-closing-tags-in-html/
# [KevinMarks] the modern HTML spec does define a consistent way of creating a tree from the implicitly closed tags.
# [campegg] btrem, all good! It was poor phrasing by me 🙂
# [aciccarello] Link previews are my main use case for metaformats
# btrem [KevinMarks]: it does. And that's a solution to the implicit closing problem that is obviously far better than xhtml. Which, for the record, I thought was a failure from pretty much day one and never once used.
# btrem To put it another way: xhtml was a very wrong solution to some very real problems.
Guest6_ and lazcorp joined the channel
to2ds joined the channel
# [aciccarello] Yup, one of the many warts of markdown
# [aciccarello] Personally, I like markdown as a plaintext format but there are lots of rough edges
# [aciccarello] _...ducks out of the markdown debate_
# btrem checks
# btrem 2021. So three years.
# btrem I didn't know until very recently that two spaces at the end of a line made a <br> element. So maybe, yeah! :-D
# btrem My text editor of choice strips extra spaces at the end of a line, so there was no way to accidentally stumble on that "feature."
# 
[qubyte] If markdown doesn’t work for you, make your own! https://qubyte.codes/blog/marqdown
# redblobgames Thanks to Igalia!
to2ds joined the channel
# sknebel [snarfed]: https://snarfed.org/2024-04-02_52710 has a weird link on the image and looks very empty when federated to mastodon ;)
[aciccarello] joined the channel
# [aciccarello] Beautiful photo though
# [aciccarello] New channel #indieweb-bikeshed /sarcasm
# [aciccarello] Safety first
[tantek] joined the channel
# 
bacardi55[m] @btrem (https://matrix.to/#/%40irc_libera_btrem%3Amtx.bacardi55.io) In case you don't know, you can use a \ at the end of the line in your markdown to add a <br> (like you, I remove all spaces at the end of the line, so that's the way it works for me)
# [campegg] Something has gone weird with my DigitalOcean droplet… starting about 40 mins ago, Python started running at 100% CPU usage. A reboot and a hard shutdown/restart didn't help. Any ideas on how I can track down what's happening? I haven't touched anything on it for over a week now, so it's not new code or anything like that.
# [campegg] (And sorry, I know this is more system admin and isn't directly IndieWeb-dev-related!)
# [aciccarello] [campegg] did the droplet have an automatic update or something?
# [campegg] [tantek] yeah, I think it might… I like -server better than -frontend/-backend; the former is more sysadmin, while the latter two both roll up into building sites
# [campegg] (👆 re: -dev vs -server distinction)
# bacardi55[m] Can you see what python script is running ? (eg with a ps faux | grep python)
# [campegg] [aciccarello] not sure… will check
# [campegg] That's a good question to2ds
# [campegg] Thanks for the tip bacardi55[m]! It looks like a cron job gone awry that's doing to the damage
# [campegg] *doing the damage
# [campegg] Now I just have to try and work out why it's gone rogue after running perfectly for months
# [campegg] bacardi55[m]++
to2ds joined the channel
barnaby joined the channel
# [campegg] to2ds it looks like a cron job (that runs a Python script) that just isn't finishing, then the next one spawns a new python process and so on. I don't know why it just started happening, though… there hasn't been any change to the script. Will need to spend a bit more time investigating after work
[schmarty] joined the channel
# 
[schmarty] lockfiles++
# [qubyte] While markdown is far from perfect, I think we’re being a little unfair to it here. Its original intent was for prose, not poetry. Anyway, HTML is a poor fit by default too. When it comes to poetry, nothing is off the table, right? Playing with alignment and spacing is common, and HTML doesn’t have a built in vocabulary. That stuff needs styles to be applied to achieve. markdown++
# [campegg] poemdown? marketry?
# [campegg] PFM? (poetry-flavored markdown)
# [campegg] 😆 +1 for po'down
# 
capjamesg Also see https://www.coem-lang.org/
# [qubyte] I’m reminded of LilyPond. https://lilypond.org
# [Paul_Robert_Ll] So… tabs or spaces? _ducks_
[benatwork] joined the channel
# [KevinMarks] If you use mysql from node you really want a finally
# [qubyte] At least until explicit resource management makes it into V8. https://github.com/tc39/proposal-explicit-resource-management
# [campegg] I managed to track down what was going on with my server; the cron job I use to process incoming and outgoing webmentions was hanging because of a malformed outgoing webmention. Thanks again for the pointers that got me looking in the right place, bacardi55[m]++
# bacardi55[m] Great that you found the issue!
# [campegg] This is starting to push against my long-held—and very strong, almost visceral—dislike of portmanteaux 😉
# [campegg] [qubyte] if I'm going to hate on something, I'm going to do it right, lol
# [campegg] Please, no. (Also, I think we're getting into #indieweb-chat territory here)
# btrem to2ds: I can configure VS Code to not strip spaces at EOL. But I find that config useful. So I never accidentally left two spaces at the end of a line, and never discovered the markdown "feature."
# btrem bacardi55[m]: I've never needed to put a <br> in my markup. And if I did, I have my setup configured to allow html, so if I had, I would have just inserted the <br> tag.
# btrem Back to my netlify question, if that's ok: if I drag-and-drop a folder with only images into a new netlify site, will the site show the images a la apache? As folders with images in them? If not, that would explain the 404 I get when I try to view the site.
to2ds joined the channel
# btrem That's a bit tedious. :-(
# btrem :-D
# btrem Is this your own editor? Make a "poetry" option that turns off the two spaces are a period.
Guest6233 joined the channel
ttybitnik joined the channel
to2ds joined the channel
# Loqi ok, I added "side project idea: go through "popular" "simple" npm packages and write-up one-file code solution alternatives that work by copypasta instead without all the ceremony/overhead/vulns of yet-another-npm dependency" to the "See Also" section of /User:Tantek.com/tinbox https://indieweb.org/wiki/index.php?diff=94481&oldid=93467
[Murray] joined the channel