#dev 2026-02-03
2026-02-03 UTC
#
[Al_Abut] Now that I think about it, I was more keying off of the idea of explaining decentralization. That was the cardinal sin of Mastodon when they had all the initial Twitter exodus attention. The idea that they needed to “educate users” by forcing them into picking a server (without a default fallback option that didn't come until later) is a breathtakingly bad piece of UX. One that they recognized and publicly admitted but it was too little
#
[Al_Abut] too late.
#
[Al_Abut] If I remember the timeline correctly, Bluesky was still a private beta and so they got to quietly watch Mastodon drop the ball hard so hard that their niche trajectory was practically set in stone. That's why I'm not surprised that they're orders of magnitude larger than Mastodon despite coming on the scene later.
#
[Al_Abut] There might be a parallel to be made there with telling people interested in having their own site to pick a domain name or hosting provider.
#
osteophage ulhar4409[d]: Re: Bluesky, there are some links on the wiki page expanding on that (in agreement with you).
#
osteophage [Al_Abut]: Re: growing fast despite launching later, Bluesky had a lot of other PR advantages, but yeah I do think having only one website to choose from was (in a way) one of them.
#
osteophage Not sure how coherent those last two posts were for folks across the bridge, but the first was a reply to Ulhar and the second was a reply to Al_Abut.
ggarcia246, doesnm_, oodani, gRegorLove_, gRegorLove__, gRegor, vidak, grufwub, rgbivvv, DanQ and barnaby joined the channel
#
[jeremycherfas] Ha. And a propos of that, I really do like WithKnown as one of the most competent IndieWeb-ready CMS around, but lawks! is it a mess to tinker with. Having successfully migrated to a new host and persuaded it to use the CSS stylesheet it always used in the past, it still cannot find photos. Maybe later.
#
[jeremycherfas] I wonder, again, how portable “my” data really are, with no export available. Maybe from the RSS? Or scrape the HTML? Or maybe Jottit will have an importer.
doesnm_ joined the channel
#
[simonbc] No importer yet, but I've been thinking about it. Jottit stores everything as markdown in git, so importing is really just converting to markdown files. If your site has an RSS feed, that's probably the easiest path.
#
[simonbc] I'd be happy to script a conversion. Of course totally up to you if you want to import it into Jottit or some other platform 🙂
ax562 and [lazcorp] joined the channel
#
[jeremycherfas] Right ow, I would leap at the chance, although the honest truth is that I have not yet looked into Jottit in any depth. Off to do that.
#
[jeremycherfas] One of the great things about WithKnown is that it is easy to have a distinct feed for each type of post, which could make importing easier.
#
[jeremycherfas] Linking to my existing custom domain is obviously great. And from what I see, I could host the built site in my S3-compatible storage. Is that right?
#
[Al_Abut] [jeremycherfas] I don't know much about withknown (pun intended) - what do you like about it?
oodani joined the channel
#
[simonbc] Distinct feeds per post type would make importing cleaner: we could map each to a separate collection (a collection in Jottit is just a folder of pages sorted by date, typically used for a blog).
#
[simonbc] Custom domains: yes, supported.
#
[simonbc] Fair warning: Jottit is still experimental, but stabilizing quickly.
#
[simonbc] S3: Jottit hosts your site, but you can export anytime, both the source markdown and the built HTML. So you could self-host on S3 if you wanted. You always own your files.
blacklight and jjuran joined the channel
#
[jeremycherfas] Thank you [simonbc]++
#
[jeremycherfas] I’m doing the best I can to diagnose, and will soon publish a plaintiff cry for help. I may well give up and take the plunge with something experimental. The issue now is that I have no idea where all my uploaded images are, and I suspect even the RSS feeds would not find them (though I have not yet tried that).
#
[jeremycherfas] [Al_Abut] It is a lovely CMS made by [benatwork] of this parish that is fully equipped for IndieWeb goodness and works out of the box and is no longer supported. It also has a database structure that is way beyond my pay grade.
vidak, blacklight and rgbivvv joined the channel
#
khurtwilliams Fixed IndieNews submissions issue. Issue was content negotiation between ActivityPub plugin and XRay parser. Solution: mu-plugin based on David Shanske's workaround, updated for current plugin version.
#
IWDiscord <khurtwilliams>
ax562, [schmarty], saccharin and pfych joined the channel
[pfefferle] joined the channel
#
[pfefferle] we discussed that some years ago already...
#
[pfefferle] not trivial 😞
#
[morgan] Indiewebify++
barnaby joined the channel
#
[pfefferle] 🙂
gRegor joined the channel
#
[morgan] This is something I'm about to Google search more... but basically I wanted to ask if anyone uploaded a video to their own website anytime recently and had any experience making their own captions and WebVTT stuff, if anyone had any indie experience with that
#
[morgan] Context , currently there's a video of me on YouTube that I actually like a lot, and I feel weird though that I've resorted to letting YouTube auto dub it for captions and seem to rely on it when YT is not really optimal for many considerations
KE0VVT, [jeremycherfas], [artlung], [lazcorp], [schmarty], [snarfed] and vidak joined the channel
#
[schmarty] [morgan] i have a couple of posts from experiments i did for the This Week in the IndieWeb Audio Edition some years ago: https://martymcgui.re/tag/captioning/
#
[morgan] Perfect, great reference material. If I have any interesting experiences I'll try and document it for the blog about what it can be like. Seems that it's possible to export vtt stuff from YouTube so I can potentially use the text from it and make some corrections
btrem joined the channel
#
[schmarty] morgan: there are tools for converting various subtitle formats, like SRT, to WebVTT, too. here are a couple of links to subtitle editing tools i didn't get around to trying out: https://aegisub.org/ and https://sourceforge.net/projects/gnome-subtitles/
#
btrem [morgan] I did closed captioning (as well as a translation and description) for a video on my site https://btrem.com/2021/02-sledding
KE0VVT, redshuffle and blacklight joined the channel
#
[morgan] Thanks all regarding the vtt and cc question. I also had a brief demo website to share that I made today, the idea is to share some HTML5 boilerplate, it's my initial edition of it I just tested it now
#
[morgan] I'll share it later when finished with the vtt stuff
KE0VVT, ax562 and barnaby joined the channel
#
btrem With regard to using <textarea> for text that you want to offer to users to copy: there might be a bug in Firefox. When I tried using a `<textarea readonly>` for my public key -- styled with `user-select: all` -- not only does it not highlight all text on click, it seems to block any highlighting at all. On Chromium, it doesn't block highlighting, but it also doesn't highlight all on click. I'm on Ubuntu fyi.
[Mike_Mai] joined the channel
#
[Mike_Mai] why not just regular text with a "copy to clipboard" button?
#
[Mike_Mai] i always stick with "if it's not a form, don't use form elements".
#
[Mike_Mai] i disagree with that. a screen reader user going through would be pretty confused about a rogue textarea that they can't even type. everybody already knows how to highlight regular text and copy.
#
[Mike_Mai] as far as i know pre also doesn't communicate semantics to certain screen readers, so you are not really getting that "this text is special" consistently. if you were to provide that context, you should provide that context to everyone. design some surrounding context.
#
[Mike_Mai] i'll mention this but i haven't tested myself with screen readers, but `contenteditable` on a regular text element also can allow someone to click inside to select all.
#
[Mike_Mai] that is my https://mikemai.net/mcds/components/accordion/ of providing code snippet for people to copy 😄
#
[Mike_Mai] that seems like enough context to me. you don't need textarea or pre.
#
[Mike_Mai] gotcha. that makes sense
#
[artlung] This got mentioned the other day (using `user-select: all`) and it's been bugging me since the last time I tried it in Safari (a while back) it didn't work. I notice I need `-webkit-user-select: all` as well. Made a codepen with some demos. I have no particular use for this but I was, you know, curious.
Kupietz joined the channel
#
Kupietz Hey, is anyone here experienced with python's yfinance or yahoo finance api? I'm trying to find a list of available keys for the ticker.info dict but documentation appears sparse.
blacklight joined the channel
#
Kupietz NVM. Apparently different tickers return different dicts, there's no standard answer. I'm trying to reliably get the current day session close during the PM session and it's looking like it's going to involve a little more than just grabbing a key from API data. Off to fetch and parse some history...
ruta123 joined the channel
#
[tantek] btrem, your observations re: textarea are what I see too. I filed https://bugzilla.mozilla.org/show_bug.cgi?id=2013428 last week accordingly
barnaby joined the channel
#
ruta123 hello~
#
KE0VVT Hi.
#
neoruta :)
#
KE0VVT Using mdBook and Zola for generating my site. Still need to add webmention stuff. Sigh.
#
neoruta > Using mdBook and Zola for generating my site. Still need to add webmention stuff. Sigh.
#
neoruta interesting, my blog is also zola based, idk how webmention work still. is ur website open source any chance?
DanQ, DanQ1 and [aciccarello] joined the channel
