#dev 2024-04-04
2024-04-04 UTC
# 
[aciccarello] I just sent myself a private message from a mastodon account. We'll see what bridgy does
# [aciccarello] I don't see anything in my bridgy fed notifications so I assume they are silently ignored
# [aciccarello] "Same with DMs and private/followers-only posts; it ignores those."
AramZS, [tw2113] and [snarfed] joined the channel
[jeremycherfas] joined the channel
# 
[jeremycherfas] Where are you hosting now GWG ?
# 
[Jan_Tuomi] I've been thinking about how to format my /now page. Currently it's just a static pile of text, with a "currently" and "recently" section. I can see that the "recently" section grows larger over time and at some point I would have to drop the oldest stuff. I would like to keep an archive of old nows though.
# [Jan_Tuomi] This site has a nice solution to the problem: https://derekkedziora.com/blog/dynamic-now-page. TL;DR special notes with a "now" category tag would cause a publishing routine to trigger that replaces the now page with the new one. Old now pages live on and can be linked to from the newer now page.
# [Jan_Tuomi] However, this approach requires that, during rendering of a page, there is a database of other posts that can be queried to construct a link to the previous now page dynamically. Trivial to do with a proper SSG with a proper database of course. However my current pandoc-based publishing setup renders one page at a time without any knowledge of other posts. All links are static.
# [Jan_Tuomi] is very limited.
# [Jan_Tuomi] I was thinking that maybe I could improve my setup so that it does compilation in multiple passes. First pass would collect paths and markdown frontmatter from each page into an in-memory structure, and the second pass would then render each page like it does now, although now with an in-mem database that can be queried. Does this sound sane? I would also probably have to implement a better templating engine since the builtin pandoc one
bterry and [Jan_Tuomi] joined the channel
[contact898] and to2ds joined the channel
[qubyte] and to2ds joined the channel
bterry1 and [Murray] joined the channel
Guest6 joined the channel
# jeremycherfas !tell GWG I'm hoping I can avoid having to rebuild from scratch. I use Dreamhost and they are pretty good about that sort of thing.
[Scout] joined the channel
# 
GWG jeremycherfas: My VPS, codenamed Wisdom1, has been online for as long as I've been part of the community. Also running CentOS, which switched to a rolling release. I need a long term stable release distribution. So...if you can access my website now, the migration is already over. I have 15 more websites to move this month before I shut down Wisdom1 for good
# jeremycherfas I see the site!
[lcs] joined the channel
# 
[KevinMarks] An AP server written in C that runs in the file system (no db) https://codeberg.org/grunfink/snac2
[asuh] and to2ds joined the channel
ttybitnik, Guest6, [Paul_Robert_Ll], Guest6_ and to2ds joined the channel
jonnybarnes, [KevinMarks], geoffo and jjuran joined the channel
# [aciccarello] Speaking of self closing tags...
ttybitnik joined the channel
# 
capjamesg "We do not know of any solution that could guarantee (e.g., through cryptography) that a given piece of content was or was not generated (partially or entirely) by AI systems. That gap unfortunately leaves a systemic risk in terms of misinformation and spam that should be of grave concern for the health of the Web as a content distribution platform and of society as a whole." 😦
# capjamesg I like the work the BBC/Adobe, et. al. are doing with Content Credentials: https://www.bbc.com/mediacentre/2024/content-credentials-bbc-verify
# [KevinMarks] R&D wrote more about C2PA https://www.bbc.co.uk/rd/blog/2024-03-c2pa-verification-news-journalism-credentials
[aciccarello] joined the channel
# [aciccarello] I agree, checking if something is written by AI feels like it's missing the point. Whether it's correct seems more important (but also hard).