#dev 2021-05-18

2021-05-18 UTC
sparseMatrix joined the channel
#
sparseMatrix
'sup y'all
#
sparseMatrix
@[barnabywalters] I'm noticing the title and subtitle/summary on my site seems to be fairly randomized too (/not/ the intended presentation, heh).
#
sparseMatrix
fragile doesn't begin to describe this.
#
sparseMatrix
It's due to it being only half-way integrated into the system as a service.
#
sparseMatrix
Also, taking it out of the virtualenv to get it running under proepr perms has upset my automated content update scheme, so now I have to do that manually until I have it resolved, which apparrently is a risk fraught process lol
#
sparseMatrix
I suppose I might have to slow down a little bit with this project, as tomorrow I'm picking up 20 hrs a week contract programmer/admin work
#
sparseMatrix
which I actually need pretty desperately, so hopefully that works out
alex11, [tw2113_Slack_] and [girrodocus] joined the channel
#
[girrodocus]
Good luck with the work. I saw a YouTube video about PM2 and node.js and the guy said PM2 would also work for Python. Could that be an option for you instead?
#
GWG
Working on Response Type Discovery. Trying to figure out how to design this code.
#
jacky
that's a thing?
#
jacky
how would that be different from post type discovery
#
GWG
jacky: It's in the PTD Spec.
#
GWG
It is a subset of the Post Type Algorithm, which is for discovering the type of a response.
#
jacky
ah okay
#
jacky
what is post type discovery
#
Loqi
Post Type Discovery specifies an algorithm for consuming code to determine the type of a post by its content properties and their values rather than an explicit “post type” property, thus better matched to modern post creation UIs that allow combining text, media, etc in a variety of ways without burdening users with any notion of what kind of post they are creating https://indieweb.org/post-type-discovery
#
GWG
The code currently used by me is due for replacement as part of the next version. Rethinking how to improve it.
#
GWG
Sometimes, I have to stew for a bit.
[tw2113_Slack_], [KevinMarks], [Ana_Rodrigues], barnabywalters, [Murray], [chee], barnabywalters_, ShadowKyogre, reed, ytqb[m], smacko[m] and [schmarty] joined the channel; ShadowKyogre left the channel
#
@astralwave
↩️ I guess URLs and webmentions works as a Web 2.0 backport, kind of. But URLs are like page refs (hyperlinks), not block refs (transclusion). Twitter is an exception because tweets are atomic by design.
(twitter.com/_/status/1394647628311322632)
astralbijection[, zootella, unrelentingtech, jimpick and AkyRhO joined the channel; ShadowKyogre left the channel
#
@kevinmarks
↩️ The combination of webmentions with microformats gives a way to provide structured responses https://indieweb.org/responses You can use a fragmention to refer to a piece of text too
(twitter.com/_/status/1394652111607062533)
sumner6, oenone_, strugee, stacktrust_, [snarfed] and ShadowKyogre joined the channel
#
Loqi
[Daniel Aleksandersen] How to stop Bing ranking XML feeds over your webpages
ShadowKyogre and [manton] joined the channel; ShadowKyogre left the channel
#
[manton]
[dianoetic_net] Your Hugo theme looks great, congrats! I was thinking about making it work as one of the default themes for Micro.blog customers to choose from. Let me know if you have any concerns about that, or if I should wait to play around with it.
mtinsal, ShadowKyogre and [chee] joined the channel; ShadowKyogre left the channel
#
barnabywalters
jacky: regarding issues #1 and #2 in webmention-ecosystem: is there some user-focused discussion of what problems they’re solving somewhere? they seem very vague and plumbing-focused, and seem to be jumping to the conclusion that JSON is necessary, rather than considering if human-readable HTML+mf2 would work
ShadowKyogre and sparseMatrix joined the channel; ShadowKyogre left the channel
#
sparseMatrix
lol digital ocean and censys.io are scanning my application layer with penetration testing tools
#
[manton]
My 2 cents is the use-case for retrieval is to use JavaScript to show mentions on a blog post, so JSON is convenient for that.
#
GWG
barnabywalters: Did you watch the video?
#
[manton]
Especially if the goal is to have potentially multiple “rendering” libraries that could output mentions in different ways.
#
barnabywalters
there’s a video?
ShadowKyogre left the channel
#
GWG
From the event this weekend
#
GWG
Video and notes
#
barnabywalters
oh a video of the meeting, okay. I was watching the etherpad
gRegorLove joined the channel
#
sknebel
barnabywalters++ good point though re explicitly capturing usecases
#
Loqi
barnabywalters has 2 karma in this channel over the last year (5 in all channels)
#
aaronpk
i think we kind of skipped over explicitly documenting the use cases since we were starting with the assumption of people's existing uses of webmention.io
ShadowKyogre, lanodan and ben_thatmustbeme joined the channel; ShadowKyogre left the channel
#
GWG
But it's the same argument for why Microsub doesn't use html+mf2, the fact that projects like this need a limited subset of Microformats vocabulary...of course you could do that in HTML as well.
#
jacky
barnabywalters: I added a comment in one of them regarding content negotation
#
jacky
so like one could ask for text/html+mf2 (or just text/html) and get that
#
jacky
that said, I can add a note about that to the ticket to help scope it
#
aaronpk
i think the real answer is the consumers of this data isn't expected to be humans, it's expected to be software
#
barnabywalters
neither of the issues makes it clear that they apply specifically to people implementing webmention.io-like servers, and client software to consume data from those servers
#
jacky
hm okay
#
barnabywalters
read from the POV of someone implementing their own webmention endpoint, they’re very confusing
#
jacky
I'll up that
#
jacky
*update that
#
jacky
I definitely smacked it together after the popup
#
barnabywalters
although I think that even for that use case, there are benefits to providing formatted replies in a human-readable format
#
aaronpk
there are benefits, but i don't think that part needs to be standardized
#
barnabywalters
e.g. people could link to that page if they wanted to make the replies available but not show them directly on their website
#
Loqi
[jalcine] #1 Defining the Retrieval Format for Webmentions
#
aaronpk
like it's fine for people to do that on their own in any way they want, without a standard
#
aaronpk
also remember that we are trying to adapt to what people are actually doing in the wild today, not plan for what people might theoretically want to do
#
Loqi
[jalcine] The reasons above (needing to maintain compatibility and allowing for an option in the kind of format that they'd like to use) does nudge me towards using some sort of content negotiation here. However, for the simplicity of this, it seems like defin...
#
aaronpk
it does, but people don't use that as far as i can tell
#
jacky
stats++
#
Loqi
stats has 1 karma over the last year
#
jacky
one could contest that the lack of its use is a discoverability thing but /shrug
ShadowKyogre joined the channel
#
aaronpk
the conversation was really centered around "people are doing X with webmention.io today, how do we enable them to do that same thing but on more things than webmention.io" -- i think that context maybe didn't get captured in the notes well enough because it was right at the beginning
#
barnabywalters
it certainly didn’t make its way into the GH issues
#
aaronpk
agreed, those need some more context for sure
#
barnabywalters
when I first read #2, I thought “retrieving webmentions” referred to looking up the status of individual webmention requests
#
barnabywalters
i.e. if they were queued and the sender wanted to get a response for them for some reason
#
jacky
is very open to changing the copy of those issues
#
GWG
barnabywalters: That's a topic I wanted to discuss at some point... some standardization for status pages, but we kept our focus on this particular aspect.
ShadowKyogre left the channel
#
barnabywalters
jacky, aaronpk, GWG, [manton] I added a bunch of context to https://github.com/indieweb/webmention-ecosystem/issues/1, let me know what you think
#
Loqi
[jalcine] #1 Defining the Retrieval Format for Webmentions
sparseMatrix joined the channel
#
sparseMatrix
I'd guess from reading over that discussion GH that you guys would like think a capabilities endpoint is probably overkill.
#
Loqi
[jalcine] #2 Defining the Retrieval Source for Webmentions
#
jacky
hmm I don't see anything there
#
barnabywalters
it let me edit the original issue for some reason
#
jacky
oooooooh
#
jacky
perfect
#
jacky
you're part of the org
#
jacky
Yeah, I think I like this rewrite
#
jacky
if one didn't have a context of what's going on, I think it sets up the scene well
#
jacky
only thing is that the response types are something that's not explicitly defined by Webmention (so that parsing type _could_ be an implementation detail tbh)
#
barnabywalters
where did I imply that parsing response types were defined by the spec? definitely not here “Other than verification, update and delete behaviour, it leaves the details of retrieval, storage and presentation up to the implementor's discretion.”
ShadowKyogre joined the channel
#
jacky
I think it's done by the explicit definition of those post types " that another post has responded to it somehow, where the response could be a plain mention, a reply, a repost, a like, etc"
#
jacky
like I _think_ someone could interpret the Webmention spec to _also_ do this but I might be reading too deeply into that
#
barnabywalters
yeah, that comes right after “The most common use-case is…” though
#
jacky
ah of course
#
jacky
then it's all good
#
barnabywalters
I wanted to make it clear that while the webmention spec is minimalistic and could be used for all sorts of things, handling responses to a h-entry is the most common use case, and the use case on which the discussion is focused
#
jacky
I think I'm getting GitHub fatigue lol
#
jacky
yeah that's a good way to capture that
[KevinMarks] joined the channel; ShadowKyogre left the channel
#
barnabywalters
I couldn’t resist linking to https://snarfed.org/1-million-webmentions as a citation for “most common use case” :D
#
barnabywalters
I guess we’re overdue another one of those posts, now that bridgy alone has sent >2M webmentions
#
jacky
maybe for 10M
#
barnabywalters
we definitely need a new one in time for webmention’s 10th birthday next year
#
jacky
ha yeah
#
GWG
Maybe cupcakes?
#
jacky
lol with the WM mini logo? that'd be cool
#
doosboox
searchmysite.net and wiby.me seem neat. Anyone here knows anything about them?
#
Seirdy
doosboox: searchmysite.net supports indieauth if you pay.
#
doosboox
Seirdy: cool
#
doosboox
https://github.com/searchmysite/searchmysite.net/blob/main/src/indexer/indexer/spiders/search_my_site_parser.py <-- I'm trying to understand the indexer. I've never seen BeautifulSoup in use before.
#
barnabywalters
doosboox: which part are you confused about?
[snarfed] joined the channel
#
[snarfed]
definitely fun! disappointing to see that it still limits to 500 indexed pages per domain though. keeps it from being very useful as a search engine 😆
#
jacky
haha agreed
#
jacky
wonder if it's like an affordability thing
#
jacky
oh or maybe like a 'freshness' thing?
#
doosboox
barnabywalters: not confused, just haven't understood it all yet :D
#
[snarfed]
disk is cheap, pay as you go cloud hosting is cheap, honestly i don’t get the 500 page limit. 🤷
#
doosboox
[snarfed]: crawling and indexing isn't cheap though. Eats bandwidth and CPU
#
[snarfed]
eh, i’d argue it’s still cheap, esp since crawling is I/O bound (not CPU), and it’s only looking at HTML, not assets
#
[snarfed]
esp considering current cloud compute prices
#
[snarfed]
everyone’s in a different situation money wise, granted, but I kinda doubt they did the math and realistically determined they couldn’t afford to crawl more pages per domain
#
jacky
now I kinda wish they _did_ explain that on the site, lol
[chrisaldrich] joined the channel
#
jacky
that said, I'm really glad that they leverage IndieAuth for this
#
jacky
it's like one of those IndieWeb-adjacent-without-mentioning-the-IndieWeb things that just pop up heh
ShadowKyogre joined the channel
#
jacky
hm that's not _as_ informative as I'd like it to be
#
jacky
but tbh that's okay
#
jacky
still a bit cheaper than I expected
#
[snarfed]
oh nice, good find! that’s much more public info than i expected
[tw2113_Slack_], ShadowKyogre, barnabywalters_ and [jeremycherfas] joined the channel; ShadowKyogre left the channel
#
[jeremycherfas]
I’d be willing to pay $7 a year to enable 9 other sites to be included.
#
doosboox
alright, played around with BeautifulSoup4 and watched Spider-Man Far From Home at the same time. Both are simple and fun to deal with.
#
doosboox
I can definitely see myself using BS4 for a project
#
[chrisaldrich]
I'm pretty sure the creator of searchmysite.net has been in the IndieWeb chat before...
#
[chrisaldrich]
what is searchmysite?
#
Loqi
It looks like we don't have a page for "searchmysite" yet. Would you like to create it? (Or just say "searchmysite is ____", a sentence describing the term)
#
[chrisaldrich]
what is search my site?
#
Loqi
Search My Site provides simple search for personal and independent websites - to improve the ability to find independent and personal websites, and provide a search as a service for site owners https://indieweb.org/Search_My_Site
#
[chrisaldrich]
Michael Lewis was the name... yes.
#
Seirdy
[snarfed]: it might also be to keep big sites with lots of pages from dominating search results
#
Seirdy
doosboox, jacky, [snarfed]: many of the decisions of Wiby and SMS (like SMS' 500 page limit) revolve around them not being "normal" search engines but being more like "discovery" engines. users don't use them the way they use Google, users want to surf and find a new website. So it makes sense to cap a website's representation so that it doesn't overshadow smaller sites.
#
Seirdy
atl that's my interpretation
[dianoetic_net] joined the channel
#
[dianoetic_net]
[manton] Thank you, that’s a nice compliment! Please feel free to have at ‘er. If you come up with any improvements I’d be delighted if you PR them to the theme!
#
doosboox
Seirdy: makes sense
ShadowKyogre joined the channel
#
[chrisaldrich]
Most personal sites may not have more than 500 pages until you get into crazies like me with over 30K+ posts... 😉
#
[snarfed]
hrm doesn’t really make sense to me. discovery and representing smaller sites sounds like great goals to me, but orthogonal to limiting indexing per site. if you only index 500 pages from a big site, you have no idea if you got a representative sample, and you often won’t have relevant pages for a given query at all
#
[snarfed]
(you’d need to put work into your ranking algorithm, of course, but if you really want a decent search/discovery engine, that work is worthwhile)
#
[chrisaldrich]
500 posts on my site only gets you back to January (for public posts at the moment), and only May 2nd if it were to include all my private posts....
#
[chrisaldrich]
It does look like the 500 limit would just about cover all my articles though...
ShadowKyogre, maxwelljoslyn, Kendall[m] and [Murray] joined the channel; ShadowKyogre left the channel
#
[Murray]
yeah, Michael Lewis was pretty active last year, talked a lot about the building of searchmysite
#
[Murray]
also, ooph, I'm always interested to see where people draw the line at "not that expensive"; my personal hosting costs just over £50 a year and that seems fair. I'm currenty looking at upgrading, which would go to around £100-120 a year and that feels too steep, but at least I'd get some genuinely useful functionality out of it. If someone told me a side project was getting to 5x that, I'd shut it down immediately 😅 (tbc, very grateful for
#
[Murray]
those who run such projects, I use a bunch that I'm sure cost more than that, just find these types of things interesting; I consider myself extremely well off now, but that kind of spending on something makes me anxious still)
#
@CrowderSoup
↩️ #IndieWeb is cool. You'd need to add a micropub endpoint to your blog that could handle standard requests + get IndieAuth set up which is fairly trivial unless you want to build your own. I'm slowly trying to add all these to my own blog. https://indieweb.org/
(twitter.com/_/status/1394784368694992896)
#
@jamesvandyne
@iamhirusi I think I've broken sending webmentions on my blog, but saw your post about handling django & .env files. I highly recommend using django-environ to read/convert your .env files. Full response is on my blog at https://jamesvandyne.com/03a30d90-1d13-41c7-9862-0f2b4b5ef881
(twitter.com/_/status/1394784319525113858)
#
[chrisaldrich]
FullResponseIsOnMyBlog++
#
Loqi
FullResponseIsOnMyBlog has 1 karma over the last year
#
jacky
that's a domain name
#
[chrisaldrich]
register it quick... that sounds like a good micropub client domain name
#
jacky
tbh!! lol
#
jacky
but I can't, I've been told by the head of household that I have to stop with that lol
#
jacky
after she saw how much I pay for black.af, I have to let one go before getting another
sparseMatrix joined the channel
#
sparseMatrix
casual observation: it seems that in this age of everything on the cloud, self-hosting begins to take on a pretty narrow set of activities
#
sparseMatrix
*begins to describe a pretty narrow set of activities
#
sparseMatrix
mostly centered on systems administration
#
sparseMatrix
unless one insists on local hardware as part of the definition
#
sparseMatrix
many of you guys seem to wrestle a bit internally with the philosophy of independent ownership/operation vs. cloud hosting and centralization
#
sparseMatrix
e.g., single points of failure. but really, strrrictly locally hosted stuff is a pretty big SPOF too
#
jacky
that depends on how you view the 'point'
#
jacky
like if you look at something like GCP or AWS, then if they go down, thousands of people go down
#
jacky
ignores how this is guaranteed to happen since multi-cloud deployments and systems are a bit taboo by the industry by design
#
jacky
but like if it's just _my site_, then I'm the only one failing to be - only my site is affected
#
jacky
other peoples' site can still be up
#
sparseMatrix
was just reading up on that jf2 working notes document
#
sparseMatrix
really good stuff there
ShadowKyogre joined the channel
#
sparseMatrix
I keep coming back to this question: is my site a blog, or is it a social media client? to me, this cooks down to, "is it a client or is it a server?" of course, there is nothing to be gained by forcing roles like that into existence, but it seems that as the interaction between author and reader continues, the role gets quite blurred. Maybe there is something to be gained by designing an app with elements of both?
#
sparseMatrix
I keep envisioning a tiny webserver running on a phone serving an hcard
#
sparseMatrix
just like, you know, the most basic level of such an operation
ShadowKyogre left the channel
#
sparseMatrix
there is so much that would get in the way of doing that, like right now, but things change
#
sparseMatrix
ways to do stuff are found
#
sparseMatrix
in 1998 I was taking free java classes on sun microssystem's website in between applying for jobs on monster.com
#
sparseMatrix
I had a bunch of contact info in a palm pilot that I was having to transcribe each time I applied
#
sparseMatrix
it was frustrating, so I wrote Bill Joy an email -- I told him he needed to make a cell phone running nothing but a java vm
#
sparseMatrix
it would have stripped down linux kernel in it
#
sparseMatrix
sound familiar? lol
#
sparseMatrix
it was pretty damned unlikely at the time
#
sparseMatrix
now that shit is everywhere
#
sparseMatrix
I'm not telling you this bc I'm trying to claim the smart phone as my own idea
#
sparseMatrix
I'm telling you this because I want to make the point that some things are inevitable
#
sparseMatrix
I figure probably me and about 10000 other guys had tha idea within a few weeks of each other
#
sparseMatrix
some of them were in a position to do something about it
#
sparseMatrix
so anyay, cool story right
#
Loqi
😃
[snarfed] joined the channel
#
[snarfed]
hah, funny story indeed!
#
[snarfed]
what constitutes self hosting, whether to, how much to, etc are all common and useful conversations here…
#
[snarfed]
...but fwiw, indieweb doesn’t necessarily require or advocate self hosting, whether all or in part
#
[snarfed]
we advocate that you a) use your own domain, and b) own your data, which generally means that you can export it from your server software (etc). those two are generally enough to ensure that you own your identity and destiny online
#
sparseMatrix
It's something that keeps coming back around for me - as I've made a pretty good income hosting some niche applications
#
sparseMatrix
There are a lot of seeming opportunities here to host data and/or provide relevant services
#
sparseMatrix
but I keep asking myself whether that is in philosophical conflict with indieweb
#
jacky
if it was, then micro.blog would be a bad thing ;)
#
aaronpk
what is self-hosting?
#
Loqi
It looks like we don't have a page for "self-hosting" yet. Would you like to create it? (Or just say "self-hosting is ____", a sentence describing the term)
#
sparseMatrix
: D I'm so new to the eco Jacky, I can only imagine what that is
#
aaronpk
what is self hosting?
#
Loqi
It looks like we don't have a page for "self hosting" yet. Would you like to create it? (Or just say "self hosting is ____", a sentence describing the term)
#
aaronpk
what is hosting?
#
Loqi
Web hosting can be the primary regular cost in maintaining an IndieWeb site; this page lists several options from free on up depending on your publishing needs, like a static, shared, private, or dedicated server https://indieweb.org/hosting
#
jacky
what is pine.blog
#
Loqi
Pine.blog is a reader and blog hosting service that supports Webmention https://indieweb.org/Pine.blog
#
jacky
that one too
#
aaronpk
i feel like we need what [snarfed] just said on a wiki page. he always says it so well and this question keeps being asked
#
[snarfed]
aw thanks aaronpk!
#
sknebel
right, we generally are explicitly not against services, we just have opinions about what services should do
#
sparseMatrix
@[sknebel] oh, completely understandable, and logical that those making the most use of the thing should go about refining it
#
sparseMatrix
and if I may say, doing a proper job of it
#
sparseMatrix
myself, I'm still sniffing around getting the lay of things
#
aaronpk
self hosting is the practice of running the software for your personal website on hardware under your own physical control, typically on a server at home
#
sparseMatrix
@aaronpk: that's me right now
JackyAlcin[m] joined the channel
#
sparseMatrix
with the exception of stupid router tricks, it isn't a lot different technically from doing it at e.g., digital ocean, at least day-to-day
edsu, astralbijection[ and Kendall[m] joined the channel
#
sparseMatrix
micro.blog looks pretty awesome
#
[snarfed]
it is! it’s our recommended indieweb site host for people who aren’t technical and don’t want to write code or do sysadmin
#
sknebel
aaronpk: well.... that definition is not that clearcut IMHO :/
#
sparseMatrix
pine.blog looks pretty awesome too. I see they also incorporate reddit. I've been wondering how I'd go about that, given reddit's explicit policy of anonymity.
#
sparseMatrix
I'm a pretty prolific commentor on reddit, but I'd really not want anyone knowing who I am there.
#
sparseMatrix
You could say I'm pretty vocal about politics ;)
#
[snarfed]
you may be interested in Bridgy’s Reddit support: https://brid.gy/about#which
#
sparseMatrix
the last thing I need is a bunch of crazy ass republicans showing up at my front door lol
#
sparseMatrix
hahaha yeah, no that wouldn't work for me, linking to my site would be doxxing myself
#
Loqi
hahaha
fredcy_ joined the channel
#
sparseMatrix
stuff I write on my own blog will keep me in enough trouble no doubt
#
sparseMatrix
I actually had to redact a couple of posts today, things that were written when the site was meant to be private and never should have been crawled
#
sparseMatrix
so if you're reading any of my journals and hit a page that sayss 'redacted' you'll know whats up
drmr and astrojl_matrix joined the channel