#microformats 2018-07-09

2018-07-09 UTC
#
aaronpk
oh gosh, apparently the server is trying to email rohit a *lot*
#
aaronpk
"server reached MaxClients setting, consider raising the MaxClients setting"
#
aaronpk
I had a feeling it was something like that
#
tantek__
I wonder why its trying to mail rohit though
#
tantek__
also lame that a single app (mailing?) can slow down everything else on the server
#
tantek__
wonders if there are os-level ways to throttle (CPU, memory) any particular process so it doesn't eat up everything
#
aaronpk
I don't think the email was slowing it down
#
aaronpk
apache is getting too many requests at some point, and then it gets stuck
#
tantek__
apache was trying to email rohit?
#
aaronpk
not sure what was trying to email yet
[kevinmarks] and [jgarber] joined the channel
#
Loqi
[kartikprabhu] #108 deepcopy error on <use> inside <svg>
#
sknebel
commented
KartikPrabhu, [miklb], tantek__ and [manton] joined the channel
#
@davidcoles
Updating the website. First step: Restoring the old website from http://archive.org. Enjoy some 2010 era website featuring handcrafted valid XHTML 1.0 Strict, microformats, RSS/Atom, PGP, OpenID delegation and my uncut Edinburgh hair. http://dcoles.net/
(twtr.io/1fq9hRFfVWf)
#
@davidcoles
Updating the website. First step: Restoring the old website from http://archive.org. Enjoy some 2010 era website featuring handcrafted valid XHTML 1.0 Strict, microformats, RSS/Atom, PGP, OpenID delegation and my wild Edinburgh hair: http://dcoles.net/
(twtr.io/1fq9thVbhNw)
[cleverdevil], barpthewire, wakest, gRegorLove, [kevinmarks], tonzyl, [tantek], [colinwalker], [manton], [kim_landwehr], [eddie], bradenslen, sketchess, Lovi, nitot_, tantek__, KartikPrabhu, [jgmac1106], tantek and [jgarber] joined the channel
#
gRegorLove
tantek, do we need +1s for adding proposed h-entry properties, or is it ok as long as we include the publishing/consuming examples? e.g. https://github.com/microformats/h-entry/issues/10#issuecomment-401185757
#
Loqi
[gRegorLove] We have 2+ publishers for `p-read-of` and `p-read-status`: * https://gregorlove.com/2018/06/want-to-read-welcoming-the/ * https://martymcgui.re/2018/06/12/105334/ * all indiebookclub profiles, e.g. https://indiebookclub.biz/users/gregorlove.com ...
#
gRegorLove
Issue 11 and 12 are also for proposed/draft properties
#
tantek
gRegorLove: more importantly - I think we need more consideration beyond just one use-case at a time
[Lovi] joined the channel
#
tantek
also multiple properties for a single use-case seems inefficient (AKA leads to vocab bloat) so I'm hoping we can either figure out how to solve more use-cases with same/similar vocab, or use fewer properties, or both
#
tantek
in particular, is this really so narrow about just books? or any printed matter? or any text (including online)? or is this overlapping with /watch posts? or /listen posts?
#
tantek
all the other proposed property additions went through months of experimentation / experience, and use in different contexts over time before getting slowly solidified
#
gRegorLove
Ok, makes sense
#
tantek
just look at all the u-photo vs e-content discussions over the years we had to have just to get that one "simple" property working properly
#
gRegorLove
u-listen-of is #11, heh
#
tantek
so I'd advise not "rushing" all the -read- props, because they seem to be focused on a very specific use-case, which may end up needing them, or may not need them because we can cleverly re-use other properties, or rename/expand them to handle more media types
#
tantek
do we really need separate ones for each media type like that?
#
aaronpk
i have a similar situation with my food/drink posts
#
tantek
or worse, separate *pairs* of them?
#
aaronpk
and exercise/workouts/transportation
#
gRegorLove
Didn't intend to rush it through the process, but thought it would be good to show it's proposed since it is in use. I can see how that might encourage people to use it outside the use-case discussion, though, making things tricky.
#
gRegorLove
consume-of? heh
#
tantek
there's also the architecture astronomy extreme, which ends up with properties like "about-of" for all the *-of uses, and some abstract property like "action" or "activity" which just pushes the vocab problem down a level into a single property
#
tantek
that's potentially worse because it makes it *look* simpler while actually making it much *more* work to implement because now you have to deep dive one more layer to do the "actual" work
#
tantek
ideally we'll find a balance somewhere in between the short-term one property per use-case, and the useless layer of abstraction one property for everything which just pushes the problem down a level
KartikPrabhu, [jgmac1106] and [tantek] joined the channel
#
gRegorLove
Hm, interesting
#
gRegorLove
Did I make indiebookclub too soon? I'm wondering if I should be worried about bigger consumers like micro.blog adding support for read-of now, if it might change in the future.
#
gRegorLove
I guess worst case, if a new, common property is decided on, ibc could still send `read-of` for legacy consumers.
vivus and tantek joined the channel
#
KartikPrabhu
gRegorLove: *-x-* is a thing for experimental properties for a reason ;)
#
gRegorLove
I thought there was some recent instance where that was recommended against?
#
tantek
context. it's something IETF has had problems with
#
tantek
it's a problem if there's no obvious way to standardize vs reject things, and -x- things linger for *years*
#
tantek
then the -x- thing becomes part of the practical standard. see email headers
#
gRegorLove
Ahh, right
#
gRegorLove
Well, I'm not really sure what to do at this point. listen-of / watch-of / read-of are being published and consumed without the -x- so I'm not sure adding -x- to read-of at this point helps?
#
tantek
how are you determining "are being published and consumed"? are there specific things listed on each of their pages?
#
gRegorLove
Yes, listed on the github issues.
#
Loqi
[gRegorLove] We have 2+ publishers for `u-listen-of` (more listed on https://indieweb.org/scrobble#IndieWeb_Examples) and at least 1 consumer, so this can be added as a proposed property. I propose adding: > **u-listen-of** - indicates this h-entry is a scr...
#
tantek
the issues aren't really a good place to keep track of "current state" - each of their pages e.g. indieweb.org/read /watch /listen should be listing the current set of indieweb examples and consuming software
#
Loqi
[gRegorLove] #12 Upgrade u-bookmark-of from proposed to draft property
#
tantek
and the issue should link to those sections of those pages
#
tantek
rather than attempting to duplicate the information
#
tantek
until we've actually crossed a meaningful threshold
#
gRegorLove
...they do
#
tantek
no. indieweb.org/read has zero consuming things listed. "consum" is not in the text.
#
tantek
publishing != consuming
#
tantek
"posting read posts to your site" = publishing
#
gRegorLove
I'm consuming micropub posts from ibc, as is schmarty. I'll update my example on there.
#
gRegorLove
As is micro.blog as of IWS
#
tantek
when you say "I'm consuming micropub posts" I don't think you understand what "consuming" means in this context
#
tantek
something being a micropub server != consuming
#
tantek
consuming is something actually parsing the post from a permalink or feed and doing something interesting / different with it
#
tantek
the whole setup, ibc > mp > your site is all "just" publishing
#
tantek
none of that is "consuming"
#
gRegorLove
I thought it meant consuming the `read-of` property whether it was parsing from HTML, or a property in an mp request.
#
tantek
the former, not the latter. the latter has never meant consuming in any context
#
tantek
handling an API request != consuming
#
gRegorLove
ibc does parse a URL for read-of properties now, so that's consuming. Updating now.
#
gRegorLove
It's on the documentation page, just not the indieweb wiki yet: https://indiebookclub.biz/documentation#query
#
tantek
when you write an email in Gmail, is Gmail "consuming" your email? No. Gmail is helping you publish that email to whoever you send it to. When those other email addresses receive it and parse the email, then they are consuming it and presenting it for a user to consume
#
tantek
is that for reposting read posts from other people?
#
gRegorLove
Not "repost" strictly speaking
#
gRegorLove
Create your own read post for that book
#
tantek
oh cool, like a I'm also reading this kinda post
#
tantek
gRegorLove: in particular I'm concerned about "read-status" which seems very special purpose
#
tantek
whereas nearly the same values would make sense for a "watch-status" property
#
tantek
which usually means there shouldn't be two properties, instead, there should be one with more generic values
#
tantek
instead of "to-watch", "watching", "finished"
#
tantek
or "to-listen", "listening", "finished"
#
tantek
and this "status" could be broadened beyond consuming media to apply to something like "tasks" / "to-do" items too
#
tantek
"to-do", "doing", "finished"
#
tantek
in fact a reading a book could be a task / to-do for someone
#
tantek
as could seeing a movie, listening to a music album etc.
#
gRegorLove
Yeah, the primary reason for read-status at the time was so the mp client wouldn't have to either a) deal with i18n for the text "want to read," "reading," etc. or b) query the mp endpoint for options. Agreed there could be a better property name that isn't so specific.
#
gRegorLove
b) is what I was initially thinking it would need to do, when I started working on this at IWS 2017.
#
tantek
agred with not trying to standardize plain text of "want to read," "reading,"
#
gRegorLove
but then IWC Austin, Aaron helped me realize I needed to take a step back: https://gregorlove.com/2017/12/my-project-for-todays-indiewebcamp/
#
Loqi
[gRegor Morrill] My project for today's IndieWebCamp Austin was initially to update my fork of Quill for posting read posts. I started by getting my fork of Quill updated. As I was working through some questions about how to include the action being taken on the boo...
#
tantek
what I'm saying is that when you start doing any kind of format design (which property proposing is doing), you need to start thinking about the ecosystem the format is in in general so you don't just grossly expand it with lots of minor variants which makes it harder for everyone publishing and consuming everything in the long run
#
gRegorLove
Yeah, understandable
#
tantek
this is in particular a key part of *micro*formats
#
tantek
in contrast to XML/RDF/LD *mega*formats which really have a "don't care" philosophy about a huge complex sprawling mess (see schema org for one such mess by just a single (effectively controlling) company!)
#
tantek
if anything we've tried to *reduce* number of properties over time across formats
#
tantek
witness the h-event vs h-entry discussions and collapsing
#
tantek
attempt to eliiminate p-description etc.