gRegorjamietanna, noticed you're using e-summary on some posts instead of p-summary which is why Monocle doesn't pick them up (see permalink, but post content is blank). Have you considered using p-summary? Alternately, could file an issue on Xray.
[tantek]xray should be picking that up, because mf2 parsers are automatically supposed to provide both the HTML and plain text versions of any e-* properties
[tantek]so if consuming code is dependent on the "plain text" version of a property, it needs to check for "only" a plain text value, and the case where there's both HTML and plain text, and grab that plain text value
[Jamie_Tanna]Thanks gRegor, tantek, aaronpk - want me to raise a feature request? I'd spotted this before with Monocle/Aperture not picking up some posts (when I first posted `e-content`), but hadn't gotten around to raising an issue with Xray
[manton]Personally I don’t think I want to “own” the venue data. I’d rather a client look up venues from a multi-user, centralized service that can be improved. I just want to own the post data.
[dshanske][manton] I have query for location name enabled and I have it access an external API. It just means that the micropub client doesn't need to handle that
[dshanske]So, you send coordinates and it offers a geo property and/or a places one...so a reverse lookup and/or a venue search. Limit and offset would apply.
[dshanske]Also, if you want to change it, only WordPress and Drupal support it, and only Indigenous for Android on the client side, so the proposal could be refined
[manton]I’m considering starting a new project that will be based on OSM data, with a places API on top. It would be kind of a standalone, open source web app but I would host a version of its database inside Micro.blog so that things work well by default with Micro.blog clients. Still in the experimenting phase, so not 100% sure yet.
[manton]As we’ve discussed, OSM data is very incomplete. I don’t think that will change without multiple clients that people are actually using that can contribute new places back.
[dshanske]My goal in having the micropub endpoint query is that you don't need to know where your site gets the data. So it works regardless of your new database.. but if it is usable by me, I'd build it into my site
sknebel(I asked above if anything supports that because I could throw up a prototype "find stuff around X,Y in OSM that might be relevant" service fairly easily, but people need to use it somehow
[manton]There are pros and cons with keeping things at the site level. On the one hand, privacy. On the other, the places data can’t improve for other people unless it’s hooked into a larger data set that improves.
sknebelhm, for evaluation one could make something that people can throw in their checkin feeds and it'll show for each location what objects it knows in that area
[schmarty]this should maybe be a blog post but it also feels like _strong opinion bait._ the "Moresquare" idea has me thinking about writing yet-another Micropub client and that has me again wishing there was some kind of boilerplate Micropub client project.
[schmarty]the Python ones have been nice and stable but have also gone stale w.r.t. updates to IndieAuth, etc. Kapowski has really suffered from dependency churn and I'm now kind of afraid to work on it.
[schmarty]i had high hopes about node because it seems like it's becoming pretty easy to host apps for free on glitch.com etc. but it feels like the bits rot too fast for my tastes.
ZegnatDo Python dependencies update less often in general, or is it because of npms tendency to just throw messages about outdated packages all the time?
[schmarty]Kapowski uses express, but it that has its own dependency trees with high churn. the other big one is node needs a good HTTP client library like axios or node-fetch or whatever, and those also churn.
[schmarty]i agree it is an interesting feeling. thinking about updating any of these projects feels like a big task but the node project is somehow extra daunting.
[schmarty]fetch in the standard library is exciting. i was just about to say i think Python just kind of has more goodies standard and with interfaces that change more slowly.
ZegnatNode added fetch in 17, they are now on version 18, and even though version 16 is still listed as LTS it is good to remember that when they stop updating 16 it is 18 that becomes LTS. So if you are writing something new today it is probably fine to go v18 only.
[tantek]in terms of a "centralized" community project for venues, you're right that OSM is kinda spotty in that regard, because venues aren't really the focus for OSM IMO
ZegnatIt does look like Flask “churns” less fast. But still, it too has had 4 updates this year. While only 5 the whole of last year. If that trend continues, I guess flask will be “churning” soon too.
[tantek]I can't remember the domain right now, but capjamesg will remember it. He setup MediaWiki specifically for collecting IndieWeb coffee / breakfast venue reviews and organizing them, and set up IndieAuth login there as well
[schmarty]Zegnat: i think i'll need to start over with Kapowski either way, so i am tempted to just not fool with node for it since i wasn't able to keep up with the churn.
ZegnatExpress seems to have only had 3 updates this year. So Flask is churning faster? XD Not a race of course, I am just trying to figure out what makes people go with different languages
[schmarty]i should also clarify that i don't necessarily think of updates as "churn". if i can bump a dependency and nothing breaks and i can go on with my day, that's great!
[schmarty]anyway all of this was to get to asking folks: what your favorite minimal PHP setups / frameworks that could get all the IndieWeb-maintained PHP libs around IndieAuth etc plugged in to make a nice boilerplate project for making new Micropub clients? 😂 😅
[schmarty]in classic "check aaronpk's work first" i see that he has sometimes used Laravel and Slim but sometimes goes right for PHP League's PSR-X libraries.
[manton][tantek] Yeah, having folks that are kind of “city experts” to help add places and curate the data would be really nice to get things off the ground. It’s kind of the Gowalla reboot model I think, too, with their idea of “street teams” that build things up during beta.
[tantek][manton]++ that is a very good framing of it, having "city experts" contribute & curate per city would be great. Similarly for "travel experts" that can offer opinions informed by experiences across different cities.
gRegorI haven't written one from scratch, but yeah, wrapping my head around the precise meaning of the parsing spec, holding that in my brain, and converting it to code can be medium involved I'd say.
Zegnatcapjamesg: re the search talk in #indieweb, I build that once upon a time. https://github.com/Zegnat/indiewebsearchengineproofofconcept It would take a search query, go and execute the search on a handful of IndieWeb WP sites, and then combine all the results.
gRegor I think it's XRay, I checked: https://www.jvt.me/mf2/2022/07/nvjcm/ and https://xray.p3k.io/parse?url=https%3A%2F%2Fwww.jvt.me%2Fmf2%2F2022%2F07%2Fnvjcm%2F&pretty=true
[Jamie_Tanna] Thanks gRegor, tantek, aaronpk - want me to raise a feature request? I'd spotted this before with Monocle/Aperture not picking up some posts (when I first posted `e-content`), but hadn't gotten around to raising an issue with Xray
capjamesg An interesting story: https://gist.github.com/aconbere/1982a5eb17b77817017a3da50914732f
[dshanske] [manton] Yes
[manton] I’m considering starting a new project that will be based on OSM data, with a places API on top. It would be kind of a standalone, open source web app but I would host a version of its database inside Micro.blog so that things work well by default with Micro.blog clients. Still in the experimenting phase, so not 100% sure yet.
[schmarty] sknebel: i do post checkins! just haven't gone to many places recently. did a bunch yesterday because why not.
Zegnat capjamesg: re the search talk in #indieweb, I build that once upon a time. https://github.com/Zegnat/indiewebsearchengineproofofconcept It would take a search query, go and execute the search on a handful of IndieWeb WP sites, and then combine all the results.