GWGaaronpk: How do you want to store your weather data? I see your notation about adding it to compass as an open issue
snarfed, dougbeal|imac, [tantek], [kenbauer], tw2113, jjuran, danyao, ingoogni, cweiske, [Rose], [xavierroy] and KartikPrabhu joined the channel; mblaney1 left the channel
aaronpkGWG: in my posts it's stored as another property along with everything else. I also have a separate Compass database that stores the weather info every five minutes or so, I think it stores the JSON that Atlas returns
[Rose]The Austrian government has a whole website of free APIs you can use, but it seems weather doesn't make an appearance except as a historical "statistical book"
[Rose]Though that does take you over the 1000/month Dark Sky allows, but every hour is still quite accurate, and if you make a larger call you could use it to backfill other data
[grantcodes]!tell doubleloop You are exactly right about if I build granary integration into together then it is only in together. An ideal way would probably be to have an external client that could work on it's own. But perhaps with a simple rest api or something that other clients can integrate to make life easier for their users?
[Rose]Aha, figured out the API for ZAMG, you get a CSV file and the contents change every hour. Why they don't use JSON like everyone else I have no idea, but it's a start!
ZegnatSure. We probably have a weather page you can add to. Or add -developer to the wiki page name and start a new page just for development purposes (like /Webmention has)
[jgmac1106] and [kimberlyhirsh] joined the channel
[eddie]It seems like all the "serverless" providers max out at 10mb per payload (which in practice ends up about 5mb of actual file size). which is not very big for a media endpoint
[eddie]The other option was to try to upload directly to say Amazon s3, so I thought my Micropub endpoint could create presigned s3 upload urls whenever apps do the config query
ZegnatI mean, there is a reason why it is defined seperately from the micropub endpoint. Specifically because you may want to use separate storage and so on
[Rose]Or perhaps clients should automatically check to see if all is well from time to time, which if the timeout could be set in q config might alleviate Eddie's issue
[eddie]That's true, and we don't currently have a solution for that outside of OAuth's refresh tokens. But we haven't even thought about how those should work
ZegnatYou could write an entire post, hit submit, only for the client to find your token has expired. I wonder how that is handled today. It would suck for you to be suddenly logged out and for the post to be lost
ZegnatEh, that sounds like a bad feature. I often write draft emails on the train, where I would not have access to a server either because of lack of connectivity
[Rose]And for something like Quill, if it finds the endpoints are invalid I would hope we could extend it to give the user a .txt file download they can save
[Rose]Maybe not the most elegant of solutions, but it means they can easily copy/paste the data back in after re-auth if the draft can't be maintained any other way
[eddie]for OAuth 2.0, when providing an access_token that expires, it will return an "expires_in":3600 attribute. So if we started paying more attention to that in Micropub clients, we could also add that same thing to the Config query
[eddie]I know Nintendo Switch is using something like Open ID so it's not pure OAuth 2.0, but it's similar and I know they were sending back 401's when my tokens were expired
ZegnatAnd of course “log them out” may just mean you disable the post button and remove the token from storage. The client doesn’t need to throw up a boot screen and lock away drafts
aaronpknever assume the user has an internet connection, and also don't assume that just because a token expired that it means you should completely reset the app state
[eddie]Yeah I think one challenge is because most stuff started out to be Micropub web apps without much state it didn't matter if you logged out and back in
aaronpkI'm just excited I can write offline apps without having to deal with the native development platforms now :-) no offense to native app development, it's just overkill in some cases
[eddie]There's something nice about being able to just load up a website. I just find a lot of "web app" experiences provide the functionality but they feel clunky. That is the main thing that pushes me to native apps
aaronpkI have two that I use regularly now, Teacup and Airplane Time. They're not perfect by any means, but they are pretty simple and would be overkill as native apps, and also I probably wouldn't have been able to make a very good native app experience anyway
[eddie]from what I've been reading (i haven't tested it yet) if I write a function in serverless that redirects to another url, but then I sent an upload of 10mb to it, it would get a 500 error because of the size rather than the 302 redirect
LoqiIt looks like we don't have a page for "serverless" yet. Would you like to create it? (Or just say "serverless is ____", a sentence describing the term)
[schmarty]aaronpk: last week i got a shortcut working that lets me make a read post out of any goodreads.com book link from safari by sharing it to the shortcut. iti s raddd.
[eddie]it's tough, this might mean I just can't use serverless in this way but with my outage on DreamHost's compute platform makes me want to get as much off of my single box as possible
[schmarty]haha yep. i was able to crop the image, send it to my media endpoint via shortcuts, and edit the post to use the new image (had to do that part by hand...)
[eddie][schmarty] I added my investigation into Lambda and S3 media endpoint usage to the wiki. Feel free to comment on if there seems to be any missing info
sknebel[eddie]: if you're in AWS land, can't you make a fixed url through the loadbalancer layer (whatever that's called in that case, api gateway?) that sends traffic to the changing s3 url?
aaronpk[eddie]: yeah most of the AWS stuff does not expect you to be using it standalone. They kind of expect you to use all their products together, like adding a load balancer in front of ec2 servers
LoqiIt looks like we don't have a page for "Airplane Time" yet. Would you like to create it? (Or just say "Airplane Time is ____", a sentence describing the term)
[jgmac1106], [kenbauer] and [Rose] joined the channel
LoqiIt looks like we don't have a page for "offline first" yet. Would you like to create it? (Or just say "offline first is ____", a sentence describing the term)
[eddie]I have a new Media Endpoint on Lambda road to research. Apparently Lambda@Edge (aka a lambda function running on amazon's CDN) can catch error statuses and do things like return a redirect
[eddie]so I haven't looked into it quite enough yet, but I'm wondering if I can set up a url that is my "media endpoint" and then if I catch the 50-something error when the payload is too big and in that case return a redirect to my presigned s3 url
[tantek]offline first is a method of developing a website that once a user has visited it once on their device, it will always provide at least some content ([[posts]]) and or features (like [[creating]] posts) when they revisit even without internet access, using technologies like Service Workers and Local Storage.
LoqiIt looks like we don't have a page for "offline first" yet. Would you like to create it? (Or just say "offline first is ____", a sentence describing the term)
[tantek]offline first is a method of developing a website that once a user has visited it once on their device, it will always provide at least some content ([[posts]]) and or features (like [[creating]] posts) when they revisit even without internet access, using technologies like Service Workers and Local Storage.
calumryanThe closest we have is possibly https://indieweb.org/Progressive_Web_App but not Offline first, likewise no page for Mobile first? They’re quite broad terms for methodology as well as technologies to cover
@johanbove ↩️ Done! Implemented the Microformats to my Jekyll blog posts and also added http://webmentions.io links. Excited to have joined the #Indieweb! (twitter.com/_/status/1113704670701264896)
[kevinmarks] working on that
Zegnat Oh, wow, SMHI is putting all their data under CC0 apparently. I am going to look into this. https://opendata.smhi.se/apidocs/
[grantcodes] !tell doubleloop You are exactly right about if I build granary integration into together then it is only in together. An ideal way would probably be to have an external client that could work on it's own. But perhaps with a simple rest api or something that other clients can integrate to make life easier for their users?
@timweston ↩️ @HorizonsOne @jefflipack @MoJoDisney Yep! And lots of indieweb stuff has that. See https://indieweb.org/Webmention (twitter.com/_/status/1113788088688496640)
[jgmac1106] adafruit micrpub weather rock...not publishing weather on my blog until i build that
[eddie] Alright tried to sum up these thoughts in a extension issue (https://github.com/indieweb/micropub-extensions/issues/17)
[schmarty] bringing back the issue of pre-signed S3 form URLs - that sounds like plumbing that the serverless endpoint should hide
KartikPrabhu what is serverless
Loqi It looks like we don't have a page for "serverless" yet. Would you like to create it? (Or just say "serverless is ____", a sentence describing the term)
@meyerweb Is there a way in Firefox to add a site to the content-blocking exceptions list WITHOUT using the address bar dropdown? (twitter.com/_/status/1113830765463318528)
[tantek] offline first is a method of developing a website that once a user has visited it once on their device, it will always provide at least some content ([[posts]]) and or features (like [[creating]] posts) when they revisit even without internet access, using technologies like Service Workers and Local Storage.
calumryan The closest we have is possibly https://indieweb.org/Progressive_Web_App but not Offline first, likewise no page for Mobile first? They’re quite broad terms for methodology as well as technologies to cover