Is OAuth2 working for anyone?  I'm getting this (sensitive bits removed for obvious reasons):

$ curl -s -d'' 'https://auth.hackaday.io/access_token?client_id=${ID}&client_secret=${SECRET}&code=${CODE_FROM_CALLBACK}&grant_type=authorization_code' | python -m json.tool
{
    "error": "invalid_access_code",
    "error_description": "Invalid access code"
}

I've also tried POSTing that in the body, although the examples use URI query parameters and that's what my code does now:

https://github.com/sjlongland/hackaday.io-spambot-hunter/blob/master/hadsh/hadapi/hadapi.py#L262-L274

Did my API account get blocked again for too many queries?  Just noticed my website's skull counter isn't working any more.  It is registered as "

is that seriously a bug with double quotes for comments? I will add test behind " test

interesting, yeah it removed what was in the double quotes.  It is registered as (no quotes) skull counter widget

Silly question… how do you view the "Things I've built" part of profiles from the HAD API?  I'm seeing some spambots using that to spruik their wares and that's a blind-spot for me.

when I compare likes / skulls count from the API vs. the page, I get a difference of 869(API) to 926(page). I guess you don't count in page likes? 

Hi! I'm trying to make a simple HackChat-IRC gateway, but the server keeps returning 500 to this: https://api.hackaday.io/v1/message/2369?api_key=API_KEY . I can get the list of conversations I have, checked that the HackChat convo is there, but this particular page doesn't want to work for me. I also can't seem to be able to send messages, either, same 500 error. (BTW, why the message receive and send APIs are so different?)

Not exactly API question, but is there any chance of being able to add tracking to projects? Like ability to add google tracking id or some other third party tracker? So that I could gauge what brings most traffic(sources etc) as currently I am totally blind & clueless :D Thanks

can't you add &reference=HackadayProjectLink or something to your page and log stuff like that?

Thank you for the suggestion davedarko it somehow skipped my mind :) although it does have its own pro & cons :D eg can't afford to pay for hosting or lose the domain ... (sort of unstable times I live in)...

{"code":400,"type":"Bad Request","message":"API key incorrect. Please visit https:\/\/dev.hackaday.io\/ to register an API key"}

Is the api still maintained/usable? I'm getting a 500 error whith the Objective-C example app when it tries to fetch "http://api.hackaday.io/v1/feeds/global?api_key=***"; (I filled the Client_id, secret_id and api_key parameters). Any thoughts?

There seems to be no option for getting "myFeed", just the global feed.  Is myFeed something that in the future roadmap?

I am getting a 404 when i redirect to the authorize link  
https://hackaday.io/authorize?client_id=CLIENT_ID&state=&response_type=code

not sure if this is new idea, but for those who post on here and also maintain another site with a blog, it is tedious posting to two places.   This code makes project logs appear as blog posts on another site.

  https://hackaday.io/project/12657-hack-a-blog

Is there a way to GET projects by tag? Seems like search/projects....&search_term=TAG doesn't do the trick

Never mind, I take it back :P My bad.

I'm playing around with the api, is there no publish/draft state on the pages?

in api docs... https://dev.hackaday.io/doc/api/get-comments

missing example API for call "comments/events/:id"

No prob!  I was looking into a way to keep up better with the Hacker Channel chat -- see if anyone has written to/about me without having to scroll back through so many messages manually.

I have a python script that I was using that would read the messages from the page source.  I could clean it up and post it in a private project next week.  I can add you to the project if you think it is something you would find useful.

Definitely! 

this sounds similar to #HackaDump :-)

Finally got around to looking into scraping the Hacker Channel -- it's not so easy.  It's written in Angular, so all of the database -> text stuff is done dynamically in Javascript.  Which, AFAIK, means using something like Selenium/Webdriver that can interpret the JS.  

So it's doable, but nontrivial. I may hack at it if I get time, but the cleanest solution is a backend API.

Yeah, I am using selenium, but I was going to try it with mechanize to see if I could get it to work.  I think I will have time this afternoon to get something posted.

Is the "message" part of the API working?  I've been playing around, and can get the list of my conversations from http://api.hackaday.io/v1/message , but then when it comes time to see an individual convo (indexed by 'id' from above), I get a 500 internal server error back.

It says in the docs that the server will error out if I don't have permissions for the conversation, but I've double-checked the numbers with the normal web interface and the list from /message.

I should be able to read message 15700 at : http://api.hackaday.io/v1/message/15700 right?  Or am I doing something obviously wrong?

unfortunately, that part doesn't work. it's in our todo list : )

Is there any where that I can check my rate limits? I.e. I want to be able to see how close I am to the 30k api limit.

@Ivan Lazarevic @Edison Chen My API key for "skull count widget" has exceeded its monthly limit and now my skull counter is broken.  Not sure if there is a better way to implement it, what is the monthly limit for mine?  I have had 3000 pageviews in the last 30 days, my counter only pings the API once very 30 seconds and it only does it while someone is actively on the page.   Do you see there being a better way to implement this?

@Ivan Lazarevic Any word on this?  what is my monthly API limit?

monthly limit for read is 30000 for now. it might be increased in the future. 

you should consider caching api response instead sending api call for each page view.

Hey Alex - just to reiterate on this one. Monthly limit for number of API calls per given API key is 30000. We have checked the counters and the number of issued requests with your API key over the last month is 32941, which is hitting the max requests limit, hence the issue. Hope this helps.

I changed my code so the theoretical maximum monthly API calls is now 21,600 (one every two minutes).  I actually was caching the skull count and only pinging API if the cache was older than 30 seconds.  If someone opened my site and just sat on it for several weeks without closing the browser, maybe that would cause this issue.  I have ajax that checks the cache every 10 seconds while someone is on the page, if cache was older than 30s it gets a fresh value from the API.  The hope was that someone could go to hackaday.io and skull the project, return to my page and see it update quickly.  Oh well, no big deal.  Would be ideal if the API pushed changes to me somehow so I don't have to continually check, that's not an option in the API at this point though.

You guys should check this project out, should be interesting: https://hackaday.io/project/5803-hackadayio-analytics

The release of the hackaday API has obsoleted a global feed scraper project I *just* did, but I figured I'd post it here in case anyone is interested: 

https://hackaday.io/project/5611-hackadayio-global-feed-scraper