IRC log for #rest, 2016-03-11

https://trygvis.io/rest-wiki/

All times shown according to UTC.

Time	Nick	Message
00:15		fumanchu joined #rest
01:33		tbsf joined #rest
01:36		fumanchu joined #rest
03:28		tbsf joined #rest
06:02		bigbluehat joined #rest
06:10		angular_mike___ joined #rest
06:10		fumanchu joined #rest
06:48		Coldblackice_ joined #rest
06:57		timg___ joined #rest
07:17		timg___ joined #rest
07:21		bigbluehat joined #rest
07:53		bigbluehat joined #rest
08:33		gigo1980 joined #rest
08:41		keex joined #rest
08:53		graste joined #rest
10:10		interop_madness joined #rest
10:23		fumanchu_ joined #rest
10:38		timg___ joined #rest
11:29		keex joined #rest
14:19		tbsf joined #rest
15:21		quimrstorres joined #rest
15:24		tbsf joined #rest
15:32	pdurbin	Has anyone heard of an API where when you want to hit it twice you have to put a 500 millisecond sleep in between the two calls or the second call times out? Is there a name for this type of throttling?
15:32		tbsf joined #rest
15:56	pdurbin	here's the way a coworker expressed it:
15:56	pdurbin	"While in the process of testing I noticed that if I send in GET and POST requests in fairly rapid succession say 3 or more in a row, the third or fourth fails to respond and my application expecting a response hangs. Is there a limit to the number of request that can be made over a short length of time? Have other applications run into this issue and how have they addressed it?"
16:06		StatelessCat joined #rest
16:16	* fumanchu_	wonders if that's accidental throttling simply due to poor performance on the server
16:17	fumanchu_	if those are being sent from a browser, it's common to hit a browser-imposed limit on connections and blame the server though
16:18	fumanchu_	like, 2
16:20		tbsf joined #rest
16:21	trygvis	sounds like your db connection pool is full
16:22	trygvis	isn't that limit at least 8 now? or even much larger
16:28	pdurbin	yeah, I was thinking it was a design decision but maybe the server is having trouble handling load or something. it's just strange that adding a half-second sleep seems to help consistently
16:34	trygvis	I've been wondering about adding a 10 second delay for requests on resources that should be cached
16:34	trygvis	or for bursts greater than one
16:34	trygvis	that'll teach them!
16:35	trygvis	but I'll need to add a way for clients to poll for changes first
16:36	fumanchu_	in http3, server calls you
16:37	trygvis	:D
16:37	fumanchu_	to pushing you the advertises
17:18		keex joined #rest
17:41		Left_Turn joined #rest
17:45		gigo1980 joined #rest
19:11		tbsf joined #rest
19:42		timg___ joined #rest
20:05		tbsf_ joined #rest
20:51		fumanchu joined #rest
22:06		keex joined #rest
22:32		tbsf joined #rest
22:47		timg___ joined #rest
22:59		wCPO joined #rest
23:04	wCPO	Is it bad practice to hide some field depending on if the user is authorized or not?
23:06	whartung	no
23:06	whartung	in other fields, we call that "Security"
23:07	wCPO	whartung, thanks! I'm trying to "design" my rest api.. Before I code it.
23:07	whartung	nah, code first, then design it -- how do you know your design works if you ahven't coded it!??
23:09	wCPO	whartung, it more a kind of conversation/rewrite in rest (and go).
23:10	whartung	yea I was being facetious
23:14		Macaveli joined #rest
23:23	wCPO	whartung, initial proof of concept: http://sprunge.us/hTAe I need to fix some of the names, the json is request
23:24	wCPO	It is for some Debian digital signage box, which then call "/v1/players/:player_id" and use that as config.
23:31		tbsf joined #rest
23:44		anth0ny joined #rest
23:57	whartung	ok wCPO

https://trygvis.io/rest-wiki/