Time |
S |
Nick |
Message |
04:13 |
|
|
mgomezch joined #rest |
08:48 |
|
|
shrink0r joined #rest |
08:56 |
|
|
azr joined #rest |
09:10 |
|
|
azr joined #rest |
09:29 |
|
|
interop_madness joined #rest |
09:35 |
|
|
Left_Turn joined #rest |
10:50 |
|
|
ewalti joined #rest |
11:35 |
|
|
azr joined #rest |
12:05 |
|
|
mezod joined #rest |
12:25 |
|
|
azr joined #rest |
12:41 |
|
|
shrink0r_ joined #rest |
12:43 |
|
|
dEPy joined #rest |
13:38 |
|
|
aGHz joined #rest |
13:56 |
|
|
azr joined #rest |
14:15 |
|
|
saml joined #rest |
14:18 |
|
|
Andre-B joined #rest |
14:19 |
|
Andre-B |
in a restful api, with a search endpoint, if the search does not yield any results, shall I return http status: 200 and empty array, or http status 204 no body. |
14:19 |
|
Andre-B |
having a nice discussion about this right now with a coworker |
14:21 |
|
trygvis |
I would return 200 and empty body |
14:21 |
|
trygvis |
s,body,array, |
14:21 |
|
gamache |
200 with [] seems less surprising to me |
14:23 |
|
gamache |
though I tend to make my query responses look more like {results: []} with some other attributes on there too, like _links |
14:24 |
|
trygvis |
yep, me too |
14:26 |
|
|
shrink0r joined #rest |
14:29 |
|
|
nkoza joined #rest |
14:39 |
|
|
shrink0r_ joined #rest |
14:45 |
|
pdurbin |
yeah, I return 200 with {"status":"OK","data":{"items":[],"count_in_response":0 ... |
14:45 |
|
pdurbin |
http://guides.dataverse.org/en/latest/api/search.html |
15:14 |
|
|
mezod joined #rest |
15:22 |
|
|
azr joined #rest |
15:26 |
|
|
ralphschindler joined #rest |
15:36 |
|
|
azr joined #rest |
16:33 |
|
|
lemur joined #rest |
16:49 |
|
|
Mxyzpltk joined #rest |
17:06 |
|
trygvis |
pdurbin: why the "status" field? |
17:07 |
|
trygvis |
and the data-field, I would think just the value of "data" should be the payload |
17:09 |
|
fumanchu_ |
trygvis: http://beachpackagingdesign.com/wp-content/uploads/boxvox/6a00e54f0014bd88340134892d2314970c.jpg |
17:14 |
|
* fumanchu_ |
has seen client frameworks that don't provide access to HTTP status nearly as easily as they do to the payload |
17:28 |
|
pdurbin |
trygvis: we always return a status. another example: {"status":"ERROR","message":"q parameter is missing"} |
17:31 |
|
fumanchu_ |
pdurbin: what HTTP status is returned with that payload? |
17:32 |
|
pdurbin |
400 Bad Request |
18:46 |
|
|
adaro joined #rest |
18:52 |
|
trygvis |
if you have a rest framework that doesn't give access to the status code, use another framework |
18:52 |
|
trygvis |
tunnelling the status code in the document is really suboptimal design |
18:55 |
|
pdurbin |
trygvis: as I said, we return 200 or 400 depending |
18:57 |
|
trygvis |
it's still a wtf |
19:00 |
|
whartung |
could just be extra verbage to clafify details of the status code trygvis |
19:01 |
|
whartung |
or, if could be a direct affrontage to your family heritage. Hard to tell from here. ;) |
19:02 |
|
|
shrink0r joined #rest |
19:04 |
|
pdurbin |
trygvis: what if it said "message"? If leave off the "q" (query) at https://api.github.com/search/repositories it says "message": "Validation Failed" (and has an array of errors with "q... missing") |
19:05 |
|
pdurbin |
maybe you just object to it being "status" instead of "message" |
19:06 |
|
pdurbin |
I guess "status" has a special meaning for HTTP. |
19:17 |
|
|
tr3online joined #rest |
19:29 |
|
|
Andre-B joined #rest |
20:04 |
|
|
Andre-B_ joined #rest |
20:16 |
|
|
rosstuck joined #rest |
20:38 |
|
|
shrink0r_ joined #rest |
20:38 |
|
trygvis |
pdurbin: feels better, then you're adding information to the http status code |
20:39 |
|
trygvis |
I don't object to the name itself, mainly that it seemed to be a copy of the data from the http layer |
20:39 |
|
trygvis |
(but I would probably go for another name myself :)) |
20:43 |
|
pdurbin |
trygvis: I'll point the guy who wrote the code to this conversation and see what he thinks. thanks |
20:47 |
|
trygvis |
;) |
20:52 |
|
|
blahdeblah joined #rest |
20:52 |
|
|
blahdeblah joined #rest |
21:11 |
|
pdurbin |
I mean, we're supposed to ship real soon now so who knows if we'll change anything. |
21:11 |
|
pdurbin |
whartung: which reminds me, more success with shib these days: http://irclog.iq.harvard.edu/dataverse/2015-03-16#i_17107 |
21:12 |
|
whartung |
nice |
21:12 |
|
whartung |
oh I read your command presentation -- that was pretty good |
21:13 |
|
pdurbin |
oh! glad you liked it. it was with the same guy who added the "status" message :) |
21:13 |
|
whartung |
:) |
21:13 |
|
pdurbin |
so I'll only have to send him one link to these logs |
21:14 |
|
pdurbin |
whartung: you don't think we should have just used Spring Security? sometimes I wonder |
21:14 |
|
whartung |
I can't speak to it, I don't use spring, not famliar with it's capabilities |
21:28 |
|
|
shrink0r joined #rest |
21:36 |
|
|
Andre-B joined #rest |
21:49 |
|
|
ralphschindler joined #rest |
22:08 |
|
|
ralphschindler joined #rest |
22:11 |
|
|
ralphschindler joined #rest |
22:15 |
|
pdurbin |
whartung: did you write your own permission system? did we talk about this already? |
22:15 |
|
whartung |
I don't think so |
22:15 |
|
whartung |
"yes" lol |
22:22 |
|
pdurbin |
I mean, in Django or whatever there's some sort of permission system built in you can just use. |
22:22 |
|
whartung |
no doubt |
22:23 |
|
whartung |
no, we wrote our own roles/priveleges system |
22:23 |
|
whartung |
…of course... |
22:24 |
|
pdurbin |
like you do |
22:24 |
|
whartung |
it's the way I roll |
22:24 |
|
whartung |
or, role even |
22:24 |
|
* pdurbin |
groans |
22:27 |
|
pdurbin |
:) |
22:27 |
|
pdurbin |
whartung: and you'd do it again? nothing off the shelf is good enough? |
22:28 |
|
whartung |
I dunno. I'd have to look. I wouldn't use Spring, that's just me. So I'd have to find something else. |
22:29 |
|
pdurbin |
I've never used Spring but some people and work won't touch it. |
22:29 |
|
whartung |
well, the simple truth is that Spring is like getting a piece of personal electronics in Europe and bringing it home, finding that it won't fit your plugs |
22:31 |
|
whartung |
I just don't have the time and energy to "learn spring", especially for something like a collection of roles and identity. |
22:31 |
|
pdurbin |
you could move! ;) |
22:31 |
|
whartung |
what we USE is the JEE Principal and UserInRole |
22:31 |
|
whartung |
yea, exactly |
22:31 |
|
whartung |
that's what it would be like |
22:31 |
|
whartung |
moving to europe |
22:31 |
|
whartung |
everythings the same, but it's all different at the same time |
22:32 |
|
whartung |
they drive on the wrong side of the road over there |
22:34 |
|
pdurbin |
it's hard to tell what temperature it is, how much anything weighs |
22:34 |
|
whartung |
damn skippy |
22:34 |
|
whartung |
food tastes funny too |
22:36 |
|
whartung |
(Paris has the worlds best ham and cheese sandwiches…I dunno what it is, every block and 3 or 4 stores that sells them, they're ham..and..cheese….and they're awesome. I had one in the airport before I left, it was awesome too -- in the AIRPORT! High cuisine (/snort) indeed!) |
22:36 |
|
pdurbin |
so healthy tasting |
22:39 |
|
whartung |
pdurbin: this is our daily existence in the Stripes channel. https://www.dropbox.com/s/udo5b7wgdspky02/Screenshot%202015-03-16%2015.39.11.png?dl=0 |
22:42 |
|
pdurbin |
heh. this happens all the time in #circuits |
23:23 |
|
|
SupaHam joined #rest |
23:28 |
|
pdurbin |
trygvis: if you want to argue with that guy I @mentioned him here: https://twitter.com/philipdurbin/status/577580004583804928 :) |
23:56 |
|
|
adaro_ joined #rest |