greptilian logo

IRC log for #rest, 2014-09-10

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

| Channels | #rest index | Today | | Search | Google Search | Plain-Text | plain, newest first | summary

All times shown according to UTC.

Time S Nick Message
00:06 `0660 joined #rest
00:44 `0660 joined #rest
01:15 Arahael joined #rest
01:15 Arahael So, I want to write a restful api.
01:16 Arahael But I want to add search functionality to it - how can I do this while remaining restful?
01:16 Arahael Or is such a search function orthogonal to rest?
01:29 warehouse13 joined #rest
01:47 talios 'lo
01:48 talios I don't think so - you'd just have a search resource that returns links to other resources.
01:50 `0660 joined #rest
01:55 Arahael As orthogonal to REST?
02:00 talios no, as REST.    search is just a resource, that happens to return a representation of a query, with links to other resources.
02:01 Arahael But that representation can change - and the client has to construct the uri?
02:02 talios in our system for example, if we GET /api/search?s=talios we return a JSON document linking to matching documents.  representations change over time - like all resources.
02:02 talios some other resource might link to your search service with a URI-Template, i.e. Link-Template: <http://localhost/api/search{?term}>; rel="search"
02:03 talios then the client just resolves the template provided by the link-template, indirection and freedom for the server to change its location.  Sure, the rel and the template values are fixed...
02:04 Arahael Where can I read about uri templates?
02:10 Arahael (SPecifically, those link templates)
02:10 talios http://tools.ietf.org/html/rfc6570 is the RFC for them
02:12 Arahael Sweet, thanks for that.
02:13 Arahael So my REST landing page would probably provide an index to the various resources, including the search resources, and allow the client to start navigating from there. Is that a good approach?
02:18 systmkor joined #rest
02:21 tr3onlin_ joined #rest
02:25 pdurbin sounds discoverable, which is good
02:32 talios sounds good
02:38 begriffs joined #rest
02:38 Arahael Does providing an xml representation give any advantage to a REST API?
02:53 talios over JSON?  XML is harder to consume nicely from Javascript, or so I hear.
02:53 talios I guess it depends on your target users.  XML gives you DTDs/schemas you can exploit - if you care
03:19 DrCode joined #rest
03:21 systmkor1 joined #rest
03:28 tr3online joined #rest
04:06 proteusguy joined #rest
04:07 Arahael I guess it really gets down to that.
05:01 systmkor joined #rest
05:08 systmkor1 joined #rest
05:41 systmkor joined #rest
06:44 talios joined #rest
06:59 tr3online joined #rest
07:09 fumanchu_ joined #rest
07:16 rosstuck joined #rest
07:21 fumanchu joined #rest
07:26 xyz_ joined #rest
07:26 xyz_ hi
07:27 xyz_ i was trying to get the yahoo mail using yahoo yql rest api
07:27 xyz_ but i am getting error like  {"error":{"lang":"en-US","diagnostics​":null,"description":"Authentication Error. The table ymail.messages requires a higher security level than is provided, you provided ANY but at least USER is expected"}}
07:27 xyz_ any idea what it means?
07:46 graste joined #rest
07:54 shrink0r joined #rest
08:00 tr3online joined #rest
08:01 interop_madness joined #rest
09:12 martinfilliau joined #rest
09:15 shrink0r joined #rest
09:45 fumanchu_ joined #rest
09:48 Left_Turn joined #rest
10:46 Jarda joined #rest
11:36 jcromartie joined #rest
11:44 Left_Turn joined #rest
12:39 jcromartie joined #rest
14:11 Mxyzpltk joined #rest
14:21 rosstuck joined #rest
14:43 blongden Hi all - I've made a load of updates to the vnd.error draft based on feedback thus far... would appreciate anyone with a little time to go through it and pick holes... https://github.com/blongden/vnd.error
14:43 blongden It occured to me that I first created the draft over two years ago. Time flies.
15:02 benedikt left #rest
15:24 pdurbin "a simple way of expressing an error response in JSON" ... interesting
16:00 pezra joined #rest
16:46 fumanchu joined #rest
16:52 fumanchu joined #rest
17:02 fumanchu_ joined #rest
17:03 systmkor joined #rest
17:17 fumanchu joined #rest
17:20 fumanchu_ joined #rest
17:30 fsvehla joined #rest
17:53 cigarshark joined #rest
17:58 pezra joined #rest
18:05 proteusguy joined #rest
18:30 tr3online joined #rest
18:41 tr3online joined #rest
18:42 shrink0r joined #rest
19:06 pezra joined #rest
19:12 jcromartie joined #rest
19:32 danizord joined #rest
19:45 saml hey, a resource has prev and next  (something like doubly linked list)
19:45 saml if I PUT /resource/2,  it might affect up to three resources.  /resource/1, /resource/2, /resource/3    (if /resource/1 's next was /resource/3/)  does it make sense?
19:46 saml PUT becomes non-idempotent
19:46 saml PUT /resource/2     has side effects
19:47 shrink0r joined #rest
19:57 shrink0r joined #rest
20:00 fumanchu_ saml: so maintain the order as a separate resource
20:02 saml example?
20:02 saml /resource-orders
20:03 fumanchu_ sure. that can return an array or graph of links to resource/1, /2 etc..
20:04 fumanchu_ that also allows you a powerful separation between initialization (the PUT to resources/2) and activation (adding it to the array or graph)
20:05 fumanchu_ and allows for multiple orders: per-user, or different indices for different needs
20:19 saml there are over 500k resources  like /resource/1
20:44 pezra joined #rest
20:45 fumanchu joined #rest
21:08 Mxyzpltk joined #rest
21:48 systmkor1 joined #rest
21:50 talios joined #rest
21:52 talios joined #rest
22:21 systmkor2 joined #rest
22:32 tr3online joined #rest
22:35 pezra joined #rest
22:40 tr3online joined #rest
23:03 pezra joined #rest
23:22 systmkor3 joined #rest

| Channels | #rest index | Today | | Search | Google Search | Plain-Text | plain, newest first | summary

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