| Time |
S |
Nick |
Message |
| 00:24 |
|
|
sfisque joined #rest |
| 00:36 |
|
|
sfisque joined #rest |
| 00:49 |
|
|
fuzzyhorns joined #rest |
| 01:43 |
|
|
wsiqueir joined #rest |
| 02:43 |
|
|
fuzzyhorns joined #rest |
| 03:05 |
|
|
baweaver joined #rest |
| 03:33 |
|
|
baweaver joined #rest |
| 04:10 |
|
|
baweaver_ joined #rest |
| 04:21 |
|
|
wsiqueir joined #rest |
| 06:44 |
|
|
rosstuck joined #rest |
| 07:30 |
|
|
baweaver joined #rest |
| 08:11 |
|
|
interop_madness joined #rest |
| 08:20 |
|
|
graste joined #rest |
| 08:55 |
|
|
quimrstorres joined #rest |
| 09:00 |
|
|
Coldblackice joined #rest |
| 09:01 |
|
|
quimrstorres joined #rest |
| 10:59 |
|
|
quimrstorres joined #rest |
| 11:19 |
|
|
shodan` joined #rest |
| 11:20 |
|
shodan` |
Hey, I was wondering if it's bad practice to make a client send a json/xml object to the server. This feels like the most reasonable thing since I'm building an object with both properties and a list of items with properties |
| 11:44 |
|
|
interop_madness joined #rest |
| 11:54 |
|
|
quimrstorres joined #rest |
| 12:29 |
|
pdurbin |
shodan`: can you show us what you're sending? http://danceb.in |
| 12:44 |
|
shodan` |
not really, it's properitary, drafting an api for a customer atm |
| 12:45 |
|
shodan` |
it's a collection of items and lists of objects |
| 12:45 |
|
shodan` |
so {"thing":{"blah":true}, "beep": false, "stuff": []} |
| 12:51 |
|
pdurbin |
shodan`: a little bit like this maybe: https://github.com/IQSS/dataverse/blob/v4.1/scripts/search/tests/data/dataset-finch1.json |
| 12:51 |
|
shodan` |
yeah |
| 12:52 |
|
shodan` |
I kinda wanna get it right because giving support will mean a 2+ hour drive |
| 12:52 |
|
shodan` |
and then a 2 hour drive back |
| 13:03 |
|
shodan` |
pdurbin: ping? |
| 13:03 |
|
|
dEPy joined #rest |
| 13:11 |
|
pdurbin |
shodan`: yeah, down with 2 hour drives |
| 13:33 |
|
|
shodan` joined #rest |
| 13:55 |
|
|
quimrstorres joined #rest |
| 14:00 |
|
|
darkangel joined #rest |
| 14:04 |
|
darkangel |
If you have "articles" and "players", where the article can reference one or more players (and you wish to save this link), what are the options for modelling this? I'm not sure about /players/{id}/articles, bc it makes it look like the article is "owned" or "created" by the player. |
| 14:57 |
|
|
fuzzyhorns joined #rest |
| 15:00 |
|
fumanchu_ |
darkangel: I would tend to keep the same form but be more descriptive. Something like /players/{id}/references, or even /players/{id}/articles/mentions separate from /players/{id}/articles/authored, depending on what other kinds of links there are (or might someday be) between players and articles |
| 15:01 |
|
darkangel |
fumanchu_: Ya, I was thinking about something like: /players/{id}/linked/articles |
| 15:02 |
|
fumanchu_ |
sure. I'd be a bit careful with that since "link" is so generic. If there are multiple kinds of links, you'd either want those link types in the URL or as an attribute in the entities |
| 15:04 |
|
darkangel |
fumanchu_: references is also quite generic, but I understand your point. :) |
| 15:04 |
|
fumanchu_ |
:) |
| 15:06 |
|
darkangel |
fumanchu_: There are also features, videos, and streams, so: /players/{id}/?/features, /players/{id}/?/videos, /players/{id}/?/streams |
| 15:07 |
|
|
rickharrison joined #rest |
| 15:07 |
|
darkangel |
It's sort of a "link" resource that you're adding, not the feature itself. |
| 15:09 |
|
darkangel |
I don't think there's any "perfect" way of doing this. |
| 15:09 |
|
sfisque |
if you look at the uri as a location in a federation, rather than a hierarchical ownership model, the original uri looks fine (/player/{id}/article) |
| 15:10 |
|
darkangel |
sfisque: Until players are able to create articles :) (may never happen, just an example) |
| 15:27 |
|
|
interop_madness joined #rest |
| 15:36 |
|
sfisque |
aye, but i think you miss my point. the uri does not necessarily HAVE to be an analog of ownership, even if ownership is "part" of the schema. |
| 15:37 |
|
sfisque |
just because A ->(n) B , does not necessarily mean A owns B, just that A relates to B in a potentially 1->n relationship |
| 16:05 |
|
darkangel |
sfisque: Yep, I understand. I guess I just see "ownership" as the default relationship type, and might prefer to reserve that for later use. I'll think about it more. Thanks. :) |
| 16:11 |
|
|
quimrstorres joined #rest |
| 16:20 |
|
|
foist joined #rest |
| 17:11 |
|
|
quimrstorres joined #rest |
| 18:01 |
|
whartung |
random, unrelated -- but, anyone familiar with tcpdump/wireshark? |
| 18:12 |
|
fumanchu_ |
only slightly |
| 18:12 |
|
fumanchu_ |
lots of inspecting HTTP packets while working on CherryPy ;) |
| 18:15 |
|
whartung |
I think I found what I need -- just trying to eliminate all of the TCP gobbledy gook and just get my data packets. |
| 18:15 |
|
whartung |
I'm capturing with tcpdump then loading those files in to wreshark. THere I can "Follow TCP Stream" and I can export that as C arrays (which is good for my task). |
| 18:15 |
|
fumanchu_ |
right |
| 18:16 |
|
whartung |
I with I could just do it with tcpdump, but I can't figure that out |
| 18:16 |
|
fumanchu_ |
I always follow https://www.wireshark.org/docs/wsug_html_chunked/AppToolstcpdump.html |
| 18:16 |
|
fumanchu_ |
tcpdump -i <interface> -s 65535 -w <some-file> |
| 18:16 |
|
whartung |
yea, that's what I do |
| 18:17 |
|
|
DebolazX joined #rest |
| 18:20 |
|
|
DebolazX left #rest |
| 19:00 |
|
pdurbin |
whartung: so you're all set with wireshark? |
| 19:05 |
|
whartung |
seem to be |
| 19:05 |
|
whartung |
still a pain... |
| 19:05 |
|
whartung |
"how hard can this be" |
| 19:16 |
|
|
Left_Turn joined #rest |
| 19:26 |
|
pdurbin |
:) |
| 19:55 |
|
|
metasansana joined #rest |
| 19:55 |
|
|
metasansana joined #rest |
| 20:45 |
|
|
Left_Turn joined #rest |
| 20:52 |
|
|
rosstuck joined #rest |
| 21:53 |
|
|
Coldblackice joined #rest |
| 22:03 |
|
|
fuzzyhorns joined #rest |
| 22:41 |
|
|
linux_dr joined #rest |
| 22:53 |
|
|
quimrstorres joined #rest |
| 23:23 |
|
|
cvander joined #rest |
| 23:25 |
|
|
quimrstorres joined #rest |
