On 2009/03/02 23:59:30, rodrigo.gallardo wrote: > Message templates are of reduced utility without template expansion ...
16 years, 11 months ago
(2009-03-03 13:15:16 UTC)
#5
On 2009/03/02 23:59:30, rodrigo.gallardo wrote:
> Message templates are of reduced utility without template expansion as in
> Activities, but we seem to have overlooked that during the spec proposal
phase.
>
> Should we request a spec clarification adding a templateParameters field, or
> would that be better left as a proposal for the next spec version?
supporting standardized templating for message submission does seem like a good
idea. We can raise this point with the spec list and see what the response is.
Looks mostly fine. A couple of nits and one comment on operation naming. http://codereview.appspot.com/20106/diff/51/1026 File ...
16 years, 11 months ago
(2009-03-03 19:24:30 UTC)
#6
Looks mostly fine. A couple of nits and one comment on operation naming.
http://codereview.appspot.com/20106/diff/51/1026
File
java/social-api/src/main/java/org/apache/shindig/social/opensocial/model/Address.java
(right):
http://codereview.appspot.com/20106/diff/51/1026#newcode65
Line 65: private static final Map<String, Field> lookup = new HashMap<String,
Field>();
technically should be concurrent map.
http://codereview.appspot.com/20106/diff/51/1029
File
java/social-api/src/main/java/org/apache/shindig/social/opensocial/service/MessageHandler.java
(right):
http://codereview.appspot.com/20106/diff/51/1029#newcode88
Line 88: @Operation(httpMethods = "POST", bodyParam = "data")
is there a better name than "data" for the body?
http://codereview.appspot.com/20106/diff/51/1029#newcode89
Line 89: public Future<?> post(SocialRequestItem request) throws
SocialSpiException {
if the operations are more CRUD like I've been using create/update for the
method names instead of get/post which means they will be exposed to RPC as
messages.create and messages.update and by implication to oslite as such which
is more inline with the JS APIs. See ActivityHandler etc.
updates applied, going to commit to trunk. http://codereview.appspot.com/20106/diff/51/1026 File java/social-api/src/main/java/org/apache/shindig/social/opensocial/model/Address.java (right): http://codereview.appspot.com/20106/diff/51/1026#newcode65 Line 65: private ...
16 years, 11 months ago
(2009-03-03 21:42:10 UTC)
#7
updates applied, going to commit to trunk.
http://codereview.appspot.com/20106/diff/51/1026
File
java/social-api/src/main/java/org/apache/shindig/social/opensocial/model/Address.java
(right):
http://codereview.appspot.com/20106/diff/51/1026#newcode65
Line 65: private static final Map<String, Field> lookup = new HashMap<String,
Field>();
On 2009/03/03 19:24:31, louiscryan wrote:
> technically should be concurrent map.
actually it's only modified at class instantiation, so should be fine as-is
http://codereview.appspot.com/20106/diff/51/1029
File
java/social-api/src/main/java/org/apache/shindig/social/opensocial/service/MessageHandler.java
(right):
http://codereview.appspot.com/20106/diff/51/1029#newcode88
Line 88: @Operation(httpMethods = "POST", bodyParam = "data")
On 2009/03/03 19:24:31, louiscryan wrote:
> is there a better name than "data" for the body?
entity? hard to say since this data can be a message collection or a message...
http://codereview.appspot.com/20106/diff/51/1029#newcode89
Line 89: public Future<?> post(SocialRequestItem request) throws
SocialSpiException {
On 2009/03/03 19:24:31, louiscryan wrote:
> if the operations are more CRUD like I've been using create/update for the
> method names instead of get/post which means they will be exposed to RPC as
> messages.create and messages.update and by implication to oslite as such which
> is more inline with the JS APIs. See ActivityHandler etc.
agreed. done.
Issue 20106: OpenSocial Message API changes
(Closed)
Created 16 years, 11 months ago by Paul Lindner
Modified 15 years, 9 months ago
Reviewers: louiscryan
Base URL: http://svn.apache.org/repos/asf/incubator/shindig/trunk/
Comments: 28