lp:~robru/friends/urlshorten

Branch merges

Propose for merging

No branches dependent on this one.

Merged into lp:friends at revision 115

Ken VanDine: Approve on 2013-01-09

Diff: 187 lines (+54/-76)

5 files modified

friends/main.py (+1/-2)
friends/service/dispatcher.py (+30/-1)
friends/service/shortener.py (+0/-69)
friends/service/templates/com.canonical.Friends.URLShorten.service.in (+0/-3)
friends/tests/test_dispatcher.py (+23/-1)

api

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related bugs

Related blueprints

111. By Robert Bruce Park on 2013-01-05

Merge URLShorten into Dispatcher class.

URLShorten only had the one method, so it didn't really make much
sense for it to be separate.

110. By Robert Bruce Park on 2013-01-03

Prevent multiple instances of friends-service from running.

109. By Robert Bruce Park on 2013-01-03

Revert commit 95 because it breaks support for Quantal.

108. By Robert Bruce Park on 2013-01-03

Revert commit 98 in order to restore compatibility with Quantal.

107. By Robert Bruce Park on 2013-01-02

Revert revision 94 because it broke compatibility with Quantal. Sorry.

Please re-apply r94 once Quantal is EOL! ;-)

106. By Robert Bruce Park on 2012-12-18

Time thread execution.

105. By Robert Bruce Park on 2012-12-18

More explicit FacebookError exception class.

104. By Robert Bruce Park on 2012-12-18

Slightly improve error messages reported by Uploader class when there
is a zero-length response from the server.

(Thanks to Barry Warsaw for pointing this out).

103. By Robert Bruce Park on 2012-12-18

Fix docstring in Dispatcher.Upload.

(it now demonstrates making the mainloop, and corrected the callback
call signatures).

102. By Robert Bruce Park on 2012-12-14

Reduce memory usage by storing only row index ints rather than
DeeModelRowIter objects, which wastefully contain references to many,
many, many unnecessary python objects.

In my limited testing, this change dropped memory usage by ~10MBs for
a DeeModel containing 1660 rows, and I confirmed with the objgraph
module that triggering a refresh no longer results in any additional
python objects being kept around in memory -- in theory we have
abolished all memory leaks! In practise however, C code that we run
through GObject Introspection is free to leak as it pleases and this
tool would not detect it.