Merge lp://staging/~mbp/bzr/198646-http-boundary-2.3 into lp://staging/bzr/2.3
Proposed by
Martin Pool
| Status: | Superseded | ||||
|---|---|---|---|---|---|
| Proposed branch: | lp://staging/~mbp/bzr/198646-http-boundary-2.3 | ||||
| Merge into: | lp://staging/bzr/2.3 | ||||
| Diff against target: |
137 lines (+88/-2) 4 files modified
bzrlib/errors.py (+13/-0) bzrlib/tests/test_http.py (+66/-0) bzrlib/transport/http/__init__.py (+1/-1) bzrlib/transport/http/response.py (+8/-1) |
||||
| To merge this branch: | bzr merge lp://staging/~mbp/bzr/198646-http-boundary-2.3 | ||||
| Related bugs: |
|
| Reviewer | Review Type | Date Requested | Status |
|---|---|---|---|
| Robert Collins (community) | Approve | ||
|
Review via email:
|
|||
Commit message
cope with proxies that drop the connection during a multipart response (bug 198646)
Description of the change
bug 198646 is that some buggy squids can close a connection in the middle of a mime multipart request, before the boundary line; users sometimes hit this and can't easily change their corporate proxy. hat tip to Henrik for the clear description.
This builds on a patch attached to that bug to handle this more cleanly, and it adds a test that does trigger the behaviour.
The test is a bit copy-pasty, but it does cover the case and it's not awful.
To post a comment you must log in.
Revision history for this message
| Martin Pool (mbp) wrote | # |
Revision history for this message
| Martin Pool (mbp) wrote | # |
Preview Diff
[H/L] Next/Prev Comment, [J/K] Next/Prev File, [N/P] Next/Prev Hunk
It might be nice to note that retries will happen in the log, in case folk are confused/concerned by bzr being unexpectedly slow.