https://bugs.winehq.org/show_bug.cgi?id=39380
Bug ID: 39380 Summary: test.winehq.org: Provide a way to know why a report was rejected Product: WineHQ.org Version: unspecified Hardware: x86 OS: Linux Status: NEW Severity: normal Priority: P2 Component: www-unknown Assignee: wine-bugs@winehq.org Reporter: fgouget@codeweavers.com Distribution: ---
When WineTest submits a report that report can get thrown away. Either because it's too big, because it has too many failures, because too many tests were skipped or because of other reasons unknown.
The WineTest.exe is never notified of this because the report analysis is run asynchronously with respect tot he submission process. But annoyingly there is no way for the submitter to know why the report was rejected even later, levaing him to guess as to what needs fixing.
One way to plug this hole would be to send an email with the reason to the address field in the report. But that could probably be abused to send emails to arbitrary persons.
A very simple alternative would be to create a per-commit text file containing one line per ignored report with the report's tag, timestamp of when it was analyzed and the reason why it was ignored.
https://bugs.winehq.org/show_bug.cgi?id=39380
--- Comment #1 from François Gouget fgouget@codeweavers.com --- The errors can now be seen at the URL below.
https://test.winehq.org/data/errors.html
However: * There is no link to that page. * Due to bugs in dissect, a lot of reports can still end up being lost.
https://bugs.winehq.org/show_bug.cgi?id=39380
François Gouget fgouget@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |RESOLVED Resolution|--- |FIXED
--- Comment #2 from François Gouget fgouget@codeweavers.com --- As far as I can tell there is no bug in dissect that could cause reports to be lost: every error case goes through mydie() which stores the error for use to build the errors page.
There are a bunch of (soon to be fixed) cases where dissect ends up manipulating undefined values and those may cause the report to be put in the 'Other' column. But it's not lost or rejected in that case.
Furthermore the main index page now has a link to the rejected reports page.
So this is all fixed now.
https://bugs.winehq.org/show_bug.cgi?id=39380
François Gouget fgouget@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Assignee|wine-bugs@winehq.org |fgouget@codeweavers.com Status|RESOLVED |REOPENED Resolution|FIXED |---
--- Comment #3 from François Gouget fgouget@codeweavers.com --- I figured out what happens.
When dissect rejects a report no build directory is marked as outdated. As a result gather has nothing to do and thus the main index files don't need an updated. So the cron script skips re-running build-index. However build-index is also responsible for updating errors.html.
So when a report is rejected one will not know why until another report is accepted.
https://bugs.winehq.org/show_bug.cgi?id=39380
François Gouget fgouget@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|REOPENED |RESOLVED Resolution|--- |FIXED
--- Comment #4 from François Gouget fgouget@codeweavers.com --- This is fixed:
commit f7b0a4cc7e3a7f48a4f987ef4907ac753f5ff692 Author: Francois Gouget fgouget@codeweavers.com AuthorDate: Wed May 17 11:27:08 2017 +0200 Commit: Alexandre Julliard julliard@winehq.org CommitDate: Wed May 17 18:41:25 2017 +0200
winetest: Run build-index and/or build-errors as needed.
If a report has been accepted then run both gather and build-index to refresh the index files. If a report has been rejected run build-errors to update errors.html. And if there is nothing new both can be skipped.
Signed-off-by: Francois Gouget fgouget@codeweavers.com Signed-off-by: Alexandre Julliard julliard@winehq.org
https://bugs.winehq.org/show_bug.cgi?id=39380
Rosanne DiMesio dimesio@earthlink.net changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #5 from Rosanne DiMesio dimesio@earthlink.net --- Closing fixed website bug.