Monthly Archives: November 2015

My UNCONFIRMED Bug Reports

Zur Mozilla-Bugzilla HomepageUntil now I have been used to report new bugs with status UNCONFIRMED, even if I was absolutely sure that the report was concerning a real bug and complete enough so that a developer can take on. Reason is that I believe in the two-man rule. Often a review by an other user adds information, may be concerning additionally affected Operating system or an aspect of the problem not yet mentioned or something else. But in the SeaMonkey project, due to very limit “human QA resources” and the lack of interest of most SeaMonkey users to contribute to the project, the bug review process  horrible slow, even Bug reports what can be reproduced easily do not find a user who is willing to do a test whether he can reproduce the problem.

So in future I myself will change bug status to NEW after 1 week of waiting if I think that all requirements for starting bug fixing process are fulfilled (and nobody confirmed the bug).

Status update

Long story short, SeaMonkey’s upload and download hosts have changed and with that, some of the configuration parameters needed changing in the backend.  No problem.  Can’t be that hard.  Changed the configs… boom.  Green.

If only things were so simple in life.

MC on #seamonkey pointed out that there hadn’t been any uploads to the server since <some date>.    Looking carefully at the download site, ‘lo and behold,  none of the uploads to the servers are actually getting there.

But the tree was green!  Even the step on the buildbot master said green.  The log within that step said “upload complete”.  How can that be lying?

It wasn’t lying and the uploads didn’t make it.  (You have to be here to see my facial expression… complete confusion.)

Then I realized that “it wasn’t lying and the uploads didn’t make it” isn’t a contradiction.  It’s contrary to what I believe, but what I think happened is that the ‘receiver’ of the uploads discarded them quietly.

So  all those green results are lies.

Self-deprecating humor aside, I think I’ve fixed the Linux and OSX64 side.  Windows, on the other hand, is problematic.  Apparently the system I’m using has no access to the new upload host.  Soooo.. that will need to be fixed.

So all in all, been busy.

Lesson learnt?  Green really isn’t green until one confirms that it’s green; otherwise, it could be fuschia, red or purple.

:ewong

QA: New Whiteboard tag “[easyconfirm]”

Zur Mozilla-Bugzilla HomepageFor the SeaMonkey project I now use a new Bugzilla Whiteboard tag “[easyconfirm]” for UNCONFIRMED Bugzilla bug reports, which can be easily reproduced also by Bugzilla newbies. So these bug reports are ideal for some first bug confirming attempts and to gain some first experience with bug confirming and QA process on our bug tracking system. Continue reading

SeaMonkey 2.39 additional Release Notes

Additionally to the very outdated “official” Release Notes here you find some more information concerning changes with SeaMonkey 2.39.

New features

Please see Wiki-Page. I do not know which Firefox42 – features made it into  SeaMonkey

A Bugzilla query shows  SeaMonkey 2.39-specific RFE-Fixes – for “Bug 1153577 – Users should be able to hide the menu bar and show it with the ALT key” there might be a problem with Ubuntu/Unity.

Bugfixes

  • Bugs in SeaMonkey 2.38 what have been fixed for SeaMonkey 2.39.
  • Bugs which have been fixed in  Thunderbird and Firefox since SeaMonkey 2.38. Not all bugs in Firefox or Thunderbird do affect SeaMonkey, but also it is not sure that all fixes will work in SeaMonkey.

New bugs

The “Known -Issues“- list is outdated, here an attempt for a more realistic overview.

Possible new bugs (“Regressions“)

Old and new bugs

A horribly long list, what does not consider that not all Thunderbird– and Firefox-bugs are relevant for SeaMonkey. And many of  them are UNCONFIRMED or outdated (have been fixed in between).

Download-SeiteThanks to all who made possible this release.

Rainer

SeaMonkey 2.39 is out

Compared to 2.38,  this was a bit of a struggle to get 2.39 done.  Made some mistakes along the way; but, with a sigh of relief, SeaMonkey 2.39 is out.

Now, on to getting the infra back on track… a very slow and albeit, overwhelming at times, process.

I’m sure everyone has had the feeling.  Kind of like sitting at a desk, and having a whole pile of paperwork/homework to do and not knowing where to start since they seem to have their hands up saying “I’m important! You need to finish me first!”  Know that feeling?

And it doesn’t help it that Trunk is now busted.

:ewong

SeaMonkey 2.39b1 information

Here some additional information concerning the new beta release (official release notes here unfortunately are very outdated) .

There are some possible  Regressions compared to 2.38 (bugs, what appeared after release of  SeaMonkey 2.38).  It would be great if some users would do bug triaging  with the UNCONFIRMED ones  from the list. (1)

🙂 But, of course, we do not only have possible regresses, but also real progress, several   Bug-Fixes.


(1) A Bugzilla-Account is required to proceed these queries. Please consider that SeaMonkey might be affected by bugs and bug-fixes for  Firefox, Thunderbird, Core and others, none of the related bugs will appear in resulting bug lists.

SeaMonkey 2.39b1 is out!

After fixing some problems stemming from infra-code changes from 2.38’s release and some hiccups in the network, we’ve finally released 2.39b1.

Here’s hoping the next release would be less painful. 😛

:ewong

SeaMonkey 2.39b1 delayed…

… until we find out what’s troubling the final step.

:ewong