The roads I take...

Here's a summary of SeaMonkey/Mozilla-related work I've done in week 39/2008 (September 22 - 28, 2008):

• Automated Tests:
  The long-standing punycode test failure on Mac OS X 10.4 could be worked around for now, followup bug was filed for the failure behind it that should actually be investigated.
• Build System, Organizational Stuff:
  The packages file for Windows is now updated to match reality.
  A number of version number updates were filed for various tools and sites to deal with 2.0a1 and trunk changing to 2.0a2pre.
• SeaMonkey 1.1.x Releases:
  SeaMonkey 1.1.12 could be pushed on Tuesday finally, I'm listening to bugs and posts about JS regressions in Gecko 1.8.1.17, we'll see what the plans for a 1.1.13 can be based on that.
• SeaMonkey 2 Alpha 1:
  I went through the painstaking work of creating Alpha 1 builds late last week, after the last blocker landed and we now have only one, new prefwindow left.
  The candidate builds are up on the FTP servers for smoketesting, we are a bit low on tests right now, but didn't hear of any major problems yet. Once we get good smoketesting coverage on all platforms and have no major issues reported, we'll push the release live - hopefully this week.
  There are no localized builds for this alpha, but we will be providing a few experimental language packs to get a bit more testing of locale switching code and some preliminary testing of a part of the localization content in the hg source L10n repositories.
• German L10n:
  The German L10n could be fixed in time for the cut of 2.0 Alpha 1, so this will be one of the experimental language packs available with the Alpha.
  I found a too long label in add-ons manager during this, I hope we can get that in shape for later releases.
• Theme Work:
  EarlyBlue 2.0 Alpha 1 is now available from AMO in the sandbox (needs login) and I guess it's actually good to have an alpha in sandbox only (for public download, visit my download page). I discovered an issue of check/radio menu items not showing any icon, which I only fixed locally for now, not sure if I'll re-release the theme's alpha for that.
• Various Discussions:
  Toolbar customization, leaks, L10n dashboard, Alpha 1 release notes, Windows icons, installed-chrome.txt removal, etc.

We have builds for Alpha 1 now, and a final release is only waiting on test coverage so that we are confident to ship it to a broader public for even more testing. Of course, this is the first Alpha, so don't expect it to be a smooth ride, though I guess on Vista it might even feel more fitting than 1.1.x already. I'm currently working on release notes and the download page, which will all see modifications in style compared to the 1.x equivalents. We probably go through more iterations of those pages for future preview releases, depending on feedback we get, so we'll arrive at hopefully much improved download and relnote pages for 2.0 final.

I'm really looking forward to feedback to all those new things coming about with Alpha 1 very soon!

Yesterday, I finally could upload SeaMonkey 2.0 Alpha 1 candidate build 1 for (smoke)testing by our community QA people.

For me, this has been a complete new experience. All previous (1.x) SeaMonkey releases have been done on my private tinderbox machines, were cvs- and xpfe-based, had no crash reporting or automatic update support - things that all changed now. 2.0 Alpha 1 was created on Mozilla-hosted build machines, is hg- and toolkit-based, has a working open-source crash reporting system that needs build symbols uploaded, has has support for automated updates to later releases. That's the bright side.

The dark side is that I have a half-automated process for doing 1.x release, but neither Mozilla nor I have any working automated process for hg-based releases yet.
For 1.x, I have a script that gives me list of commands for version updates and tagging that I only need to copy and watch the output for doing stuff properly, and I just could copy the tinderbox configs, modify them to run only once and do a clobber and full rebuild, pull the release tag and upload to the candidate directory (the latter two are the only needed changes for every release), and then run those tinderbox client processes on the three machines (Linux, Mac, Windows). I even stuffed an additional script into the Linux config to create source tarballs and put them in the right place for being uploaded by tinderbox' upload step.
For this release though, I had to do everything - yes everything manually. And I learned painfully what that means (even though I had FF 3.1a2 build notes as a base):

• I pulled comm-central locally into a new directory, performed a client.py checkout and had to manually enter the commands for tagging cvs in mozilla/extensions/typeaheadfind, mozilla/extensions/wallet, mozilla/extensions/irc and mozilla/extensions/venkman and tagging hg repos in mozilla/extensions/inspector and mozilla/ before adding the last patch, update client.py for the tags to pull and tag comm-central itself. I didn't know how to correctly merge this back to the last tip checkin so I reverted and pushed - before I realized I forgot to change the version from 2.0a1pre to 2.0a1 for the release and Callek went in and corrected this and moved the comm-central tags for that (thanks again).
• On all the build slave machines that create SeaMonkey 2 nightlies, I created a builds/release directory and pulled comm-central from the tag there and ran client.py checkout, copied the .mozconfig from the nightlies with replacing the "nightly" update channel to "beta", did set the MOZ_OBJDIR environment var and ran the client.mk build - and repulled and restarted the builds after the tag update for the version messup. I went to sleep , logged in the next day, and realized that Windows hadn't built palmsync, so I tried a number of times until I had figured out how I needed to set the env vars to get a rebuild with that extension.
• I set a bunch of env vars and ran the symbol build and upload steps - only to discover that both Mac and Linux issued a number of strange-looking warning during that, including errors for 6 libraries on Mac. None of the Mozilla build people could really help me with those warnings, so after getting them again in re-runs of those steps, I decided to accept them and go with the results we got out. For the 6 libs on mac, this means that we actually have incomplete symbols for Intel Mac, which is a known failure of symbol dumping on OS X 10.4 - I hope it can be worked out for future releases.
• Following that, I ran the build steps to create packages - only to realize I couldn't run that in either a ssh login or a screen console on Mac but had to run it a third time from a graphical terminal window via vnc so that hdiutil would succeed and we'd get a dmg finalized.
• Fortunately, the .mozconfig contained the configs for update-packaging from the beginning so I could run the buildstep to create MAR files for full updates, which we'll need to create partial update files if we want to offer automated updates to the next release.
• Directories on the FTP staging server had to be created manually and scp run manually on the machines to upload the packages.
• For a source tarball, I pulled the tagged tree again locally, removed all .hg directories (so the packages won't be too oversized), tarred that up and manually uploaded it.

This was about 8 hours of actual work time for me, spread over two days, and with a large number of "WTF?" experiences, making me end up half-ill and very tired. See my 2.0a1 build notes for commands and detailed data.

I hope we will not need another candidate build for this release (even though I learned a number of things I hopefully wouldn't run into a second time), and I hope the automated release process based on hg will be far enough so we can use it for Alpha 2 - but I finally understand the stories about Scotch whiskey being the best friend of release engineers...

Due to server outages at Mozilla, which kicked us off IRC right when the meeting should happen and didn't allow us to see the agenda or bug lists, we're rescheduling our meeting, postponing it for 24 hours to tomorrow, Wednesday 24, same time, same station/channel.

NEW date/time:

* September 24, 12:00 UTC, 8:00am Eastern, 1:00pm UK, 2:00pm Central Europe
* Other timezones: http://www.timeanddate.com/worldclock/fixedtime.html?day=24&month=9&year=2008&hour=12&min=0&sec=0&p1=0

Agenda: https://wiki.mozilla.org/SeaMonkey:StatusMeetings:2008-09-24
Location: IRC channel #seamonkey - irc://irc.mozilla.org/seamonkey

Sorry for the inconveniences the new timing might cause for you, this is a one-time thing for this time, in two weeks we'll be back at our usual Tuesday slot.

Here's a summary of SeaMonkey/Mozilla-related work I've done in week 38/2008 (September 15 - 21, 2008):

• Automated Tests:
  Mnyromyr has created a cool test that checks several windows for duplicated XML element IDs, which even has already caught bugs with our prefwindow migration at checkin, and which has lead to numerous fixes of bugs in several windows before it even was turned on for those. This is great because it should also improve how extensions can hook into our windows and how internal code works because we avoid this class of bugs and really have unique IDs in the checked windows. When he added mailnews windows to this test, he needed to create a dummy mail account as else we would launch the modal account wizard and the windows aren't correctly exposed. We discovered that cleaning up prefs in chrome tests is problematic, so we disabled the cleanup for now.
  Input of chrome test wizards who know a good solution for actually cleaning up here would be very welcome!
• Build Machines:
  I did some work to improve reporting of L10n builders to tinderbox, but it still needs work, unfortunately I don't fully understand the code that does the reporting to the Mozilla-L10n-* tinderbox waterfalls.
  I also made L10n builders not fail uploading when the last locale failed, not do MAR files when L10n repackaging failed, and I tried turning on automatic update snippets for L10n builds, only to realize that this needs more work done.
• Build System & comm-central:
  I landed calendar/ on comm-central so the calendar guys have everything in place to switch development to hg trunk once their 0.9 release is out.
  For hg-based release automation, I made it possible to create release files with "pretty" filenames, which SeaMonkey probably won't use but Firefox does currently and will in the future. As a side-effect, that change made the paths and names of any package we create much more configurable right in the build system.
  As I'm continuously watching the hg changelogs for the Mozilla build system files we've copied as a base for the comm-central build system, I once again created a patch for porting recent changes over to our files and at the same time discovered a first easy part of comm-central build system cleanups.
• SeaMonkey 1.1.x Releases:
  The SeaMonkey 1.1.12 is only waiting for the actual release of the parallel Firefox 2.0.0.17 and 3.0.2 security updates to actually go public, which is currently scheduled for Tuesday, September 23rd, and from what I hear, this date probably will stick this time. I uploaded a number of localized builds to my staging directory to be released right with the original US English version.
• SeaMonkey 2 Alpha 1:
  Lots of triaging and poking people to get the last blockers sorted out and fixed, as well as trying to figure out how to actually create the release builds.
  We are now only waiting on the legacy prefwindow removal to get final reviews and be checked in, and then we're ready to go for tagging and creating builds. Expect those to be up for testing within the next few days.We'll need help smoketesting them on all platforms then before the release can happen.
• German L10n:
  A number of corrections to SeaMonkey trunk L10n could be done thanks to Adrian using de as a testcase for his Silme-based compare-locales work he's doing as a Mozilla intern currently. He discovered a few small glitches in my localization, and I found a few duplicated accesskeys when looking at those places.
• Theme Work:
  As SeaMonkey 2 is nearing its first Alpha release, I wanted to make sure that at least one of my two themes is ready for that, so I spent some time this weekend to update my EarlyBlue theme for the latest changes. It was harder to find out what changes I need to incorporate, as we have no possibility to query for changes in a certain directory in the hg pushlog (I need to look at suite/themes/classic and mozilla/toolkit/themes changes and update my theme for everything affecting it). This was easier with CVS and bonsai, where this was a straight-forward query. I hope we'll get this for hg as well one day, bugs are filed.
  In any case, I have completed my EarlyBlue porting (it's the theme I'm using all the time personally as well) and have put an XPI up on my download page for now, I still need to take the time to get it onto AMO as well.
• Various Discussions:
  Toolbar customization, leaks, L10n dashboard, Alpha 1 release notes, EU MozCamp, Thunderbird message view, unit test failure, etc.

I know I've been preaching that "we'll have an Alpha soon" for a very long time now. The good thing is that the code actually is practically done now. Some remaining reviews for the removal of the the old preferences window, and we're really done. It's been a long time since we released a testing version to a wider audience, the last was SeaMonkey 1.1 Beta on November 8, 2006, almost two years ago. Back then, we introduced the URL bar color change to yellow on secure sites, tab previews and message tagging as some of the largest changes. Now we're featuring a complete rework of many parts of the app due to the move from 1.8 branch xpfe to 1.9.1 trunk toolkit as well as a number of SeaMonkey-specific and mailnews updates, reworks and features. This ranges from a new Windows installer via a modernized icon set for the default theme, clear private data, undo closed tabs, newsgroup drag-and-drop reordering, add-ons manager and completely redone extension management to <audio> and <video> HTML5 tags, much improved Acid3 compliance and less memory usage, just to name a few examples. It's a really huge step forward, and it shows that we haven't been sitting around doing nothing for the last 2 years.
The SeaMonkey project is not dead, as some people might believe with our last new stable release dating back 20 months - on the contrary, SeaMonkey is more alive than it ever was under that name!

A few minutes ago, I landed the calendar code on comm-central for bug 455727.

This means that Lightning and Sunbird are moving to hg trunk development soon and that next to SeaMonkey and Thunderbird, Sunbird is the third application to be built from this community and communication repository.

Let's give the calendar people a warm welcome - we're looking forward to a good time of joint forces here!

Here's a summary of SeaMonkey/Mozilla-related work I've done in week 37/2008 (September 8 - 14, 2008):

• Automated Test Machines:
  I spent quite a bit of time this week tracking down the remaining problem on our automated SeaMonkey test machines. First I made leak thresholds work with a workaround on those test machines, so the boxes can go green even though they report leaks. We should still fix those, but it's more important to see new issues coming up by making the boxes go from green to orange or red.
  With that done, we saw a browser test fail, which actually needed a different test to be fixed to correctly clean up its listeners, but we went green on Linux and Windows except for a few intermittent failues that probably still require closer looks.
  On Mac, we had two unit test failure though, and I went into a few-hour-long session with Boris Zbarsky to debug one of those, only to find out that NSPR reported to us that machine has zero physical memory, even if I could swear it had some. This was worked around for the moment by assuming we have 32MB when NSPR reports 0, so that we create a memory cache with minimum size and get working tests.
  I also tried to get forward the other remaining Mac failure by trying to get us to exclude this test on OS X 10.4 only as we know it fails there, but those things are not easy with our TUnit test suite and xpcshell tests right now.
• Build Machines:
  The generic stage upload buildstep for Mozilla buildbots now can upload L10n update files and language packs now, so we can fully use it for upload L10n builds to Mozilla's FTP server.
  Symbol steps have been put before packaging to make OSX symbols work for SeaMonkey as well - thanks to Ben Hearsum for finding out about this and posting about it on planet (where I found out about that).
• Code Cleanup:
  The last patches for killing xpfe-style chrome versions are landed now, we finally could get rid of that old cruft.
  Additionally, the mork-based toolkit history has been removed now, the only remaining history implementations in the tree is the outdated xpfe mork history still used by SeaMonkey and the current places-based history in toolkit. Switching SeaMonkey to this is a wanted-seamonkey2+ bug but someone needs to look into making the history window and sidebar work with it.
• SeaMonkey 1.1.x Releases:
  The release process for SeaMonkey 1.1.12 is continuing, testing on the candidate builds looks good and those builds will probably be released, but Firefox people saw a few regressions with the synchronized FF 3.0.2 release and needed to push all those synchronized releases further out, which also affects us due to security advisories being published at the same time for all of those. The plan calls for the end of this week or start of next week for the actual release.
• SeaMonkey 2 Alpha 1:
  As the SeaMonkey-specific code froze for Alpha this week, I was busy caring that people get in their patches for blockers, triaging the remaining blocker nominations and approval requests after the freeze.
  In the Status meeting this week, we decided that we'll call the next milestone Alpha 2 for now, so I cared we get bugzilla flags for this and also for SeaMonkey 2 "wanted" bugs. The general feeling is also that we will actually need it the milestone to be an actual Alpha, though we could still go and change it to a Beta if that would feel more fitting - that's a decision to make when we get nearer to it though, for now the nightlies will go 2.0a2pre once Alpha 1 gets cut.
• SeaMonkey L10n Builds:
  Reporting our L10n builds on Mozilla-l10n-* tinderbox waterfall pages shows somewhat strange errors now, I need to further investigate that, but builds themselves are generated correctly.
  And please welcome Hungarian as the newest language being built for SeaMonkey trunk!
• German L10n:
  German SeaMonkey L10nsaw a few more updates for the surge of prefpanels being converted for the SeaMonkey Alpha freeze.
• Various Discussions:
  Leak sources, toolbar customization, L10n dashboard, Alpha 1 release notes, EU MozCamp, www.mozilla.org archiving and revewing, etc.

As I mentioned last week, SeaMonkey-specific code is frozen for Alpha 1 now, and meanwhile most of our remaining blockers have landed, the remaining blocker in SeaMonkey-specific code is removing the legacy prefwindow now that all panels have been migrated to the new window - huge thanks for everyone involved for completing this huge task!
Unfortunately, we have two core regressions we're also tracking as blocking our Alpha 1 release, one regrading Ctrl+Home/End keys in editor, another being wrong selection colors in plain text, I hope we'll see fixes for those landing soon, so we can cut and ship this long-awaited Alpha to more public testing.

Here's a summary of SeaMonkey/Mozilla-related work I've done in week 36/2008 (September 1 - 7, 2008):

• Build System Work:
  make install is now fixed for comm-central.
  SeaMonkey-specific security pref files moved to comm-central now, just before they were migrated to the new SeaMonkey prefwindow, and winhooks prefs could be killed now that shellservice obsoleted that.
• Build Machines:
  When the new shellservice UI landed (yay, we can set default application on Vista correctly now!), the default testing profiles needed a fix so mochitests don't pop up that default application window and fail.
• Code Cleanup:
  Landed a number of patches for killing xpfe-style chrome versions, the main patch can only be landed last, I'm waiting for DOMi review to complete this.
• SeaMonkey Releases:
  The release process for SeaMonkey 1.1.12 is continuing, the candidate builds look good in terms of testing on all platforms so far, schedule for September 16 still stands.
• SeaMonkey L10n Builds:
  This week, I did even more work for localized builds for SeaMonkey from our buildbots: language packs are added now, as well as the complete MARs for automated updates, but the update server still has no information about them, so you won't get them offered right now.
  Additionally, the localized nightly builds are in latest-comm-central-l10n now for consistency with where Firefox and Thunderbird are staging them.
• German L10n:
  Some updates to German SeaMonkey L10n, and some dummy/small L10n changes to test new L10n build stuff mentioned above
• Various Discussions:
  Nearing Alpha 1, compare-locales, drag and drop, Bugzilla update, etc.

We are freezing all SeaMonkey-specific code today, September 9, at midnight PST for SeaMonkey 2 Alpha 1. There is a small number of open blockers but most of those are near completion. Getting all the preference panels ported to the new prefwindow was a huge effort, and it did really go fast and well nearing this deadline, huge thanks to everyone involved in this effort!
Starting with this freeze deadline, only patches with approval-seamonkey2.0a1+ or fixes for blocking-seamonkey2.0a1+ bugs are allowed to be pushed into SeaMonkey-specific code. We can't freeze MailNews or the platform but we hope we can pick a relatively stable point of those to cut the alpha once all blockers are fixed - I hope this can be in about a week. We're really knocking on the Alpha's door now.

That said, I'll probably not be online much at the end of this week, celebrating my 30th birthday on Saturday. But no fear, I was told there's life beyond 30, so I'll probably be back and going strong after this weekend!
I'll keep on rollin' with the flow...

Wake up, Alpha.

The SeaMonkey has you...

Follow the frozen tree.

Knock, knock, Alpha.



Whoever can read this matrix code should realize by now:
Yes, we're knocking at the door of our first Alpha. It's been dozing along in our repositories long enough - the real story of its life is about to begin. The code is supposed to freeze for SeaMonkey 2 Alpha 1 on Tuesday - after that line we will only accept changes to SeaMonkey code that are either blocking the Alpha (pref panels!) or have explicit approval.

We will release the actual Alpha 1 as soon as all the blockers are fixed and we get a QA run to confirm it's ready for testing by a greater community than nightly testers. It will still be a testing-only preview of what SeaMonkey 2 has to offer, but it will be a very huge step compared to the current stable 1.1.x series.

Additionally, I think we should set up a collection of SeaMonkey-themed desktop wallpapers and SeaMonkey-themed web button images somewhere, I'd need good ideas of where to do that.



For now, I can only provide low-quality images of those two I have designed for my personal fun, but if we have a good place to put up such thing, I have the first one as a 1024x768 PNG, the second one in sizes up to 1600x1200 as PNGs.

When I opened the source to my Mandelbrot fun project, I did this because I hoped people would look into what TraceMonkey could optimize better there, and because I hoped an open example application helps others base their apps on XULRunner.

Now the second part is a long-term goal that I don't expect will be something that happens soon, but the first point has already been followed: As Boris points out in his comment, he filed two bugs on TraceMonkey, being able to use my code or variations of it as testcases, and there are good chances TraceMonkey will be improved to speed up those cases as well.

Additionally though, Boris pointed out possible ways to improve my code in an earlier post after reading it, and a guy calling himself "prefiks" implemented those ways and a bit more in a patch linked in another comment. When I applied that patch, the numeric algorithm approach with TraceMonkey enabled was sped up by a factor 5 and now I get the base image painted in 1.2 seconds on my machine - and even see updates of the image during drawing, which is what I wanted all along! I also tested it with the XULRunner of Fennec M7 on my N810, and even though it takes a bit over 90 seconds to paint this image, one can watch it build up and so it even feels usable there!

So, opening the source should help TraceMonkey being improved and even helped me making my fun side project more usable. This is the power of keeping the source open, and I hope an encouragement for other people to do the same.

Here's a summary of SeaMonkey/Mozilla-related work I've done in week 35/2008 (August 25 - 31, 2008):

• Build System Work:
  I ported a few build system fixes to comm-central and made relative l10n-base as well as the make install target work again on comm-central. The base for building the Qt port is also laid, and autoconf 2.13 detection should be improved, no need to declare the AUTOCONF environment variable with the usual binary names any more, not even on Debian or ubuntu.
  I also filed patches to move SeaMonkey-specific security files to comm-central.
• Code Cleanup:
  Cleaning up xpfe/ has been completed as far as it can for now, and I also could kill the old chrome registry. Additionally, I filed a patch for eliminating toolkit's mork-based history implementation and a few ones for killing MOZILLA_LOCALE_VERSION.
• SeaMonkey Releases:
  I started the release process for SeaMonkey 1.1.12, candidate builds are up, please help testing them so we can catch possible problems before the release, currently scheduled for September 16.
• SeaMonkey L10n Builds:
  I worked a long time this week on getting localized builds for SeaMonkey up on our buildbots, and I finally got them to work, see my SeaMonkey trunk L10n builds post in the L10n newsgroup, the localized nightly builds are up for download in the latest-comm-central directory - be aware that download and password managers are still broken, please read the linked post for more details.
• General L10n Topics:
  As the current MLP pages are badly outdated, I worked on a replacement page that points to current documents instead, now that we have the www archive which is pointed out in 404 error pages on the live site and will be linked for historical content.
• German L10n:
  I updated the German SeaMonkey L10n and German DOMi. Additionally, I checked in a patch for updating de help provided by Michael Opitz, thanks for this!
• Various Discussions:
  client.py improvements, compare-locales, drag and drop, SeaMonkey Status Meeting, etc.

We're about a week from freezing for SeaMonkey 2 Alpha 1 as well as from the Gecko/Platform 1.9.1 feature freeze, so if you're either working on or watching for some feature that needs to go into the platform for SeaMonkey 2 or are able to work on something that blocks Alpha 1, including any preference panel migration, please try to get the work done until next Tuesday!
Also, if you're spotting things in testing that we cannot ship an Alpha with, please nominate those bugs as blocking-seamonkey2.0a1.
We are nearing this first Alpha release for SeaMonkey 2, let's make it a good one!