I honestly don’t know why I even load Slashdot anymore.
I tried to make the change to kuro5hin when I tired (some years ago, now) of Slashdork’s continued duplicates and awful spelling and grammar—talk about needing an editor—but it didn’t take.
What do the cool kids read for their tech news fix these days?
(Do not say “Digg.” Digg is on my s#!+ list for being the main reason I personally have had to do a bunch of extra [release] work due to people “digging” “news” that software has been released when it hasn’t been… which brings into question the validity of anything that’s “dugged.” Also, there’s no [easy] past tense of “digg.”)

Tidbit the First:

The current Build Team Radars have been posted on wiki.m.o.
I did say radars—multiple—as we’ve published August’s monthly radar, but also a longer term, quarterly radar.
Astute readers of the two scopes may notice that the monthly radars now have ETA dates, whereas they didn’t before. You may also note that the quarterly radar has no specific dates, but has a ninth month outlook.
Being the aviation geek that I am, don’t be surprised if your hear me start referring to these as the “[Monthly] Approach” and the “[Quarterly] Center” radars¹

rhelmer has been spending his days working on the Test-Only Tinderboxen. For those in a hurry, this was adding functionality to tinderbox to download a build and test it on a separate machine. These machines would have different characteristics than the build machines (namely, they wouldn’t be VMs), and hopefully will give us more stable data.
Well, they’re finally here! Look for columns ending in “test perf” on Linux and Windows on the 1.8 branch and trunk Tinderbox pages. And this is only phase one.
Thanks, rhelmer, for your hard work on this!

We made this change some time ago, but it may still be confusing: nightly builds older than about about 60 days now get moved to archive.m.o, and moved off of ftp.m.o.
If you’re looking for older nightlies than you can find on ftp.m.o, check out archive.

____________________
¹ Approach radar has different requirements than Center radar: in Approach’s airspace, the planes are moving slower and the precision requirements for the radar data is higher (due to higher congestion in the airspace). So, less airspace to cover, but more detailed information necessary. Center radar, on the other hand, covers a much larger area, and the planes are moving in multiples of Mach, so there’s a larger space, but less precise positioning data. An apt analogy for our radars, I think.

While doing some testing today, we found a serious bug that will probably have to be addressed before Beta 2 ships.
It has to do with the migration code just barfing on large amounts of input.
As Rob Strong notes, it’s currently unclear whether or not it’s a blocker, but I wouldn’t be surprised if it turned out to be. The attachments have some pretty disconcerting screenshots in them, as does the test URL noted in the bug.
I know sspitzer is working on investigating it more; expect a complete writeup soonishly.

I’ve starting taking some time to go through the open build@mozilla-org.bugs list. The goal is to make the list reflect reality because, as most of you know, a list with over 100 bugs on it, all assigned to the same “person” is.. well… basically meaningless.
There’s more work to do, but I got the list down below 100 bugs, down from 150ish or so.
Anyway, by far the most exciting part of the session was dealing with my lowest Bugzilla bug number evar, bug 11127: “Build issues for external developers.”
Opened October 2nd of 1999, it reported that CVS is too slow (and we should provide rsync feeds to the source), NSPR debugging variables didn’t make sense, nmake was too slow, and “the build system is insufficiently documented.”
Given that CVS is now wicked fast, we don’t use nmake for anything anymore, and… well… something never change I guess, I RESOLVED it FIXED.
(Seriously though, I mentioned Devmo in the resolving comment; there’s tons of great stuff on there and wiki.m.o about the build system and its architecture; there’s been a ton of useful work in that area as well.)
I’m going to continue working through the list until it reflects reality, and I’ve already RESOLVED -> WONTFIXed some bugs.
If I happen to do this to one of your “favorite,” pet Build Config bugs, please gently re-open to let me know the error of my ways.¹
_______________
¹ All bugs before 2001 will not be considered for this special offer. They’re old and therefore obviously unnecessary.^2
2 Just kidding.

You heard it here first: the reason net-neutrality is a bad idea, and will never work… is because—are you listening, now?—”[The Internet is] a series of tubes!”

That’s right folks. Tubes. “It’s not a big truck.”

I’m so comforted by the fact that we have informed people like Ted Stevens making policy decisions.