March 25th, 2004
4:54pm: Say you love CVS. SAY IT.
<zab> "cvs rtag -d $BRANCH $REPO" didn't seem to really remove
  the tag..
<shaver>         -B      Allows -F and -d to disturb branch tags.
  Use with extreme care.
<zab> ah
<shaver> nice that it gave you an error message, eh?
<shaver> *shot*
<zab> god
<zab> these tools
<shaver> say its name, zab
<zab> THANK YOU SIR, MAY I HAVE ANOTHER?
March 23rd, 2004
12:01pm: money money money

Mark Shuttleworth is offering some Bounties, including some on Mozilla-related projects. Get hacking!

March 22nd, 2004
11:53am: get your flame on

My friends and I have strong opinions when it comes to CVS:

<phik> we only have a few really core requirements that cvs doesn't give us
<phik> atomic commits with unique names are nice (hey, grab patch-315)
<blizzard> phik: yay, atomic commits!
<blizzard> phik: i.e. patch sets
<phik> yeah
<blizzard> phik: I guess I've had to live for so long without them
    I don't miss them
<blizzard> phik: which is more or less like saying "I've been shoveling
    pig shit for so long I don't even smell it anymore"
<shaver_SFO> it's very much like that
<shaver_SFO> if the pig shit were on fire
<shaver_SFO> and screaming curses at you
<zab> and kicking you in the nuts
<shaver_SFO> zab: with cloven hooves

March 21st, 2004
11:59pm: conversational english

Had an instrument lesson on Saturday. Quick cross-country up to Laconia in New Hampshire. Didn't learn too much, but it was good to get some practice in. Came back and did the VOR 23 approach into Bedford partial panel which was kind of fun. Ended up a bit high, but not terribly so. It was a good confidence building lesson.

March 17th, 2004
3:43pm: ...or how SARS will save Social Security

The Bruce Sterling Rant-A-Thon.

Oh, when I was reading that I had to go and look up the word Lysenkoism. Strange that I was reading something else earlier that led me to a description of the phrase Sound Science which I've heard used on various talking heads news shows (I think from looking at this definition of doublespeak, but I'm not sure.) Today seems to be the day of disinformation and the dialectic. Very odd.

Start your rhetorical engines!

March 11th, 2004
11:59pm: you'll definitely be hearing plenty more music

I've been trying to track down a copy of an album that is apparently very rare. It's called Scenes from a Blue Divide (Live) and More by Richard Shindell. However, not being a guy who looks for rare CDs nor really looks too far past the rock-n-roll section of your local mega-cd-stores I have little idea where to start.

Also, why do I not own everything ever made my Sarah Harmer?

March 10th, 2004
3:51pm: native menus

Some of the hackers on the moz project have been doing an excellent job of trying to make firefox integrate well with a gnome desktop. Finally today FireFox's menus took a big step forward and they look - at least as I can tell - native. I don't think a casual observer would notice:

Problems in the menus at this point mostly relate to how they are organized as well as using the icons from the theme instead of the application-specific themes. There are also some subtle grab issues as well which might be harder to fix, but most people won't notice them[*].

*: If you have a menu open and you mouse around a page you will notice that hrefs will still highlight. This is because there's no native window associated with the menu bar and we're required to grab on something which means grabbing the entire toplevel window associated with the window. Because you are still delivering events to the toplevel window, mouseovers still work on it. It's messy and requires pretty low level layout changes that I've been unwilling to tackle.

March 8th, 2004
11:37am: "also, on the who-is-on-crack-today-front..."

Check it out. Sun is rewriting the same three apps all over again. Oh, but this time it's different because it's written in Java. The site says you will be able to run it from your favorite portal, too. I'm getting horny just thinking of it.

Seriously, though. If you wanted a cross-platform groupware solution that included things like calendar, email and instant messenger why not use one of the many Mozilla products? We've already got a great mail client, the start of a good calendar, we've had an instant messaging client in the past. We're years ahead in this game. Why rewrite the world - again? Sounds like another distraction for Sun on the road to bankrupcy. If I were a shareholder, I would be pissed that they were wasting money on that.

March 7th, 2004
11:12am: what is your super-power?

Berkshire Hathaway is a huge company. It also has one of the most bland web sites I've ever seen in my life. However, there is good content there. Of particular interest to me was the Owner's Manual. It's worth reading. I would also like to buy a share of stock in that company because they actually have a reasonable P/E ratio. However, I would probably have to sell my house to do so.

When I was in California, Mitchell turned me onto Space Ghost: Coast to Coast. Netflix delivered unto me a DVD today and I can't stop laughing. And I'm sure I'm annoying jacob on IRC, which is only a side benefit. For example:

<blizzard> I shall name you BANJO
It just goes on from there.

March 6th, 2004
9:58pm: eight for three

I had scheduled an instrument lesson yesterday because we were supposed to have good instument weather today. It's early march and we were supposed to have temperatures in the low 60s with clouds and light rain. Oh. And wind. Lots of wind.

By the time that my lesson came around (about 4pm or so) the clouds had moved out of our local area and there was a good bit of blue sky. I could still see well developed clouds to the south east, so I figured we could schedule a flight down to Martha's Vinyard and get into some pretty good clouds. I was not disappointed.

Checking the weather, reported clouds at the vinyard were 100 feet (!), good visibility and light rain. It sounded like a perfect first ILS approach in real instrument conditions. There was also a SIGMET for severe and moderate turbulence in the area. Given the winds outside, I could believe it. My instructor said that when he was with the student before me that they had been bumped around pretty good. He had kept his seatbelts tight for a reason.

Anyway, I filed, preflighted, picked up my clearance and we took off for the vinyard. It was pretty bumpy, but not terrible. It was just good skill-building weather under the hood. Once we got above 4000 or so the air smoothed right out, too, which was nice. About 30 miles or so from the vinyard we started to encounter the clouds which meant that I got to take my hood off and really fly in the clouds for the very first time. It's amazing how your training just kicks in when you're flying and you can't see the ground. We saw some beautiful virga against the nearly-full moon and flew amongst the cumulus clouds while were were vectored for the ILS 24 approach into the vinyard.

On the approach, things started to get pretty bumpy and I got my first taste of what it's like to try to fly an approach in moderate turbulence while trying to look at the instruments. There was this picture in one of my text books of an instrument panel that was blurry from turbulence and I can only say that that is what it really looks like. I was kind of surprised. But I managed to keep the plane right on the glide slope most of the way down and right at the minimums the approach lights appeared out of the gray. It was just amazing to see the airport appear out of thin air. (By the way, you do not know the meaning of the word focus until you have attempted something like this. Your world becomes very small and contains only eight little dials.)

We landed and went into the FBO for a bit and talked to the bored people in the office who were watching some really bad movie starring the Governor of California. Once I had peed and most of the high had worn off we went back out, picked up our clearance and took off for bedford again. It was well after nightfall at this point and the air was really starting to smooth out. I got to spend a little more time in the clouds before we got into clear air and I had to go back under the hood.

Smooth uneventful ride back to Bedford, except that it went really slow. Indicated airpseed at 4000 feet: 105kts. Ground speed on the GPS: 62 kts. It was good practice with the GPS which still has a lot of features I'm not completely familiar with.

We flew the GPS runway 29 approach into Bedford and it got really bumpy again under 3000 feet. Once again, good practice flying an approach under really bumpy conditions. Had to fight pretty hard to keep both heading and altitude at least somewhat close to where they were supposed to be. It was windy down at the ground, too. Winds were reported at 10kts. But I know that they were lying. It was much windier than that. Uneventful landing. However, I had this very nice conversation when I got down on the ground:

    <tower> Did you really fly that GPS 29 approach or did you
    cheat?
    <blizzard> No, I really flew that.
    <tower> That was one of the best approaches I've ever
    seen.  Good job.
Kind of nice to end an evening with an uncoerced compliment. I haven't had that much fun during a lesson in a really long time. Probably not since I soloed least year.

March 4th, 2004
4:50pm: reviewing who?

There's been talk recently about the size of people's review queues for Mozilla developers. I have to say, I'm pretty bad. Possibly one of the worst offenders. There's stuff that's been sitting in my queue for months, untouched. People bug me from time to time on IRC or over email and then I usually get to it, but sometimes not. So from now on I'm going to be reviewing at least two patches a day to try to stem the tide.

March 3rd, 2004
9:05pm: serenity now

I've spent most of the last two days of my life trying to get Mozilla 1.4.2 to build on 7.2-era ia64 systems (i.e. Red Hat Advanced Server 2.1) and it's been...challenging. The linker is no match for the huge size of the layout lib. In 1.0.2, which is the last version I built there there were two layout libs: libgklayout and libgkcontent. Those have been since combined into one SUPER-LIBRARY. Seriously. If you ever see an error as opaque as this (and gcc and friends are not lacking in shitty error messages):

../../dist/lib/libgkconbase_s.a(nsPrintEngine.o): In function
`nsCOMPtr<nsIPrintSession>::operator=(nsCOMPtr_helper const &)':
/usr/src/build/367008-ia64/BUILD/moz \
illa/content/base/src/nsPrintEngine.cpp:248:
relocation truncated to fit:
PCREL21B nsCOMPtr_base::assign_from_helper(nsCOMPtr_helper const &,
nsID const &) collect2: ld returned 1 exit status
You apparently need the magical -Wl,--relax flag added to the linker line. This was presented to me in this fashion:
<jakub> blizzard: ia64 solves this in the linker,
  have you tried -Wl,--relax ?
To which I replied, as any sane human would:
<blizzard> jakub: I have no idea what that even does
In my head there was an extra 'fuck' added to that sentence. But that being said, I figured - what they hey! Let's see what the documentation says about that option.
`--relax'
     An option with machine dependent effects.  This option is only
     supported on a few targets.  *Note `ld' and the H8/300: H8/300.
     *Note `ld' and the Intel 960 family: i960.  *Note `ld' and Xtensa
     Processors: Xtensa.

     On some platforms, the `--relax' option performs global
     optimizations that become possible when the linker resolves
     addressing in the program, such as relaxing address modes and
     synthesizing new instructions in the output object file.

     On some platforms these link time global optimizations may make
     symbolic debugging of the resulting executable impossible.  This
     is known to be the case for the Matsushita MN10200 and MN10300
     family of processors.

     On platforms where this is not supported, `--relax' is accepted,
     but ignored.
They should have saved us the bits and just stopped after the first sentence.

March 2nd, 2004
1:09pm: all errata all the time

Mozilla 1.4.2 is tagged and bagged so I've been spinning errata for our various Red Hat releases. Not sexy, but I do still enjoy it. 1.4.2 on 2.1AS should be tons of fun.

4:11pm: mozilla futures

Brendan has his slides up from Mozilla Developer Day, 2004. If you want a peek into what we're looking at for the future, it's a good place to start.

I'm particularly interested in our ongoing attempts to build in making SVG a first class citizen in the layout engine. We spend a lot of time fretting about the availability of various plugins on Linux, SVG being one of them. If we have good SVG in our web browser we don't have to depend on external non-free plugins and they are freed from what Brendan has called the "plugin prision." It's part of the content instead of being kept in a little rectangle on the screen. This means you can treat it just like HTML content, modify it with the DOM and overlay it into your web applications. I think one of the best examples I've heard of this is a real time performance monitor graph that updates client side without using images or any other gross hacks. And, because we have SOAP and XML-RPC built into the browser, it can build that information on the fly without even having to do a page reload of any kind. This could be a very powerful combination.

It's a real shame that we don't have a nice JVM or .NET style backend to manage the code to run these nice new client side apps. Brendan talked a little bit about that as well, especially his interest in supporting python for writing XUL apps, but didn't really think that would be useful for downloadable content. I think that we need something that fits into the downloadable content model, though. Brendan mentioned, and I'm probably paraphrasing badly here, that JavaScript wasn't really designed for large apps and he's right. As soon as you start building larger apps it starts getting a little clumsy. We need something better. He's been working on his JS-in-JS compiler which is a bootstrap for other possible targets, including the yet-unspecified-JS2, but I'm curious to see if we can hook up other engines that exist today.

Of particular interest is also the GNOME integration. For the past few months Brian Ryner has been doing that work and has been quite the star. It's going well, although FireFox has to come a lot farther in looking like more of a "native" application ala pinstripe on OSX. I think that's going to be ongoing work, though, and I have yet to really talk to Ben about the importance of this.

March 1st, 2004
12:30am: please be prepared to show your ID at any time
The sniffling, sneezing, aching, coughing, stuffy head, fever, how the fuck did I end up in Chicago medicine.
11:50am: still liking the monkeys, even after all these years
From: protectingtheinnocent@aol.com
To blizzard@redhat.com
Subject: Your monkey story. Been searching for you for years.

Two years ago, in the middle of sixth period History, someone gave me
a document. This was your monkey story. While my teacher babbled on
about shit that happened in the past, I unleashed your story's awesome
powers, I was nearly blown out of my desk by its' blatant ability to
entertain! I knew this was the dawning of a new era. I showed this
document to all of my friends, they all loved it as well. I had big
dreams for this piece of brilliance. Me and a friend wanted to spam
neighborhoods with it [by making copies and putting them in
mail-boxes], give it as acedemic speaches, put it in our
resumes.... Then, while arguing over something about it (possibly what
to do with it first) we broke into a fit of rage, and wrestled
eachother to the documents doom. In the heat of our nerdy
slap-fighting, we ripped up the paper! Nooo!! Hilarious monkey story,
we hardly new ye.... Now, two years later, the story has been
rediscovered via Google. I finally found your e-mail address!  Please
reply, I need to know you are alive, and not lying in a gutter
somewhere full of bullets and cocaine. The world needs your genius.
For the record, I didn't actually write said story. I think it came off of usenet or something.