The phrase of the day (of yesterday, actually) is "weakly strongly plurisubharmonic".
Today I installed Opera 7.11 for Linux, so as to try it. I downloaded the tarball that has no OS version attached to it, with static QT, since my OS (RedHat 6.2) is older than any of the options they listed on the download page. Untarred the browser, ran the install script, installed inside my homedir. Ran the browser in the foreground; got back a prompt. Not useful.
Undeterred, I downloaded the RPM; again the one not listed as being built against a particular OS, with static QT. Installed it with no problems (which was nice). Ran it as myself: "Segmentation fault". Ran it as root, on a lark: browser starts. Ran it as myself again: "Segmentation fault".
No Opera 7.11 for me, I guess, since I have no plans to spend time attempting to get this to work.
Reading the Slashdot comments on Mozilla Firebird 0.6 and the Mozillazine comments on the same, I am struck by the number of conversations which look like this:
Person trying Mozilla Firebird 0.6: "This, this, and this seem to be problem areas that could use some work."
Mozilla Firebird 0.6 user: "If you don't like it, don't use it! We don't need your stinkin' kind around here!" (I wish I were making this up, but this is nearly word-for-word what I've seen people post in reply to a number of comments.)
I asked one of these Mozilla Firebird users who this mythical "we" was that did not need to hear criticism, since he did not seem to be involved in the project in any visible way (as measured by CVS logs, Bugzilla activity, etc). His response, removing the profanity, personal attacks, and attempts to insult me, was that he was involved in "helping newbies install it" and "spreading the word".
I love his methods of spreading the word, and I'm sure that all the Mozilla Firebird front-end developers and all Mozilla back end developers are simply overjoyed to have such an effective group of advocates speaking for them in the first person plural.
Johnny changed things so that the document's onload event now waits for all the image onload events to fire before it fires. The event was already waiting for the images to load; just not for their onload events to fire. This added 30ms (about 3%) to the Tp time on btek, because the image onload events are fired asynchronously (which has to be done because libpr0n, unlike necko, will respond with data synchronously) and the Tp time is measured as the time from load start to load event firing. So the question is, how much of that total 1020ms time is spent waiting on things like this after the page is actually loaded? How much time does it take for the page to look loaded to the user (note that this could well be over 1020ms depending on when reflow happens in relation to onload events).
It's too bad that as soon as people come up with a measurable substitute for whatever it is they care about they start treating it as more important than the real thing. I'm reminded of IQ tests and SATs.
When I first typed TeX in my previous comment, I entered it as
"TeX". Then I thought to myself, "This is the Web, not a
plaintext document. Surely I can use some CSS to mark this up so that it will
look right!"
And indeed, simply doing
T<span style="vertical-align: -0.5ex;
line-height: 0;
text-transform: uppercase">e</span>X
did sort of what I wanted. Here the -0.5ex was a guess based
on what I knew the result should look like, the line-height was
needed to not cause an unsightly line-spacing increase (CSS3 has some proposed
properties for handling this globally), and the text-transform was
used to allow non-CSS clients to degrade gracefully.
There was just one problem. The result looked ugly as sin. Still does,
really. So I took a peek at what \TeX is actually defined as.
And that is:
T\kern-.2em\lower.5ex\hbox{E}X
So I was right about the vertical-align; what I was missing was the kerning. It's still missing, as you can tell, because I cannot think of a good way to do it in CSS (relative positioning is no good, since that will make the space after the "X" too big). Oh, well. Chalk up a point for 20-year-old technology.
On a related note, this is actually a case when use of the "style" attribute seems to be in order. I suppose I could set a class="e-in-TeX" on those spans and move the style into the site stylesheet, but that seems pretty silly too (though now that I have that markup in two or three places on this page that may indeed make sense). What I would like is a way to say, "Put TeX logo here," without having to repeat the icky markup for it every time. Chalk up a second point for 20-year-old technology, I guess.
Today I decided to finally bite the bullet and set up X to let me enter Russian (mostly for e-mail purposes). Surprisingly enough, this was not as hard as I had expected—the HOWTOs and other documentation have gotten a lot better since the last time I considered this endeavor (3 years ago or so).
The time breakdown for this was as follows:
Somehow, the whole procedure recalled freshman year in college in my mind. Might have had something to do with the amount of time I spent back then configuring things like mail readers, window managers, X, emacs...
Next project: find time for my annual search for a tool that converts TeX to MathML. The gotchas here are that it needs to deal
with the various AMS packages, with things defined via
\newtheorem, and with things defined via \def and
\newcommand (though I'll settle for just one or the other if both
are not supported). Oh, and dealing with \newenvironment would be
a nice bonus. (Unlike some, I do not indulge in \catcode or
\let much, so it's ok if those are not handled very well...) If
you know of such a beast, please let me know.
For a while now, there's been this tent in the middle of the quad here with a bunch of bricks spelling out "No Occupation." Seems that the "moral" thing for the U.S. to do would be to immediately pull all troops out of Iraq and let the infrastructure, civil service, police force, etc. just somehow rebuild themselves. I bet the looters would be ecstatic. As would the dictator who would manage to gain control of enough support to put himself in power.
Some people's inability to think through the implications of their pat slogans is astounding. Especially when you consider that this is one of the premier institutions of higher education in the United States, with a strong humanities core curriculum that includes numerous history classes. You'd think people would have learned something in those classes....
I now filter mail to my MIT address into three folders:
Folder #3 is getting perhaps a dozen mails a day. I don't bother to read the other two much; every couple of days I skim subjects (and senders for the spam folder) and delete all those mails.
Life is much improved.
Looks like there is this nice url-uncompress function over in
url.el. Just changing xml-rpc-request-process-buffer
to run that before doing anything else makes life happy.