Muli Ben-Yehuda's journal

… is that you get to view such lovely code snippets:

traceLogHandle = CreateFile(TRACELOG_FILE_NAME, FILE_APPEND_DATA,
FILE_SHARE_READ, NULL, OPEN_ALWAYS, FILE_ATTRIBUTE_NORMAL, NULL);

My god, the ugliness.

A new book was waiting for me at the library at work today. Normally, this would be a very happy event (I *love* new books), but right now, I’m pretty depressed. Why? Because the book is the third edition of Hennessy and Patterson’s Computer Architecture: A Quantitative Approach, and it is over a thousand pages of goodness-filled knowledge, and I HAVE NO TIME TO READ IT.

I wonder how far I’ll get with 50 pages every weekend for the next few months.

(For those of you keeping score, I am also reading Feynman’s Lecture Notes on Physics, and will be reading selected chapters of the second chapter of Introduction to Algorithms (Cormen et al) as soon as my copy arrives at the library at work. A new semester will be starting soon, too, with the associated work load, and there’s always Work).

Writing a long entry whining about lack of time for reading books has a certain delicious irony, too. I think I’ll go read my book now.

It all started a couple of days ago. It was afternoon, I was working at the office and planning on a quick dinner somewhere before heading back to the office (don’t ask). One of my pleasures when eating alone is reading a good paper or book. I searched the web for interesting papers and happened upon this wiki, with its excellent Operating Systems and Lisp sections. Browsing the Lisp section, I came upon a link to Richard P. Gabriel‘s momentous essay on the Evolution of Lisp. I printed it out and started reading. I read it during dinner. I came back to the office and read it. I went home and continued reading it. While reading, I chased people and places and projects it mentioned on the web. I stayed at home yesterday, and alternately worked and kept reading it. It opened up many trails, and I wandered down them. A whole world I only had a dim inkling of before, the world of lisp, opened before my eyes.

Sometime around noon, I was filled with a strong urge to write some lisp of my own. I started with a small interpreter, but then decided to write something useful instead. As it happens, I’ve been using a hacked up Emacs TODO mode for a few months as a short-term memory replacement. It works very well, but has a few annoying shortcoming: it didn’t keep the “folding” state when exiting and reentering the buffer (the folding state is which items are hidden and which are shown, a-la outline mode), and it didn’t keep my position in the buffer when closing and reopening it. I set out to write some lisp to fix them, and within a few hours, did. It was a pleasure, and now I want more!

You can find the hacked todoo.el on my code page.

Had my first *real* workout this morning in a long time, where *real* is defined as “Muli is about to faint if he doesn’t stop exercising Right Now”. Had a couple of good workouts last week with ladypine as well, but for some reason, I haven’t pushed myself to the brink when we were working out together.

I have too many things to do at the office today, followed by a talk to the Tel Aviv Linux Club on OLS 2004 in the evening, and then going out with friends. Good morning, World!

The excellent lwn.net has an outstanding piece in this week’s Weekly Edition on why proprietary kernel modules are bad for you. Well done, sirs! (subscriber only for the next week, but well worth the subscription).

It is thus in the interest of all users to discourage proprietary modules. It is not a question of irrational allergies to end-user license agreements or free software fundamentalism; it is, instead, a matter of creating the most stable and capable kernel possible. Had the kernel been a friendlier environment for proprietary code, the kernel we all use now would be less capable, less stable, and less portable than it is. When you see a proprietary module break, or (as in the case of the PWC driver) be withdrawn, what you are seeing is the risk which is inherent in the use of non-free modules, not irrational behavior on the part of the kernel developers.

The reiser4 and pwc threads on lkml are really bringing out the nutters out of the woods. lkml is a zoo.

and since someone asked, here’s my opinion on the pwc business:

Have you seen the threads on lkml? See the links at http://kerneltrap.org/node/view/3729 for how it got to be this way.

Basically, I think greg is 100% right (no hooks that are only useful for binary-only modules), I think Linus is 100% right (if the author wishes to remove the code he wrote, we should respect that), and I think the assorted whiners should realize that the code is GPL’d. You can read it, fork it, and if you rewrite it and offer to maintain it, it will be back in the kernel very quickly. Of course, it’s much easier to whine.

and here’s my opinion on reiser4:

nice toy. Now go play somewhere else.

Yeah, I’m harsh. Got a problem with it? write some code and prove me wrong 😉

In my never ending quest to know everything there is to know about Unix system programming, I learned something new today: man ptmx(4) for the gory details.

I have a new (old) project idea, nick-named RTMOS, or Run-Time Modifiable Operating System. I don’t like the RT “Real Time” connotations. Anyone got an idea for a better name? To put it in one sentence, I’m thinking of doing the emacs equivalent of operating system – infinitely run-time modifiable. I’m writing a rational / research proposal and will post it when it’s ready, but I need a good name now!

People who refer to themselves as hackers, without having done enough IMENHO to merit it, annoy me. I should use procmail to minimize annoyance.

hacker: n.

“It is better to be described as a hacker by others than to describe oneself that way. Hackers consider themselves something of an elite (a meritocracy based on ability), though one to which new members are gladly welcome. There is thus a certain ego satisfaction to be had in identifying yourself as a hacker (but if you claim to be one and are not, you’ll quickly be labeled bogus). See also geek, wannabee.”

Following the MD5 collisions found and the SHA-1 break rumored, orrd, who should know something about these matters, recommends using Tiger. I’m reading the paper now.