Muli Ben-Yehuda's journal

– q_skud_ has an interesting book meme going on.

3 books you use most often for reference: Bovet and Cesati’s Understanding the Linux Kernel, W. R. Stevens’s Advanced Programming in the Unix Environment, and google, because while not a book, I use it for reference far more than I use any book…

3 books you read on “high rotation”: Stephen King’s It, Neal Stephenson’s Cryptonomicon, anything by W. E. B. Griffin (yummy brain candy).

3 books you read for comfort: Stephen King’s It (knowing which character I always identify with shouldn’t be too difficult), Jeffery Archer’s The Prodigal Daughter and James Clavell’s Shogun (I guess I’m weird).

3 books you really ought to read: oh, but there are so many. Let’s see, just looking at my amazon wishlist, anything by Knuth, but especially his Art of Computer Programming (I’ve read bits and pieces, I should read all of it AND do the excersizes), Feynman’s Lectures on Physics, and the complete works of Shakespeare…

3 books you will never read: there are far too many to mention. Not because I don’t want to, but because there are not enough hours in the day.

Yesterday morning I spent a few hours setting up remote connectivity to work from home. Eventually, thanks to Oleg‘s useful tips, a bit of insight and a modicum amount of luck, it worked. This remote connectivity caused an immediate blurring of “work time” and “anything else time”, which I’m intimately familiar with from three years of working from home. Everything has its price, I guess…

Last week passed in a kind of blur. Sunday and Monday were Yom Kippur (aka Day of Atonement), and I don’t think I did anything particularly interesting. Homework, and that’s about it. Then, Tuesday through Thursday were insane days at work, where I was working on two fronts simultaneously. On one front, we have a couple of nasty hardware related bugs that we’ve been busting our brains on, so far, to no avail. On the other front, we worked insane hours to develop an idea into a research proposal, which was finally submitted on Thursday night.

If the research proposal gets accepted, I am likely to be doing very interesting Linux kernel memory management things in the near future. Keep your fingers crossed for me!

I’ve also had several interesting talks with people at the lab about free software, open source and Linux. I also agreed to give a talk on User-Mode-Linux memory management to the Linux Study Group in a week’s time.

In the “code for fun” arena, I did a little work on syscalltrack’s kernel 2.6 support (which will resume as soon as I finish updating this journal). In the “almost as fun as coding”, I’ve been helping people on lkml with their bugs and offering constructive(?) criticism of patches where appropriate.

Next week is the Sukkot holiday, and we only work half days. I hereby set myself a goal – fix at least one real kernel bug, preferably related to shpte. Not very ambitious, I know, but that’s all I’ll have time for.

Yesterday I did Mathematical Logic homework in the morning, and then ladypine and me went grocery shopping. Then I went to the office to check my email and see if there was any progress on the hardware related bugs mentioned above (note to self: get the stupid VPN dialer to work on Linux, not having access to work’s network from home is not acceptable anymore). In the evening we went to an excellent restaurant with Nadav, my former flat mate. Much fun was had. Then we went back home and ladypine helped me fathom the Calculus homework. Eventually, Calculus conquered, we went to sleep.

I seem to have shrugged off the lethargy that’s been all over me for the last few weeks (months?!). Got in a full day of work today, and while the arch-bug I’m working on isn’t solved yet, I made progress. Now off to dinner with Orna and then I have tons of work to do at home, including syscalltrack, some kernel work, and homework. No rest for the wicked!

Updated the kernel patches page with the latest patches.

On Thursday, after the first Mathematical Logic class in Tel Aviv, I stopped by the Azrieli center (aka purgatory, home of the clueless), to grab a quick dinner before heading home on the train. Since I never eat dinner alone without something to read, I stopped by the local Steimazky (book shop) and picked up a copy of Dan Simmons’ Hyperion. By the time I finished dinner, I was hooked. Late yesterday night, I finished it, only to discover that there are three(!) more books in the series, and if I want resolution, I’ll have to read them too. Damnit, it’s a good story.

Yesterday, Orna brought home my long awaited birthday package from the post office – Neal Stephenson’s Quicksilver). I’ve only just started reading it, and it’s every bit as good as I expected. Hooray for good books!

For ladypine‘s benefit, the original reference.

lj-compose now starts with the new buffer in auto-fill-mode, just the way I like it. Steps to achieve this worthy goal:

At the top of lj-write.el, define a variable to hold the hooks list:

 
(defvar lj-compose-hook nil 
    "Hooks to be called on lj-compose startup") 

Change lj-compose() to call our hooks:

(defun lj-compose (&optional profile)
  "Compose a LiveJournal update, using PROFILE."
  (interactive)
  (lj--initialize)
  (switch-to-buffer (get-buffer-create "*LiveJournal*"))
  (delete-region (point-min) (point-max))
  (or (eq major-mode 'lj-update-mode) (lj-update-mode))
  (lj--insert-initial-buffer-contents nil profile)
  (set-buffer-modified-p nil)
  ;; call our startup hooks 
  (mapcar 
       (lambda (x) (funcall x))
       lj-compose-hook))

And add the hooks you want to the initialization files (.xemacs/init.el in my case):

(add-hook 'lj-compose-hook '(lambda () (auto-fill-mode t))) 

That’s all folks!

I’ve had a lousy couple of days, lack of motivation, long hours of staring at the massive sea of crud we call the innernet. I’m trying to snap out of it, but keep getting drawn back in. I think maybe I’ll disconnect from the net over the weekend. It will be … redemption.

Hope the next entry won’t take another 6 months…