Older blog entries for mjs (starting at number 75)

I notice aaronl is picking fights with everyone who responded to his flamage, but clearly hasn't read my tips on accurate measurement of memory use, because he has neither pointed out where I'm wrong, nor retracted his misleading remarks.

"RSS" and "SIZE" are not remotely accurate measures of the actual memory load a program places on a system; they can either underestimate or overesitmate, for a huge variety of reasons, of which shared libraries are only one.

OK, slight refinement. You shouldn't measure the memory each time with `top'. Instead, run `free' and look at the free memory column in the -/+ buffers/cache row (since buffer cache is not real memory load and can always be tossed). Also look at the free swap column.

It seems most people don't know how to measure a program's memory usage. Doing a ps and looking at SIZE or RSS totally does not cut it.

Here's how to do it right (in my opinion):

First you need to make sure your system is relatively quiescent, and memory load won't be randomly jumnping around. Now run `top'. Look at the memory and swap free. Record these numbers.

Now run the program you want to check. Keep watching `top', and see what the numbers settle at. Compute the difference. This is the only real way not to be fooled by thinks like shared libraries, buffers, mmap'd I/O, memory shared by multiple kernel threads that show separately in ps, other sorts of shared memory...

Of course, this only measures one prorgram. If shared libraries are getting used a lot, it's best to measure a set of programs meant to be used together all at once, since a lot of memory will be shared between them. This is the major blunder a lot of people make. Measuring one program that uses lots of shared libraries, if you are going to use any of the others, penalizes the program unfairly for being a good code reuse citizen.

life

I think I have gotten infected with the Advogato Malaise that has been going around lately.

tv

Bring out the Iron Chefs!

Shut up, Eskil.

Shut up George.

Some highlighs from LWE:

• Nautilus looking way better than it ever did on my desktop
• StarOffice running embedded in Nautilus
• The GNOME Party
• The Eazel afterparty - for the actual hackers
• "I'm sorry. Wanna feel on mine?"
• Hanging out with all my friends from Helix and other places
• "<clahey>"
• "It's brrrrrrrroken!"
• Helix Code: The Tchotchke Company
• "You must drink, Comrade!"
• Weird card games
• Using an Eazel sticker to make my own Eazel T-shirt

mathieu, the correct term is "Darinate", not "Darinify". As in "I sure got Darinated today" or "could someone Darinate my code please?"

hacking

The Nautilus tree view should be quite nice now. In fact, it seems a lot faster than the tree or list views. But then, the code is much simpler.

Please, please add some info about what free software projects you are involved in, and how. It makes it much easier to certify you. I promise to give anyone who puts project info and asks me at least an Apprentice rating.

hacking

I finished doing a bunch of medusa stuff to make it more usable for other hackers at Eazel; hopefully the search feature will progress faster now.

weekend

Well, the party on Friday was pretty cool. We started out with a bath tub full of beer bottles in ice water and ended up with an empty bathtub. That was scary. Many cool people came. On Saturday, I went with Steph, Rebecca, Arlo and some other folks to The Tonga Room, an amusing tiki bar in San Francisco. Those umbrella drinks pack a hell of a punch. I also got stood up for a date this Sunday, but it has been rescheduled now, and the other party provided a decent excuse.

We are now officially in "crunch mode" at Eazel, heading towards the feature-complete milestone. I will probably post a bit less often.

certifying the `rms' account

ncm comments about this. I certified it, even though I am reasonably sure it is not actually Richard M. Stallman. I did it because whether or not the account is real, RMS himself deserves props.

I guess this creates the minor possibility that someone might try to use the account to impersonate RMS and say things he wouldn't say, but that has not happened yet.

rating accuracy

I think some rankings are probably inflated. However, this is a natural result of the fact that anyone with a Master rating can singlehandedly make someone else a Master. This means even a few people excercising poor judgement can throw a lot of noise into the system. I try to be very careful not to over-certify people, although I think sometimes I go overboard, and don't give people a Master rating just because I think others might not think they deserve it.