Gmail testing on the rise

Seems like the Gmail folks are stepping up their testing; I was given the notification to invite 3 people to use Gmail twice in one week! Yeah, I’ve been using it for quite a while now, but 6 invites in one week… could mean they’re now entering the post-beta stage of testing.

However, konqueror still isn’t supported, I hope they get around fixing it soon.

Linus on qmail

A thread’s going on on LKML about a glibc bug that caches the value of getpid(), which breaks programs that create threads. Someone then reported qmail doing several getpid()s. Here’s Linus’ reply:

qmail is a piece of crap. The source code is completely unreadable, and it seems to think that “getpid()” is a good source of random data. Don’t ask me why.

It literally does things like

random = now() + (getpid() << 16); and since there isn't a single comment in the whole source tree, it's pointless to wonder why. (In case you wonder, "now()" just does a "time(NULL)" call - whee.). I don't understand why people bother with it. It's not like Dan Bernstein is so charming that it makes up for the deficiencies of his programs. But no, even despite the strange usage, this isn't a performance issue. qmail will call "getpid()" a few tens of times per connection because of the wonderful quality of randomness it provides, or something. This is another gem you find when grepping for "getpid()" in qmail, and apparently the source of most of them: if (now() - when < ((60 + (getpid() & 31)) << 6)) Don't you love it how timeouts etc seem to be based on random values that are calculated off the lower 5 bits of the process ID? And don't you find the above (totally uncommented) line just a thing of beauty and clarity? Yeah. Anyway, you did find something that used more than a handful of getpid() calls, but no, it doesn't qualify as performance-critical, and even despite it's peyote-induced (or hey, some people are just crazy on their own) getpid() usage, it's not a reason to have a buggy glibc. Linus

The debate on spatial UIs continues

One more article blasting Gnome’s recent switch to the spatial finder.

Well, it’s right that we relate things spatially, we “know” where the buttons to the light bulb are, and we always expect them to be there. While this definitely does make sense, and it would also be nice to have all things spatial on the computer screen, I somehow don’t like the idea.

Our “view” is limited by the size of the computer screen: 17″ or 19″. This much a window is too less to do things spatially. It does make sense in real world, since we work in 3D and think in 3D. Our hand automatically reaches out to the gear lever on the cars, our fingers reach out to the right keys on the keyboard with some practice, but this is in the world that’s not limited to just some screen-sized resolution. Also, the “objects” we deal with, cars, keyboards, doors, etc., are different and our brain remembers and maps the various “controls” to the right objects.

However, it would be very taxing on our brain to actually map various folders of a single object — “/” or “/home” — in a 17″ diagonally-spaced window. The screen is going to run out of unique new start-locations for all the folders we have, and access. So our brain not only has to remember the namespace object to the folders, it also has to have a hash of the co-ordinates of the screen which maps to the various folders that are opened at those co-ordinates.

I haven’t used the Gnome spatial nautilus yet, but I’ve experienced the spatial explorer by way of Windows 95; and it sure wasn’t a pleasant experience.

India – Pakistan: citizens’ view

So the India – Pakistan One Day International series ended yesterday with a lot of cheering and support for both the teams by cricket-lovers from both the nations. It’s great that India won the first series in Pakistan, also proving the people wrong who called the Indian team as ‘chokers’, following the recent trend of performing well throughout some series but losing in the all-important final. This should give the Indian team a lot of confidence and joy. Thanks, you’ve done us all proud.

The Times of India today has a report on the hospitality extended to the Indians by Pakistanis: “When you say you?re Indian, strange things happen. Faces break into smiles, doors fly open, rickshaw drivers ask you home, cyber cafes waive away hours of use, tea is called for, Pepsi bottles are uncorked, invitations to lunch and dinner are proffered.”

I sure am disappointed to not have experienced this first-hand, but I’m sure this isn’t a special case. I hope there’ll be many more such tournaments and events and I’ll get to visit Pakistan some time in the future. Cricket really is a passion in both the nations and the people from both the countries have proved that we have nothing against each other — we love the sport, we love the players and we love each other.

It’s also heartening to know that the leaders of the two nations are working towards better ties and curbing the terrorist activities along the border.

I now really hope the Pakistan team tours India and gives us the opportunity to shower our affection and hospitality on them.


Had a horrible day yesterday; guess I pulled some muscle in sleep which caused my lower back ot hurt. Couldn’t even move in the morning. It got better as the day progressed; could walk normally, but with occasional jerks by evening. Spent whole day lying on the bed and reading stuff.

Software Patents

Here are some useful links for learning about software patents:

arch vs darcs

Hmm… very interesting indeed. DARCS vs arch: this is a post on the darcs-users mailing list which lists one case where darcs shines over arch. Basically, it allows you to establish (or rather, itself establishes) relationships between patches, so that if you retrieve patch-11 which is an enhancement over patch-8, it would retrieve both, patch-8 as well as patch-11. Pretty cool feature anyways, should be implemented in any VCS.

Article on SCM systems

Just read David Wheeler’s comments on CVS, SVN and GNU Arch. Probably because of a bias towards arch, he lists out the problems of arch and has asked the developers on the arch-users mailing list to look into them / confirm them. Overall, it’s a pretty good article and provides nice insights. He’s talked a bit on monotone and other versioning systems at the end.