PJ <[EMAIL PROTECTED]> wrote: > Niyam wrote: > > > > > unless, are we supposed to merely agree with all-things linus, the > > > pied-piper on the gates of dawn? > > Gora Mohanty <[EMAIL PROTECTED]> writes:
<snip> Guys, am taking the liberty to comment and explain as below. Please feel free to cross-check my facts and counter comment (if you want to!) The reason for the is as below:-- (1) Device Manufacturers often have black-magic happening in the driver code. By black-magic--> I mean, drivers for things like hard-disks etc have stuff like checks for the chip-set version--- by writing a value to a certain port-- and polling till another value can be read--- and using that to decide the version number--- which in turn will be used to set up delays and loop-iterations for busy-wait spin-locks---> optimal values for which are often discovered by trial and error-- through testing and through tech support's solutions to problems to earlier versions of the hardware-- and collected in a manufacturer's bug/solution database-- and used to optimize the next version of the driver. This expensive; highly technical; labour intensive work is often unglamorous to hacker-types. Also, this same work costs the manufacturers a lot of money-- which is necessarily spent to maintain their market position. Further-more documenting a lot of their work would make the manufacturers liable-- to be sued etc... or have the reliability of their peripherals questioned by rivals--- or even possibly have rivals use seemingly "open source techies" to shred their reputation. Hence, even though many device manufacturers are more than happy to release drivers; and as far as possible help open-source folks developing drivers--- they at the same time make sure their backs are covered-- with respect to the following kinds of issues:-- (1) legal issues (2) trade-secret related issues (3) possible rumour mongering issues (like some supposedly open-source guy claiming that XYZ's hard-disk is least reliable-- based on false arguments). That's why the above "partially proprietary" type of issues. That's also why manufacturers often issue certain chunks of their code as BLOBs of object-code/machine-language byte sequences instead of c source code. Hope that helps... Regards, N.S _______________________________________________ ilugd mailinglist -- ilugd@lists.linux-delhi.org http://frodo.hserus.net/mailman/listinfo/ilugd Next Event: http://freed.in - February 22-24, 2008 Archives at: http://news.gmane.org/gmane.user-groups.linux.delhi http://www.mail-archive.com/ilugd@lists.linux-delhi.org/
