--- Francois Gouget fgouget@free.fr wrote:
But do we really have Windows applications that mess with the timeouts? Unless we know of many applications that really do stupid things, we should let them modify the screensaver timeouts. Otherwise it is just paranoia. And if we don't trust Windows applications we should not let them run in the first place. They are many things worse than changing the screensaver settings that they could do: just imagine all you can do with an evil API like DeleteFile, or CreateWindow.
The worst thing may happen is the same mess with settings as on Windows machines. Most Windows users reinstall OS not more frequently than 2 times a year. Sure, restarting X session so often is a huge annoyance ;-)
Andriy asked:
Do you suggest another extreme - "Integrate with X
whenever
possible"? If not, where do you suggest to draw
the line between
these two extremes?
I would say that I lean in favor of the "Integrate with X whenever possible" option. But I did not study all the SPI_ settings in enough details to really say that it always makes sense. As Alexandre says there can be exceptions: just not the same ones. Though it may be that by putting our exceptions together we manage to cover all of the SPI settings ;-)
Exceptions will be not so many as I see from the first 20 implemented system parameters. For the most system parameters I just don't know what to think. E.g. what type of applications use SPI_GET/SETBEEP, SPI_GET/SETKEYBOARDSPEED, are interested in exact mouse buttons mapping?
Andriy Palamarchuk
__________________________________________________ Do You Yahoo!? Find the one for you at Yahoo! Personals http://personals.yahoo.com