From: Andrew Lentvorski <[EMAIL PROTECTED]>
Michael O'Keefe wrote:
The larger problem is that we shouldn't be using a "filename".
Hans Reiser (of Namesys) would agree with you.
What are the proposed alternatives ?
Queries primarily.
You want Joe User to write a query? I'm scared just thinking of that.
I also don't see how queries would help. If you can't even remember the
name (or enough of one to search for) of the file you stuck some info in,
why are you any more likely to remember enough data to write a decent query?
I want the ability to search files. *FAST*. And I want the index built
without turning my computer into a dog while doing it (yes, I'm talking
about Spotlight).
Why? I find I never search files- I know where I keep my stuff. On rare
occassion, I use a grep over a small number of files, usually when
programming and looking for what references a function. Maybe once or twice
a day, when I'm coding. Searching in files is just not something most
people really need. And given the proliferation of file formats, its not
really possible on an OS level unless you restrict the file types you index
over (which Murphy says would eliminate the one the user wants).
The big problem right now is that heavy disk access grinds most consumer
systems to a halt. In fact, I would argue that heavy anything grinds most
consumer grade systems to a halt.
??? Heavy disk usage can grind a single app to a halt, but other apps on
the system tend to be unaffected. Unless they too need to get files from
disk, in which case what do you expect? But I have frequently run CPU
intensive and IO intensive programs on the same box with no problem. Thats
the whole point of a multitasking OS.
It's ironic given that Linux/BSD/et al. are held up as examples of
monolithic kernels being superior to microkernels, but I think that we're
finally about to make a shift to microkernels. Too many things are
starting to make real-time demands that computers can't service. Both Vista
and OS X did *major* surgery on their kernels in order to cope with the
demands of video playback.
I fail to see how moving to a microkernel would have any effect on heavy
disk usage effecting overall system performance. Maybe I'm overlooking
something, but I see no inherent causality here. Care to explain your
logic?
Gabe
--
[email protected]
http://www.kernel-panic.org/cgi-bin/mailman/listinfo/kplug-list
