Re: [OT] Why does X need so much CPU power?
On Thu, 2003-09-04 at 04:00, Colin Watson wrote:
> On Thu, Sep 04, 2003 at 01:01:19AM +0300, Micha Feigin wrote:
> > Second is the fact that most people just use the OS they get with their
> > computer and are afraid to try and replace it. Plus, they already paid
> > for the M$ license (even if forcefully/unknowingly) so why switch to a
> > free one after you already paid for something. Same thing with macs/osX,
> > where its even harder to get linux to work.
> > Linux need to change its public image and start coming pre-installed.
>
> You're aware that this has already started to happen, right? Lindows
> are/were doing exactly this with cheap PCs sold by Wal-Mart in the US.
>
> > There is also the problem of too much options. Unlike M$ where people
> > exactly what program does a given job, on linux there are 10, and when
> > there are so many its actually sometimes harder finding the right one
> > for you or even finding what programs there are to do a given job.
> > Too much choice can sometime be as much a liability as not enough (as
> > much as I like the options).
>
> My impression of Lindows has been that they're presenting single
> applications for each niche, fixing this problem for their target
> audience. I'm quite sure others will follow.
>
> As far as Debian is concerned, we have our niche of providing maximal
> choice and power and I think it's right that we stay there. Many of the
> early contributors to and leaders of Debian wanted to make it something
> that could work well as a base for more customized distributions, and
> that seems to be succeeding quite nicely. To those who say that Debian
> has too much flexibility and choice for Mr. Average, I say: that's OK.
> Not everything has to cater for Mr. Average, and that still doesn't stop
> us doing useful things one level back and catering for the people who
> cater for Mr. Average.
>
> > Don't take from this that I don't like linux. I think its much better
> > then M$ and there are no alternatives for me for some of the things that
> > it offer, but its exactly those things that make, at list for the
> > moment, to be a non-option for the Joe-Public m$ user.
>
> I think there's a lot of work to do before we're ready to replace the
> major proprietary operating systems completely, but the situation is
> improving year by year so I don't see any grounds for despair. The
> balance is still swinging Microsoft's way, but is beginning to tip with
> news of organizations like the city government of Munich and major banks
> switching over, which erode the document format lock-in that Microsoft
> Office has had for many years. Once organizations are no longer locked
> in to what the organizations they deal with use, the balance can only
> tip further.
>
> In my opinion, it's only after that happens when we need to be ready for
> home users. Office use leads this kind of thing, and is easier because
> businesses can afford to hire sysadmins and provide basic training to
> smooth over the wrinkles. Only after that happens on a large scale do
> you start getting lots of office workers thinking "hey, I wonder if I
> could use this to handle things at home?", and so on.
>
> So I don't think it's necessary to prophesy doom because there are still
> problems that would confuse those who aren't so technically literate.
> We've got time to work on these, and it makes sense to be realistic
> about our audience in the meantime so that we don't do a disservice to
> those who are already interested and capable.
>
I completely agree to you comments, I was just trying to point where are
the problems with hitting the home niche. I never said that linux has
arrived at a point that it ready to solve those problems.
Even when it finally is it will probably be distributions like lindows
providing customized distributions.
I quite like the options debian give me and I try to start up m$ on my
computer as little as possible, usually just to keep a backup of my
girlfriends files in case she may want to work on them away from home.
On that note, the debian package tree could be organized a little better
it order to help finding the alternative programs for performing given
tasks. This could be done with a field in the deb file linking it into
some virtual tree. The problem with the current state is that its not
always clear where to look for a given program. For example you may look
for a sound editing program in any of kde/gnome/x11/sound. Also, the
current implementation allows for only one value, where sometimes
several are applicable.
I am aware that this can be mostly done using search terms, but this
method has its advantages, since searches to produce too many/little
results, depending on the terms and description included. Also,
sometimes you just have an idea of what to look for but you don't know
exactly the term defining it.
> Cheers,
>
> --
> Colin Watson [cjwatson@flatline.org.uk]
--
Micha Feigin
michf@math.tau.ac.il
Reply to: