[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: SEUL: Interest abounding...

At 14:51 1998-01-13 -0800, Erik Walthinsen wrote:
>The basic idea is for all existing groups to collaborate and create a single
>core distribution, upon which all distributions could theoretically be
>based.  It could be called (if you wish), the "official Linux core".  It
>would consist of a stable set of core utilities, libraries, and basic
>programs, and would be versioned as a single block.

This would be a Good Thing!

It's evident from the discussions on usenet and mailing lists that things
like difference in libc versions and SysV vs BSD style init cause problems
for many people.

I also think that it's a logical step to extend the coordination of linux
from the kernel to this `core´. I'm sure there will be discussions about
how much to include in this, but I suppose it's reasonable to start small
and let it grow with time.

In a long term perspective I think it should include things like default
X window manager and system administration tools etc. Certainly package
installer. With all respect for pluralism, I think it would be a major
advantage if people could buy or download whatever linux distribution they
desire, install it and find themselves in an environment that is familiar to
anyone who has used vanilla linux. Then they can reconfigure it as much as
they desire! ;-)

So far Linux has mainly copied features in other unix versions, getting more
and more of the functions of competitors like Solaris and AIX. It's about
time for Linux to take the lead -- not only in performance and stability, but
also by breaking new ground in user interface and features.

There is certainly enough work for many teams and distributors without
maintaining duplicate systems for basic functions. Who cares about `office
software´? (The Wine project obviously needs a double adrenalin injection.
If Sun is so eager to compete with MS they ought to make the Wabi source free.
That would be a boost for linux/unix.) But the Linux killer application will
be something that doesn't exist on any platform today.

After 50 years of development, computers still don't understand what we say.
You should be able to talk to a computer, and it should be able to interpret
instructions and write documents from your dictation. OK, such systems exist,
but they certainly don't work well enough. With the close ties between linux
and the research world, this ought to be a possibility.

A home computer should obviously be able to control the home, making sure
that the stove if turned of when you leave home, and warning you if the
electricity usage is unusually high etc. With all the geeks around who play
with both hardware and linux, this should certainly be doable. OK, it is a
reality today in places, I'm sure, but not as a _product_ (AFAIK).

With x86 based `Biscuit PC´ board sized 6 * 4 inches well capable of running
Linux, and with wireless networks etc, we have a whole new world to explore.
There are already embedded systems with Linux in the industry, and these
devices are getting cheaper. Why should a firewall or a web server be bigger
than a palm sized box that you plug to the wall with some cords. Why can't I
access my home alarm system via WWW? Would be simple with Linux.

I'm sorry that I wandered away so far from the original thread, but with
all the amazing possibilities we have with Linux, it seems so absurd that
so many people are arguing about petty details such as package formats. I
hope there will be a core linux that handles these things well enough, and
that the rest of us can do more creative things.

If we manage to make a few Linux killer systems or killer apps, I'm sure
the big software manufacturers will start porting -- at least if there is
a generic linux core so that they don't need to stay compatible with half
a dozen different distributions. Just as Word has replaced WordPerfect as
the major word processor I'm sure Word will in time be replaced by something
that isn't owned by makers of the big ugly operating system. They will port.


Magnus Lycka, S/W Engineer, M.Sc.E.E; Folktrov. 6C, 907 51 Umea, Sweden
Tel: +46(0)90 198 498, GSM: +46(0)70 582 80 65, Fax: +46(0)70 612 80 65
<mailto:magnus.lycka@tripnet.se>         <http://www1.tripnet.se/~mly/>