[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
SEUL: I want SEUL first edition to be called SEUL 97
I don't want SEUL2050. I want SEUL be issued this year. In order of
avoiding Linux becoming vaporware we need this:
1) Not be too ambitious: this is the key for starting a chain reaction.
2) Try to find solutions requiring a MINIMUM of work
1)NBot be too ammbitious. How to start a chain reaction.
I want SEUL and I want it fast. A fast SEUL will allow us to start a
chain reaction in the same way than in the Linux project.
Linus said: "Instead of writing a great operating system I made a
small one, not so great. Then I made it free. That turned it into a
great operating system". Because Linus was smart enough to make
something with attractive features (virtual memory) but small enough
to be written by a single man in a few months. Moreover this kernel
worked. So he attracted people to work on his project and with this
team of programmers he made the Linux we know. Had he started an
ambitious kernel we would be now using W95 (shudder).
We face two problems: limited resources and scarceness of user
friendly software. So I think than we must concentrate on using
existing software whenever possible for SEUL 1 because that would
allow us to issue SEUL 1 in a few months and start a chain reaction.
Because if SEUL attracts to Linux people who are unable to cope with
it that will have an effect both on our numbers, on the software
available and in the way of thinking of the other distribution
authors. On our numbers because we will attract people. On the
software available because developpers of both free nad commercial
software will begin to think than easiness of use is not a useless
feature for a Linux user. On the authors of distributions because
they will begin to think in end users when building their new
With these elements (new software, more resources, better
distributions) we will be able to produce better releases of SEUL in
2) Trying to find solutions requiring a MINIMUM of work.
When deciding about what distribution we will use as base, what window
manager or anything else we must be able to MAKE COMPARATIVE REVIEWS
and to decide not on the basis of what we are using or what we like
but on the basis of what software already includes what we want.
I will include two examples of what we are doing wrong now. Please nobody
First: Someone has called a vote for the distribution who will be our
basis. Sorry but in my country we don't conduct elections without an
electoral campaign to allow the voter to know the candidates. If you
like Debian make a complete review of it in order we know its
strengths and weaknesses. And remember the thing we are loooking for
is not if Debian is a great distribution (it is), the quality we are
looking for is if it makes a good starting point for SEUL: how much
work to make it user friendly and nice (the user will want to
configure ugly software besides it looks unprofessional).
Second: When I pointed the screen configurator in KDE as being a good
thing someone pointed than we could write a similar tool for
AfterStep. Well I like AS so much I am using it now but I do not want
it for SEUL. Using it would mean spending resources writing a
configurator just to give him one of the capabilities of KDE. Until it
gets a configurator of its own (I have heard of ASCP but still not
tried) AfterStep must be discarded from the WM competition.
Do not start unnecessary projects. An exmple of this si the question
of kernel compiling.
There is a trend of discussion to build software for making kernel
compilings easy. So we would be spending lots of programming effort
for the sake of allowing the user do kernel compilings. And no matter
how user friendly this software the fact is than we would be giving
him plenty of rope to hang itself (remmeber he does not know what is
an SCSI disk). But WHY IN THE HELL WOULD HE NEED TO RECOMPILE THE
KERNEL? In 2.0 if we use modules we can provide him with an universal
installation kernel and make him choose between less than ten
production kernels. If he chooses wrong he just can try the next one.
The kernels can be as fast and as small as the ones compiled by the
user. With the benefit of having been compiled by knowledgeable
people. AND THAT WOULD NEED FAR LESS RESOURCES THAN A USER FRIENDLY
KERNEL COMPILING INTERFACE.
We will go nowhere with decisions made by consensus. Sooner or later
someone has to take the decisions. Best thing would be if we had an
undisputable leader like Linus in the Linux-kernel project.
Unfornately this is not the case but it becomes urgent subgroups are
formed, leaders are designated and we begin work. A good thing in the
interim could be having "themes of the week" to discuss a subject to
Jean Francois Martinez
==================== The Linux. Use the Linux, Luke! =======================
Simple End User Linux Mailing list
To be removed from this mailing list send a message to email@example.com
with the line
in the body of the letter.