Personal suggestion for fixing the Linux situation

Richard Gaskin ambassador at fourthworld.com
Fri Jul 16 09:08:12 CDT 2010


Damn fine post, Peter.  I've quoted it in its entirety because anyone 
here who missed it and uses Linux would be missing out if they didn't 
read it.

Excellent stuff - very exciting to see this level of energy from so many 
of us going into this, kinda like how Linux itself is made.

I like your approach of targeting a minimal install, and I think it's an 
excellent compliment to the admitted fetishism I have with the 
relatively bloated Ubuntu.  Between your slackware, my Ubuntu, Mark 
Wieder's SUSE, and there's gotta be at least one of us using Red Hat, we 
should have a good matrix of small and large and in-between distros to 
coordinate our testing and diagnostics among.

I really appreciate your willingness to dive in and help with the work 
of improving the Rev experience on Linux.  I've never been more 
optimistic about the outlook for my Rev work in Linux than I feel right 
now.  Thanks.

-- 
  Richard Gaskin
  Fourth World Media Corporation
  ___________________________________________________________
  Ambassador at FourthWorld.com       http://www.FourthWorld.com


Peter Alcibiades wrote:

> Here is how I would go about tracking down these things.  Just to recap,
> the things we are seeking to track down are these four:
>
> -- not all and only installed fonts are visible and useable
> -- revPrintField does not work properly
> -- virtual desktops don't work
> -- editor slows down, freezes and crashes
>
> Are there other priority areas?  There are other niggles, but are there
> other real basic functionality showstoppers?
>
> My suggestion for going about tracking this down is quite different from
> what most people here will instinctively want to do.  The general view here
> is that Linux is an enormously complex mix of components, so the thing to
> do is pick some large general purpose distro and standardize on it.  I do
> not believe this to be the solution.  In fact, it is a wrong diagnosis of
> the problem.  This approach, which regards each distro as a distinct OS, is
> actually part of the problem.
>
> Were I in charge of the effort I would proceed in EXACTLY THE REVERSE WAY.
> I would seek to find the minimum installation set, and within that, the one
> closest to the way packages are released by the developer, that will allow
> the reproduction of the problems.
>
> You can argue about which distro will most readily meet these requirements,
> but if you want to start from something fairly simple and mainstream and
> not start compiling the whole thing from scratch, the contender that leaps
> out at you is Slackware.  I accept, there could be an argument for going
> even further down, like Slitaz or TinyCore.  Maybe that is worth a try as
> well, but they are not, as Slackware is, deliberately as untweaked as
> possible.
>
> So I  would propose doing a minimal install of slackware, with nothing but
> the basic system and the most basic window manager, probably OpenBox.
> Maybe Metacity without Gnome desktop environment, if you want to be as
> close as possible to mainstream what it will have to run on.  But no
> Firefox, no OpenOffice, no apps at all.
>
> If you can reproduce the problems on this sort of minimal install, then you
> are much closer to the source, because you have basically ruled out all
> distro specific issues.  If not, then start to add stuff until you do get
> the problems.
>
> I understand that on this list there is a, well, a precoccupation, with
> Ubuntu as a distro for use.  This is not about use however, this is about a
> tool to get to the source of the problems.
>
> I'm prepared to do serious work on this, but am not capable of writing
> patches to the IDE myself, and before getting started on the project, would
> welcome comment, and would like us to have an agreed approach, so what do
> you all think of the above?  It would also be nice to have some feedback
> from Edinburgh, to the effect that given contributions from us, they will
> do their bit also.



More information about the use-livecode mailing list