RunRev and Linux

[Peter Alcibiades]

Its not Rev's fault that Linux has small market share.  No, of course not. 
But Richard, you cannot seriously be arguing that it is either sensible or
acceptable to release poor quality software as long as the target markets
are small?  As to its being the same half dozen people.  Well, maybe.  How
many people on the list are exclusively using Rev on Linux?

Rev's problem seems to be that it has already made the decision to have a
Linux offering.  So it has to make it a good quality one.  Our task is not
to raise the market share of Linux in order to motivate it to improve its
offering.  Our task may well be to try to help Rev in making their offering
better by testing, inputting, even contributing if we are able.  I will do
what I can if its needed.  But I'm not going to accept that because my
platform has low market share right now, I have no right to expect a quality
product until it rises.

Fewer features, maybe.  Those that it has must work.

Richmond, my suggestion to anyone testing Rev on Linux is to get as close to
bare metal as you can.  Slackware, which I do not use myself, is the closest
in this respect.  If it fails on that, you can be absolutely sure the
problem is with the app, not with some tweak of packaging.   My second
suggestion is to use the most stable, if not exactly leading edge, distro
you can find, and that is Debian Stable.  It is the most tested and
scrutinized one there is.   The reason is, then you can be pretty sure that
bugs you encounter are not coming from your distro.  Debian Stable by the
way is also not what I use myself.

I know that many people like and use Ubuntu in various versions.  What we
are looking for however is not something that pleases us.  We are looking at
a tool for diagnosis of the extent of problems in what is reported to be
more or less buggy software.  So we need to eliminate as far as possible any
suggestion that there is something about our distribution which is causing
our problems.  The closest thing there is in the Linux world to an
unproblematic distribution, though its not one I myself want to run every
day particularly, is Slack.  It will do the job we need.  No-one can
seriously say, if Rev has a problem on a basic install of Slack, that there
is some problem with the distro or setup.  If its a decent Linux app, it
will run properly on Slack.  Debian Stable is a close second in this
respect.

It is the same reason why, in the present state of affairs, I would avoid
testing on KDE4.  People may like KDE4 a lot, but up to now, if you had a
problem, it was very difficult to pin down between it and the application.

This is horses for courses.  The problem now is eliminating variability and
tying down the problems.

Richard's advice is load it up and use it.  Yes, I agree, but not casually.
do it in a disciplined way.  And on a modern machine with a modern sized
screen.  Get someone to use it in native mode, not in a VM, and on a full
time exclusive basis.  Do it professionally.  We can all potentially gain
from that. 

I suspect that quite a few of the quality problems here come from excessive
reliance on testing  on Ubuntu running on Parallels.  I use VMs, they are
wonderful.  But they are not the real thing.  I would never have found the
screen resolution and printing problems on my VMs, because I run them in a
window with different resolution settings from the primary machine.  There's
lots of stuff like this.  If people have not found it, they have not done
enough with VMs.  There's a ton of problems out there, waiting for you, at
the interface between a real OS and real hardware, which you will not find
with the virtual hardware on the VM.
-- 
View this message in context: http://n4.nabble.com/RunRev-and-Linux-tp1835808p1836199.html
Sent from the Revolution - User mailing list archive at Nabble.com.