Octave popularity

classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

Octave popularity

niles-5

The current "Linux Journal" has an article about "rlab" another
free Mat*ab implementation, that doesn't look half as good as Octave
(IMHO).  I don't think it's a good thing to split the few freeware
programmers in this field between, rlab, octave, and SciLab.  Without
outside developer interest the project will burn out the
orig. author. (jwe has done 99% of the work!)  The Linux community,
which seems to be the largest current consumer of free software, seems
largely ignorant of Octave and heavily leaning toward the other two.
The "Linux Journal" has mentioned rlab, and Scilab many times and
don't know that they EVER mentioned Octave.  Also, it seems from the
article that many rep. people (Univ. profs) are contributing
"toolboxes" to rlab.  Those are toolboxes that could of been written
for Octave.  Of course, since it's all GPL we could still "port"
them. I personally, would like to see more users and developers use my
favorite, Octave.  Therefore, I propose two things: (1) a new official
version is released soon (it has so many nice features now and it's
been so long!) (2) jwe and/or some us other people write an article
for the "Linux Journal".  My impression is that they're always
desperate for new material if you write it they'll put it in the next
issue.  (Of course, anything anyones writes should be checked with
jwe, first, to be polite. :)

What does everyone think?  Of course, no one but jwe can decide a release
date, but it's just a suggestion.

        Thanks,
        Rick Niles.



Reply | Threaded
Open this post in threaded view
|

Re: Octave popularity

Ian Searle-3
[hidden email] wrote:
>
> The current "Linux Journal" has an article about "rlab" another
> free Mat*ab implementation, that doesn't look half as good as Octave
> (IMHO).

I'm glad that is only your humble opinion :-)

> I don't think it's a good thing to split the few freeware
> programmers in this field between, rlab, octave, and SciLab.  Without
> outside developer interest the project will burn out the
> orig. author. (jwe has done 99% of the work!)

I don't agree. How the freeware programmer's split their time is outside
of anyone's control. I have been contacted at least several times over the
years about using Rlab in a commercial venture. I have even been warned
that certain parties were violating the GNU Copyleft (my team of laywers
was busy with other matters that week :-). Anyways, the point is that
these commercial ventures must compete in a commercial software world that
is certainly more flooded with competing programs than the freeware world.

Also, as freeware users, you should realize that nobody can make a lifetime
commitment to a software project. Personal factors (employment and family)
and factors outside of our control (OS and hardware evolution for example)
make it tough to guarentee anything. I am sure John is committed to providing
support and improvements for Octave for as long as he can, as am I for Rlab.
But, I suspect that in 10 years you will not want to use either anyways

> The Linux community,
> which seems to be the largest current consumer of free software, seems
> largely ignorant of Octave and heavily leaning toward the other two.
> The "Linux Journal" has mentioned rlab, and Scilab many times and
> don't know that they EVER mentioned Octave.  

I disagree again (this is getting to be a habit :-). If I had to venture a
guess, I would say that Octave probably has more users than Rlab, but less
than Scilab. Probably the most significant factor in usage levels of any of
these programs is Matlab compatibility. Both Scilab and Octave offer more
Matlab compatibility than Rlab (at least at first look).

> Also, it seems from the
> article that many rep. people (Univ. profs) are contributing
> "toolboxes" to rlab.  Those are toolboxes that could of been written
> for Octave.  Of course, since it's all GPL we could still "port"
> them. I personally, would like to see more users and developers use my
> favorite, Octave.  Therefore, I propose two things: (1) a new official
> version is released soon (it has so many nice features now and it's
> been so long!) (2) jwe and/or some us other people write an article
> for the "Linux Journal".  My impression is that they're always
> desperate for new material if you write it they'll put it in the next
> issue.  (Of course, anything anyones writes should be checked with
> jwe, first, to be polite. :)

I would like to see more toolboxes written for Rlab... I could look
at each Octave mfile as something that wasn't written for Rlab, but that
would be self defeating. Motivation and happiness are key factors that
are often overlooked. I did not embark upon the Rlab project to write
the dominant scientific high level language of the 20th century... I did
it for the education, and my own technical needs (this was in 1989/90).
The fact that others have benefited is great, but it was not my original
motivation. I expect the same is true with many other authors (I know a
few, so that helps), they start a project to fufill some personal need, and
it often grows into something else altogether.

I agree (first time :-), you ought to write an aritcle for the Linux Journal.
However, they are not that desparate. They have a very competent editor, who
will not take just any slop that comes along. I spent significantly more effort
on the article than I thought I would. Far more than on technical paper I have
ever written (I usually publish at least once a year at AIAA/SPIE/ASME
conferences).

However, I don't think Rlab, Octave, and Scilab authors, or users should get
into any sort of competition. All are usefull tools, with their own niche. I
believe we can all co-exist, and even derive benefit from each other without
any unpleasantness.

> What does everyone think?  Of course, no one but jwe can decide a release
> date, but it's just a suggestion.

Cheers,
--
Ian Searle
[hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: Octave popularity

niles-5
> [hidden email] wrote:
> >
> > The current "Linux Journal" has an article about "rlab" another
> > free Mat*ab implementation, that doesn't look half as good as Octave
> > (IMHO).
>
> I'm glad that is only your humble opinion :-)
....

>
> I would like to see more toolboxes written for Rlab... I could look
> at each Octave mfile as something that wasn't written for Rlab, but that
> would be self defeating. Motivation and happiness are key factors that
> are often overlooked. I did not embark upon the Rlab project to write
> the dominant scientific high level language of the 20th century... I did
> it for the education, and my own technical needs (this was in 1989/90).
> The fact that others have benefited is great, but it was not my original
> motivation. I expect the same is true with many other authors (I know a
> few, so that helps), they start a project to fufill some personal need, and
> it often grows into something else altogether.
>
> I agree (first time :-), you ought to write an aritcle for the Linux Journal.
 
> However, they are not that desparate. They have a very competent editor, who
> will not take just any slop that comes along. I spent significantly more effo
rt
> on the article than I thought I would. Far more than on technical paper I hav
e
> ever written (I usually publish at least once a year at AIAA/SPIE/ASME
> conferences).
>
> However, I don't think Rlab, Octave, and Scilab authors, or users should get
> into any sort of competition. All are usefull tools, with their own niche. I
> believe we can all co-exist, and even derive benefit from each other without
> any unpleasantness.

Please, I didn't want to make it a competition.  I also didn't want to
insult anyone elses work or be unpleasant. I guess I feel somewhat
loyal to Octave and didn't want it to go the way of Sony's betamax
format just because not enough people know it exists.  Also, I was
realize the amount of work that goes in to a product like Octave or
RLab, and just felt it was a duplication of effort, in terms of
manpower.  However, you point is quite valid that people write free
software to fill a niche and for fun and learning and it grows to be
something bigger.  I guess I would just like to see all this work
going to one product that we can all make better than all the current
packages.  However, perhaps that's not the best route.  I'd just like
to see more developers behind Octave and I want to apologize if I've
offended Rlab or yourself.

        Rick Niles.

Reply | Threaded
Open this post in threaded view
|

Re: Octave popularity

Jim Van Zandt
In reply to this post by niles-5

Rick Niles writes:

>...Therefore, I propose two things: (1) a new official
>version is released soon (it has so many nice features now and it's
>been so long!)

Yes!  Please!

It's natural to delay a new release in an attempt to "get it right".
However, (1) in the mean time a lot of folks have to contend with
problems in the official release that have already been fixed, or (2)
they give up and switch to something else, (3) the machinery for
handling releases get rusty, and (4) an official release will be
installed by more people, and any problems will therefore be
discovered, and probably fixed, sooner.

I know beta releases are available for the asking.  I've had to step
down from beta testing myself for quite a while, for lack of gcc
2.7.2.  You still can't compile Octave with the latest Slackware
distribution of Linux, as far as I know.  This suggest another
argument (4a) there's a better chance that people will go to the
trouble of preparing binary distributions that can be installed by
those without the facilities, notably compiler and *lots* of disk
space, for compiling it themselves.

I heard Linus Torvalds talk a while back.  Of course, he has very
frequent beta releases of the Linux kernel.  He says that helps with
debugging.  There are relatively few changes from one version to the
next, making it easier to figure out which change broke some feature.
(He has the benefit of "ten thousand beta testers", which also helps.)

By now, I think the beta version of Octave is enough better than 1.1.1
that most users would willingly spend the administrative effort to
install a new version.  That alone should justify releasing it.

Besides, for how many bug reports has jwe replied "It's fixed in my
current sources"?  A new release would have eliminated not only those
bug reports, but also the effort to reply to them.

I think the time is long past for a new release.

                           - Jim Van Zandt