On Thu, 06 Dec 2001 14:50:36 +0100, "Thorsten Froehlich"
<tho### [at] trfde> wrote:

>In article <3c0f6cdd$1@news.povray.org> , "Mael" <mae### [at] hotmailcom> 
>wrote:
>
>> how do you explain the success of others open source projects ?
>
>Success in what regard?  Quality or geek usage?  What about Linux kernel
>2.4.15?  I have yet to see a Linux open source project of quality comparable
>to that POV-Ray when it comes to stability, ease of portability or any
>documentation.

you can't compare the stability of a portable os kernel to the
stability of a raytracing program. It's really hard to make it stable,
because of hardware support.

>The only project close to POV-Ray is Apache, and they have at least one
>trillion worth of financial backup in form of every server vendor on the
>planet.  And their development model is fairly will controlled as well and
>taking place in a field that only needs very widespread knowledge you can
>learn about everywhere.

Of course big development projects have to be steadly controlled. But
I don't think that apache is the only opensource project that is
comparable to povray

>Contrary to this, POV-Ray is near the edge of ongoing research, i.e. in
>areas of radiosity and photon maps.  Obviously the point of entry of new
>developers is much higher than that of a webserver or kernel, to offer only
>two examples...
>> moreover AFAIK many features of pov3.5 came from patches written by people
>> not in the povteam, those same people you seem to think they can't bring
>> anything interesting. Sure, those patches may need some rewrite, better
>The rewrite thingy is the major problem.  As outlined above, ray-tracing is
>slightly more complicated than a webserver or a kernel.  

Slightly more complicated? When you write a webserver or a kernel you
have to really care about stability and security issues (expecially if
they are so widespread as apache and linux are), it's very hard to
keep track of all the work. Who cares is a raytracer crashes or if it
has a bug? The next release will fix them... I'm not saying that's ok
to have bugs in povray, I'm just saying that integrating patches in
povray is easier than in linux or apache as you don't have to care soo
much about those problems (noone will kill you for that... :P). There
are many many opensource projects that are more critical than povray,
for example atlas (blas implementation), or openpgp, almost every
linux server... And some raytracers too :P

>Unfortunately many
>people think ray-tracing is as trivial as a webserver, which is far, far
>from the truth.

I'm not experienced with webserver programming, but I don't think it's
a "trivial" task at all.

>So your argument completely missed the point.  If you disagree, ask yourself
>why the most recent GPL ray-tracer project (I think it was called Panorama)
>failed so early?  According to your theory it should have been more
>successful.....

1 fail does not make a law

Post a reply to this message