On 23/07/2020 11:41, Bald Eagle wrote:
> I'm not sure what might draw people in - but it could be anything.
> 
> There are job-search sites, and perhaps there are types of job headings that
> requests for developers could be listed under.  A shiny ray-traced graphic or
> two and a link to the HOF could pique someone's interest.

I considered reaching out to some of my contacts in .edu-land to see if
I could get some interested students. However I decided not to (at least
at this point), for two reasons: firstly, past experience has shown that
asking for volunteers from /outside/ the user community rarely results
in someone who sticks around for long enough to become helpful. The
person really needs to genuinely be interested in POV-Ray itself, and if
they are then it's likely they'll approach us (or we'll hear of some
work they've been doing).

Secondly, in depth-work on povray can be *hard* to get right. Sure, it's
easy to make tweaks here and there but to really work on the program as
a whole you need to have a solid grasp on how everything works together
as adding what appears to be a simple feature has the potential, in some
cases, to do really unexpected things *if* you don't know how it works
as a whole. To get a developer up to speed on the guts as a whole takes
time. Plus whoever it is needs to be really competent with C++ and
unmanaged languages (by which I mean having to do your own memory
management).

That said, I can see where we *could* use a student or less-experienced
developer who wanted to round out their background a bit: it would be
great to have someone formally document the inner workings in a detailed
form (i.e. at least enough so that someone who doesn't know how
everything fits together could read it and understand). This would make
getting up to speed on the codebase for anyone who wants to do so a fair
bit easier and would also help in splitting it into bits for the 4.0
rewrite.

But that's a lot of work ...

Honestly there are just so many open-source projects around the world
these days - many of which are way more glamorous than POV while also
being easier to work on - that I kind of understand why we have little
luck in attracting volunteers.

Now if I can get moray integrated with POV and all working nicely then
we'd have a bit more bling and maybe could attract some help, but even
then I have my doubts. Many people seem to think GPU raytracing is the
future and software renderers are dead but simply don't understand that
at this point GPU's are still incapable of the level of precision and
generic applicability that a fully primitive-based renderer is, and I
don't see that changing anytime soon (as there's no real need to).

Sorry if I sound negative and I'm not trying to be a wet sponge here.
I'm just tired I guess. Good reason to have someone who's a bit more
optimistic driving release cycles. But you know what? Despite the fact
this thread has been a bit, um, rowdy, it's shown me that there's still
a bunch of people who do care and has helped improve my feelings about
the whole thing. I might even enjoy getting stuck back into the code,
time will tell.

-- Chris

Post a reply to this message