Konban ha minna-san,
On Thu, 28 Apr 2005 16:30:53 -0400
Timothy Miller <[EMAIL PROTECTED]> wrote:
> We are working out our time schedules. It's still going to take a
> while, since a bunch of things are hard to predict. In any event,
> since the software is going to be MOSTLY a community effort, I was
> wondering if people would be able to help me estimate development
> times for the software bits. We'll then fit them into the work
> schedule, and that'll help figure this all out.
As nobody else answered, here one from me, but i don't think
you will like it.
Long time schedules don't work with OSS development.
Mostly because OSS work is done in free time. And as we
all know, free time is scarce. Also, the amount of free
time is highly volatile and may change from one day
to another. Thus, it's nearly impossible to plan further
than a few weeks ahead.
Although, if you have 1-3 motivated OSS programmers
you can get a few thousand loc per week without any
problems, there can be months without any line written,
due to the lack of time.
The problem for us now is, that we have to do the plan
now and we cannot rely on these changing conditions.
One way out of it would be to partialy pay a few of
the main figures for their development.
The next arising problem that TS cannot pay the whole
development could be (partialy) solved by raising
founds for it.
IMHO i would start with the hardware desgin or to be exact
with the simulator and build anything around it. Then people
could work on the software while you are developing the
hardware. And you'd get an idea how fast the software
progresses.
I'd also release an FPGA prototype as soon as possible, even
if the chip isn't finished yet, to enable the key developers
to work on the software part.
Or to say it in the word of ESR: "Release fast, release often"
Attila Kinali
--
éãåããéãåã
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)
