Re: Informal poll

On Tue, Apr 15, 2003 at 10:18:16AM -0300, Lalo Martins wrote:
> On Mon, Apr 14, 2003 at 04:54:18PM -0600, Micah Dowty wrote:
> > 
> > 1. Fixing bugs
> > 2. Adding features
> > 3. Replacing picogui's build system
> > 4. Working on the new pg2 architecture
> > 5. Other?
> 
> Sorry if I sound blunt, but mostly "other" is not an option.
> This project doesn't have enough hackers, and is completely
> voluntary-run; specially, it's run by two volunteers with
> serious motivation problems ;-) we work with what tickles our
> interest, or we don't work at all and lose steam (and go work on
> case mods, 3d games or programming languages, wink wink).

Thank you for bringing some much-needed realism to this list :)
There's a finite amount of development power to go around.. and at the
moment it's pretty small.

> 
> I believe if we want to keep this going, it's about time to
> feature-freeze; from here to 1.0 it's:
> 
> a. finish stuff we're working on (terminal, nifty, cli_py)
> b. replacing the build system, it's starting to be a major
>    bother (I'm getting tired of people unable to build 0.45
>    because automake is frooked up)
> c. svn (or eva) when and if you (Micah) feel like
> --- beta ---
> d. fixing bugs

IMO the tradeoffs to be made here are:

  1. are there things we can do to picogui that, with relatively little
     effort, could bring it more developers?

  2. do current users of picogui have any urgent needs?

  3. the current architecture is destined for replacement, we should get
     to 2.0 as soon as possible

Point (2) contributes to point (1) by making PicoGUI seem more
responsive to user needs. There seem to be a lot of people interested in
PicoGUI- the web site gets a good bit of traffic, there are almost 200
people on this mailing list. But then, there are only about 2 people
actually working on it.

> 
> The launcher is IMO something we can do without; our main target
> for 1.0 is developers and embedders, and they're gonna write
> their own launchers anyway.  (I already wrote mine, it's not hard.)

The only real reason I want a working launcher is so that PicoGUI
packages for the Zaurus can be at least somewhat usable. I have a Zaurus
that I'd like to run PicoGUI on, and it makes a relatively easy way for
people to try it out.

But again this is an issue of promoting what we have versus developing
the future.

> 
> The controllable textbox was an illusion of my part; useful,
> because it spawned the writecmd/writedata separation, but it
> proved unnecessary for what I want, once I added the cursor
> position property.
> 
> Binary packages are easily done, but I think it's rather
> pointless given the breadth of possible configuration
> combinations.  (Then again you could conceivably compile with
> everything, for the package... that's a thought

I don't see much point to making binary packages for the desktop, but it
would be good to have packages for common embedded systems (zaurus,
tuxscreen, vr3, maybe even the helio) for the 1.0 release.

> 
> I think the new build system can make things easy enough that
> people can live without binaries.

Hopefully. I know the vast majority of questions I see about picogui
stem from the build system.

> 
> Finally, documentation.  Yes, documentation is fine.  But I
> don't have time or motivation to write it, and neither does
> Micah AFAIK.  So, again I beg for your tolerance on my
> bluntness, but if you think it's important, write it, or hire
> one of us to do it ;-) (we're probably willing to work for
> hardware too, hint hint, I need a new pda, or perhaps a
> second-hand laptop)

Hmm yes... I could use rent money for the apartment I'll be moving into
this august :)

--Micah

-- 
"For every problem, there is a solution that is simple, elegant, and wrong."
           -- H.L. Mencken


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Welcome to geek heaven.
http://thinkgeek.com/sf