Re: Lua in NetBSD

[Antti Kantee <pooka%cs.hut.fi@localhost>]

On Sat Oct 24 2009 at 16:13:15 -0400, Thor Lancelot Simon wrote:
> On Sat, Oct 24, 2009 at 10:48:36PM +0300, Antti Kantee wrote:
> > 
> > Where's your careful design?  "some cool things: <list>" is not a
> > careful design.  Having a proprietary careful plan we can't see places a
> > lot of trust in you (as well-placed as it might be).  A research project
> > doing something similar is very interesting and I look forward to their
> > results with great anticipation.  But what if their conclusion in a year
> > is "it was a horrible idea and Y works much better"?
> > 
> > I don't even know why we are having this discussion.  If Lua-proponents
> > would have done their homework properly and presented a coherent proposal
> > a month later based on facts instead of could-be and can't-show, nobody
> > would have to argue.
> 
> Perhaps if you want anyone to take you seriously you should stop moving
> the goal posts.  Why would anyone go to any trouble to offer you a
> "careful design" when you've previously said -- unless I misunderstood --
> that you are only interested in seeing finished code?

You failed to quote "For the sake of argument (and that's what this is),
let's assume decision based on careful design.".  Unless my English is
failing me, that does NOT translate to "here's what I want this time".
I am not interested in seeing a 100-page corporate quality yaddayadda
document.  I was just pointing out where your arguments are total
mumbo-jumbo.

> As far as I can tell Marc wants some guidance from core, and his peers,
> on general direction: is integrating Lua with NetBSD a good thing, or a
> bad thing?  I think it's a good thing, as do, evidently, a number of
> other developers, including one who actually prototyped it and offered
> his patches for review.  We've had a reasonably useful discussion of some
> of the things we could use Lua for if we had it in userspace or in the
> kernel.  And then on the other hand you're standing across the way
> getting, it seems to me, all worked up, putting words in the mouths of
> those who disagree with you, and presenting, it seems to me, an
> ever-changing list of objections and demands.
>
> I don't think that's particularly helpful to anyone.  What do you want
> to see?  A list of proposed applications, with details?  That, you have,
> several in fact, from several developers.  A proposal for how to do the
> actual integration of Lua into the system?  For userspace, Quentin's
> offered a patch which works.  I haven't seen any comments on it.  I don't
> think it makes sense to do any work on kernel integration until we have
> userspace settled.  Or do you want to see actual working code for all the
> applications people want to use Lua for?  Because if that's what you want,
> I think it's something you're not likely to get.  Particularly not if it
> looks as if, if you get it, you might just move the goal posts again as
> you did above.

I want to see the same thing I've told you about 5 times now: implemented
and working applications.  Not an exhaustive set of applications anyone
could every implement, just some (preferably >1, since I hear they are
really easy to do).  I think this is a reasonable demand if someone asks
for objections to integration (and Marc's exact project descriptions are
"Adding Lua to NetBSD userland", "Adding Lua to the NetBSD kernel").

If someone asks "do you think it's a good idea to try out integrating Lua
to NetBSD", I will say "yes".  If someone asks "do I get an irrevocable
right to integrate Lua to NetBSD", I will say "show me the data from
your trials".