[PD] (netpd)U(Pd-extended), Pd-ext bug-tracker (was Re: elitism, software and academia)

[Roman Haefeli]

On Mon, 2007-06-18 at 19:05 -0400, Hans-Christoph Steiner wrote:
> On Jun 14, 2007, at 1:07 PM, Roman Haefeli wrote:
> 
> > On Tue, 2007-06-12 at 21:46 -0400, Hans-Christoph Steiner wrote:
> >
> >
> >> It would very nice if it was just plug and play.  It would not be
> >> that hard to do it.  I think you could spend a day on it and have it
> >> working smoothly.  It would be very worthwhile, but I think you have
> >> already spent far more time trying to help people get it going than
> >> it would take to fix things.
> >
> > reading that post a second time, i feel somehow insulted by your
> > assumption, that i consciously do not fix things, that simply could be
> > fixed spending a day for them. so, please tell me, what do you think
> > needs to be fixed?
> 
> Certainly no insult of any kind was intended.  I was just quite  
> frustrated by the experience.  We were having a network jam at the  
> end of the NIME conference, a few of us wanted to use netpd.  But  
> only Alexandre was able to get it running.

hm, i am sorry about that.

> Please don't take my comments to be saying something bad about your  
> skills or the work you put it.  It can be a hard problem to solve,  
> getting everything running smoothly, but it is certainly possible.   
> You have been very good at providing help for people to get it up and  
> running.  I'd just like to see netpd get to the point where you can  
> spend less time helping people get it running and more time improving  
> things.

as i said, i know the problems (if they could be considered as
problems), but didn't find a way around them.

> In the general terms, I think it should be quite possible to make  
> netpd "just work" on any Pd-extended install with the user just  
> opening a patch in Pd.  If you want to base netpd on pd-vanilla, then  
> you'll need to provide any externals that are needed for the various  
> platforms.

netpd actually should 'just' work with any pd-installation, though three
points are critical:

- having the right externals loaded (that is: zexy, maxlib, iemmatrix,
iemlib1, iemlib2, iem_t3_lib)

- having the correct netpd-path in the netpd-settings dialog (this one
annoys me most, because it is due to the
'open-message-path-is-relative-to-pd's-startlocation'-problem [to
mention this problem again])

-having netpd/abs in the pathes

if these settings are correct and netpd is still not working, then
something is definitely wrong.


luckily just yesterday, i had a (very simple) idea, for which i waited
for years: 
instead of opening the netpd-patches with the (for me) inconvenient
'open'-message, i want to load them as abstractions. this has the BIG
advantage, that i can specify the location of a patch (now abstraction)
relative to the parent patch (creator in this case). by doing that i can
get rid of the very unwanted 'netpd-path' setting. AND this has a very
nice side effect: when all patches are actually the same patch, i can
add a search patch with only one [declare], that is valid for all loaded
netpd-patches (now abstractions). 
in short: in future versions of netpd there will be no need for
'netpd-path' and for a -path flag anymore.
there is one critical point left: having to have loaded the right
externals. with [declare], each netpd-patch can define for itself, what
it wants to have loaded. that means, the only thing, a user will have to
care, is to have installed the needed externals (which is the case
anyway in extended)

i actually don't have time to implement all these changes and afaik the
actual stable release of pd-extended is based on 0.39, which lacks
[declare]. but when pd-extended switches to 0.40 and i'll have made the
necessary changes, things will be hopefully much easier than today for
eveveryone, the pd-extended users and pd-vanilla/external users. 

roman









	
		
___________________________________________________________ 
Der frühe Vogel fängt den Wurm. Hier gelangen Sie zum neuen Yahoo! Mail: http://mail.yahoo.de