[PD] Re: [PD-dev] launching tcl/tk scripts from pd

[Federico]

Federico ha scritto:
> Mathieu Bouchard ha scritto:
>> On Sun, 5 Nov 2006, Federico wrote:
>>> Mathieu Bouchard ha scritto:
>>> i discovered delays of toxy when dealing with messages that needs to
>>> be timing-accurate
>>>> Do you want to give me a hand making/designing this?
>>> yes! (as I said you on IRC)
>>> let me know some newz ;)
>> Ok, if I'm not staying too long in Toronto, we're having a developers
>> meeting tomorrow at Ottawa university (matju + _heri + alx1) and we're
>> going to talk about GridFlow II. Then I'd like to make a prototype of
>> that external as soon as possible, possibly using SWIG; _heri has gained
>> quite a bit of experience with SWIG. Once we have something that works
>> I'll make you try.
>>
>> However, there's a slight chance that we back off from that project (of
>> making an external for server-side tcl), so I don't really want to
>> promise anything tonight.
> 
> maybe Toronto's nights are too late for me :)
> if you came on IRC early I would be happy to hear (or partecipate) to te
> discussion
> 
> til now, I have a question wich didn't make me spleep tonight:
> also GridFlow is affected by those client/server issues?
> 
> does GridFlow run in the client? cause i'm experiencing timing problems.
> withouth making a too long explanation, i have:
> 
> [metro]
> |
> [# some operators]
> |
> [#export_list]
> 
> and it's totally inaccurate (with cpu load at 5~10% its clearly audible
> that messages or whatever are off-time)
> 
> can't wait for the dev meeting :)


ah, nevermind that

it seems i have discovered the source of problems...
my video acquisition card is generating strange warnings about IRQ....

bttv0: IRQ lockup, cleared int mask [bits: FMTCHG* VSYNC* HSYNC* OFLOW*
HLOCK* VPRES* 6* 7* I2CDONE* GPINT* 10* RISCI* FBUS* FTRGT* FDSR* PPERR*
RIPERR* PABORT* OCERR* SCERR*]
...
bttv0: IRQ lockup, cleared int mask [bits: HSYNC OFLOW VPRES I2CDONE*]


cause without loading the bttv module my patch seems to work ok.
so sorry for allarmistic bug spam :)