Forums > MaxMSP

poofs with GTK/4.5.6/XP Pro SP2a

February 10, 2006 | 4:12 am

I’m using the above on a P4 with the latest version of lobjects I could find and the latest version of the Granular Tool Kit.

The problem I’m having seems particularly bad while running any of the GTK objects or helpers but I believe it’s happened with some other patches/objects as well.

Essentially, clicking some control at some point during use causes the app. to disappear, stay resident and become impossible to terminate via the Task Manager thus requiring a hard reset.

My audio interface is an E-MU 1616m. I have tried HyperThreading on and off, different sample rates, vector sizes and a completely different PC.

I have a lot of Extras installed on both machines, like Litter, TapTools, and FFTease, I’ve double-checked that everything’s where it should be and I’ve even reinstalled.

Are there any known issues with any combination of the above configuration?

Thanks!

Matt


February 10, 2006 | 12:38 pm

On Feb 9, 2006, at 11:12 PM, matt wrote:
> Essentially, clicking some control at some point during use causes
> the app. to disappear

You are going to have to be more specific than that.

—–
Nathan Wolek
nw@nathanwolek.com

http://www.nathanwolek.com


February 10, 2006 | 2:21 pm

On 2/10/06, matt wrote:
>
> Essentially, clicking some control at some point during use causes the
> app. to disappear, stay resident and become impossible to terminate via the
> Task Manager thus requiring a hard reset.

I can only tell you I have the same sort of crashes where max stays resident
in the taskmanager. On my computer (also xp sp2) the crashes and the
sticking of the max process, are not related. I have had all kinds of
different crashes and whether max sticks seems to depend on someting else
that’s in the patch. I’m trying to figure it out right now. I have set up a
simple js to crash max and now I’m looking which objects make it stick.

On my system, waiting for about 5 minutes makes the max process finally
terminate. I keep checking my email after every crash… I’m thinking this
behavior is introduced with 4.5.6, cause it never happend before and I’ve
been working on the same patch for a long time.

If you want to try for yourself use this fatal js to crash max:

autowatch = 1;

function bang(){

x = this.patcher.newobject("toggle",122,90,15,0);

this.patcher.remove(x);
this.patcher.remove(x[0]);
}

best, T_


February 10, 2006 | 3:30 pm

i will try but so far it doesn’t seem like there’s a pattern. i will gather more information and come back with it.
thanks.


February 10, 2006 | 3:32 pm

Quote: thijskoerselman@gmail wrote on Fri, 10 February 2006 07:21
—————————————————-

> I can only tell you I have the same sort of crashes where max stays resident
> in the taskmanager.
—————————————————-

thanks. max doesn’t go away on my system (requires hard reset), and you’re right, it happens after all sorts of crashes, come to think of it, but only on the system with the E-MU interface running at 96 kHz and using the E-MU ASIO driver. On the other system, I haven’t had any Max stickiness and it’s using stock AC97 hardware running DirectSound or ASIO4ALL.
i’ll try your test, soon…


Viewing 5 posts - 1 through 5 (of 5 total)