DO_Ray

discovered some irregularities in this patch.

It is a derivate from a bigger one, which crams 224 Kbyte in a char 

Even with 10 numbers this patch doesn't work, although it should.

How come a delay command sometimes does then again does not solve this ?

On the whole i'd like to get an idea how this thread bussiness in 

how to test for malfunctioning and the like.

Asking questions on behalf of a silent majority.

discovered some irregularities in this patch.
It is a derivate from a bigger one, which crams 224 Kbyte in a char  
matrix.

Even with 10 numbers this patch doesn't work, although it should.
What am i missing here ?
-
How come a delay command sometimes does then again does not solve this ?
It's supposed to end a thread isn't it ?
-
On the whole i'd like to get an idea how this thread bussiness in  
general works,
how much of a load you can let it have,
how to test for malfunctioning and the like.
-
Asking questions on behalf of a silent majority.
regards DO_Ray.
-> PATCH ->:



help-with-datacramming

It's unclear to me what you're trying to accomplish, but assuming you are trying to gather numbers in the bottom message box, I notice one thing. When you hit the start button, the last thing you hit this message box with is a "set" message. "set" with nothing following it will clear the box.

>discovered some irregularities in this patch.

>It is a derivate from a bigger one, which crams 224 Kbyte in a char matrix.

>Even with 10 numbers this patch doesn't work, although it should.

>How come a delay command sometimes does then again does not solve this ?

>It's supposed to end a thread isn't it ?

>On the whole i'd like to get an idea how this thread bussiness in general works,

>how to test for malfunctioning and the like.

>Asking questions on behalf of a silent majority.

Chris Muir | "There are many futures and only one status quo.

cbm@well.com | This is why conservatives mostly agree,

 | and radicals always argue." - Brian Eno

It's unclear to me what you're trying to accomplish, but assuming you are trying to gather numbers in the bottom message box, I notice one thing.  When you hit the start button,  the last thing you hit this message box with is a "set" message. "set" with nothing following it will clear the box.

At 5:46 PM +0100 11/29/07, DO_Ray wrote:
>Hi all,
>
>discovered some irregularities in this patch.
>It is a derivate from a bigger one, which crams 224 Kbyte in a char matrix.
>
>Even with 10 numbers this patch doesn't work, although it should.
>What am i missing here ?
>-
>How come a delay command sometimes does then again does not solve this ?
>It's supposed to end a thread isn't it ?
>-
>On the whole i'd like to get an idea how this thread bussiness in general works,
>how much of a load you can let it have,
>how to test for malfunctioning and the like.
>-
>Asking questions on behalf of a silent majority.
>regards DO_Ray.

-- 
Chris Muir           | "There are many futures and only one status quo.
cbm@well.com         |  This is why conservatives mostly agree,
http://www.xfade.com |  and radicals always argue." - Brian Eno


Quote: DO_Ray wrote on Thu, 29 November 2007 08:44

----------------------------------------------------

> discovered some irregularities in this patch.

> It is a derivate from a bigger one, which crams 224 Kbyte in a char 

> Even with 10 numbers this patch doesn't work, although it should.

It seems to work fine for me. Did you mean for the last bang from your "start" trigger to bang the message set to the message box? What this patch does is fill up a message box with a list (in a fairly Byzantine manner) and then erases it, real quick. Either disconnect that set message, or hang a print from [prepend append] to see that everything is working.

For your bigger problem, you might need to work with Overdrive on, rather than off. And, search the Max manuals for "queue throttle." Max has different lists of things to do: the queue for things that happen "right now", the scheduler for things that happen at a certain time (eg metro; usually higher priority), servicing things like drawing the UI and checking on the mouse state, etc. They all are in a delicate balance, which can be tweaked. Look in the Options menu at "Performance Options...". Objects such as defer, deferlow, qmetro are helpful, as is delay (which moves bangs from the queue to the scheduler) and other timing objects.

Quote: DO_Ray wrote on Thu, 29 November 2007 08:44
----------------------------------------------------
> Hi all,
>
> discovered some irregularities in this patch.
> It is a derivate from a bigger one, which crams 224 Kbyte in a char  
> matrix.
>
> Even with 10 numbers this patch doesn't work, although it should.
> What am i missing here ?
>
>

It seems to work fine for me.  Did you mean for the last bang from your "start" trigger to bang the message set to the message box?  What this patch does is fill up a message box with a list (in a fairly Byzantine manner) and then erases it, real quick.  Either disconnect that set message, or hang a print from [prepend append] to see that everything is working.

For your bigger problem, you might need to work with Overdrive on, rather than off.  And, search the Max manuals for "queue throttle."  Max has different lists of things to do: the queue for things that happen "right now", the scheduler for things that happen at a certain time (eg metro; usually higher priority), servicing things like drawing the UI and checking on the mouse state, etc.  They all are in a delicate balance, which can be tweaked.  Look in the Options menu at "Performance Options...".  Objects such as defer, deferlow, qmetro are helpful, as is delay (which moves bangs from the queue to the scheduler) and other timing objects.

I miss a description what "doesn't work" as well as "should" mean.

Don't consider others are being able to read thougts. I can't... ;-)

(You could place it as comment into the patch...)

> How come a delay command sometimes does then again does not solve this ?

> It's supposed to end a thread isn't it ?

Most likely you have an order of execution problem. Best is to use trace 

to find out how each step of your patch works...

To make this thread helpfull for others, try to figure out yourself 

(take your time...). If it helps give us a quick note what the problem 

was. If it doesn't, come back with a better description what you want to 

DO_Ray schrieb:
> Even with 10 numbers this patch doesn't work, although it should.
> What am i missing here ?

I miss a description what "doesn't work" as well as "should" mean.
Don't consider others are being able to read thougts. I can't... ;-)
(You could place it as comment into the patch...)

> How come a delay command sometimes does then again does not solve this ?
> It's supposed to end a thread isn't it ?

Most likely you have an order of execution problem. Best is to use trace 
to find out how each step of your patch works...

To make this thread helpfull for others, try to figure out yourself 
(take your time...). If it helps give us a quick note what the problem 
was. If it doesn't, come back with a better description what you want to 
achive...

-- 
Stefan Tiedje------------x-------
--_____-----------|--------------
--(_|_ ----|-----|-----()-------
-- _|_)----|-----()--------------
----------()--------www.ccmix.com

On 30 Nov 2007, at 09:58, Stefan Tiedje wrote:

>> Even with 10 numbers this patch doesn't work, although it should.

> I miss a description what "doesn't work" as well as "should" mean.

> Don't consider others are being able to read thougts. I can't... ;-)

> (You could place it as comment into the patch...)

a general explanation of these terms and how they relate towards one 

difficult words from deferlow: queue, high priority, low priority queue.

difficult words from qlim: interrupt, scheduler, overdrive, 

difficult words from qmetro: backlog the queue

these come from the max reference manual.

>> How come a delay command sometimes does then again does not solve 

>> It's supposed to end a thread isn't it ?

> Most likely you have an order of execution problem. Best is to use 

> trace to find out how each step of your patch works...

> To make this thread helpfull for others, try to figure out yourself 

> (take your time...). If it helps give us a quick note what the 

> problem was. If it doesn't, come back with a better description 

What i achieved is: using jit.matrix,jit.spill and iter to quickly 

compute an expression and put the result in an other matrix.

What happend is a major stutter whilst computing, but no crash

> DO_Ray schrieb:
>> Even with 10 numbers this patch doesn't work, although it should.
>> What am i missing here ?
>
> I miss a description what "doesn't work" as well as "should" mean.
> Don't consider others are being able to read thougts. I can't... ;-)
> (You could place it as comment into the patch...)
>
a general explanation of these terms and how they relate towards one  
and other would be most welcome:
difficult words from deferlow: queue, high priority, low priority queue.
difficult words from qlim: interrupt, scheduler, overdrive,  
'interrupt safe'.
difficult words from qmetro: backlog the queue
these come from the max reference manual.

>> How come a delay command sometimes does then again does not solve  
>> this ?
>> It's supposed to end a thread isn't it ?
>
> Most likely you have an order of execution problem. Best is to use  
> trace to find out how each step of your patch works...
>
> Good luck
>
> To make this thread helpfull for others, try to figure out yourself  
> (take your time...). If it helps give us a quick note what the  
> problem was. If it doesn't, come back with a better description  
> what you want to achive...
>
What i achieved is: using jit.matrix,jit.spill and iter to quickly  
compute an expression and put the result in an other matrix.
What happend is a major stutter whilst computing, but no crash
Thanks for the help Stefan
>
> -- 
> Stefan Tiedje------------x-------
> --_____-----------|--------------
> --(_|_ ----|-----|-----()-------
> -- _|_)----|-----()--------------
> ----------()--------www.ccmix.com
>
>


> Quote: DO_Ray wrote on Thu, 29 November 2007 08:44

> ----------------------------------------------------

>> discovered some irregularities in this patch.

>> It is a derivate from a bigger one, which crams 224 Kbyte in a char

> It seems to work fine for me. Did you mean for the last bang from 

> your "start" trigger to bang the message set to the message box? 

> What this patch does is fill up a message box with a list (in a 

> fairly Byzantine manner) and then erases it, real quick. Either 

> disconnect that set message, or hang a print from [prepend append] 

> For your bigger problem, you might need to work with Overdrive on, 

> rather than off. And, search the Max manuals for "queue 

> throttle." Max has different lists of things to do: the queue for 

> things that happen "right now", the scheduler for things that 

> happen at a certain time (eg metro; usually higher priority), 

> servicing things like drawing the UI and checking on the mouse 

> state, etc. They all are in a delicate balance, which can be 

> tweaked. Look in the Options menu at "Performance Options...". 

> Objects such as defer, deferlow, qmetro are helpful, as is delay 

> (which moves bangs from the queue to the scheduler) and other 

I looked in reference manual and found explanations for qlim, 

deferlow and qmetro using the word queue, so other objects don't use 

see relpy on stefans reply for the issue.

> || michael f. zbyszynski -- molecular gastronimist

>
> Quote: DO_Ray wrote on Thu, 29 November 2007 08:44
> ----------------------------------------------------
>> Hi all,
>>
>> discovered some irregularities in this patch.
>> It is a derivate from a bigger one, which crams 224 Kbyte in a char
>> matrix.
>>
>> Even with 10 numbers this patch doesn't work, although it should.
>> What am i missing here ?
>>
>>
>
> It seems to work fine for me.  Did you mean for the last bang from  
> your "start" trigger to bang the message set to the message box?   
> What this patch does is fill up a message box with a list (in a  
> fairly Byzantine manner) and then erases it, real quick.  Either  
> disconnect that set message, or hang a print from [prepend append]  
> to see that everything is working.
>
the last 'set' is a mistake  .
> For your bigger problem, you might need to work with Overdrive on,  
> rather than off.  And, search the Max manuals for "queue  
> throttle."  Max has different lists of things to do: the queue for  
> things that happen "right now", the scheduler for things that  
> happen at a certain time (eg metro; usually higher priority),  
> servicing things like drawing the UI and checking on the mouse  
> state, etc.  They all are in a delicate balance, which can be  
> tweaked.  Look in the Options menu at "Performance Options...".   
> Objects such as defer, deferlow, qmetro are helpful, as is delay  
> (which moves bangs from the queue to the scheduler) and other  
> timing objects.
>
I looked in reference manual and found explanations for qlim,  
deferlow and qmetro using the word queue, so other objects don't use  
the queue ?
see relpy on stefans reply for the issue.
Thanks for the help , mz
>
> --
> || michael f. zbyszynski -- molecular gastronimist
> || http://www.cnmat.berkeley.edu/
> || http://www.mikezed.com/
>
>
>


Help with datacramming