[PD] ASCSEQ

[timon botez]

Do you know if theres a limit to how much data the ascseq can hold?  
Length of symbols? I need the ascseq to drip at 200 ms. (needs the  
time too spit out the font metrics. not ideal I know, and Ill be  
improving that for the future, but I have a little gig for thursday,  
so this is an interim fix) The proposed method look risky. l2s is  
currently working well...

Cheers,
T.


On 26 Jan 2006, at 11:53, IOhannes m zmoelnig wrote:

> this email goes to the pd-list too...
>
>> Im using [ascseq] to split textfiles into single characters, that  
>> are  being used in GEM text3d. I have a so-so AFM loader which  
>> assembles  the characters with a somewhat accurate spacing. I have  
>> about 2000 of  these separate text3d objects that I use to  
>> assemble typographic  displays. You know, lots of floating letters  
>> in space that every now  and so often rearrange into legible lines  
>> of text - typewriter style.  I have bypassed the problem by  
>> removing/reshuffling certain  characters in the font - have  
>> replaced space with � - asterix. I end  up with scarily long  
>> single words. And it would be nice not having to  do that. The  
>> whole mechanism is very fragile and delays cause the  whole  
>> alignment to jumble up.
>
> in your case (i expect you to have tons of text), i think it is  
> really better to use [ascseq] instead of generating long symbols  
> with spaces in between. (because of the symbol-table; however,  
> benchmarking would be interesting...)
>
>> If the ascseq had a bang for every finished message it wouldnt be  
>> a  problem.
>
> you can get this behaviour via something like
>
> [t b l]
> |     |
> (b2)  |
>       [repack 1]
>       |
>       [t b l]
>       |     |
>       |     [ascseq]
>      (b1)   |
>            (a)
>
>
> the trick is to split the list into 1-atom messages with [repack 1]
> you get a bang at (b1) each time [ascseq] has finished 1 atom
> you get a bang at (b2) each time the whole list has been processed.
>
> so with (b1) you can trigger a [32(
> repack everything into a long list with [repack].
>
>
> mfg.asd
> IOhannes