sharing data


    Apr 02 2007 | 5:53 pm
    hi - is there a way to share data between identical objects in seperate sub patches without using pattr? say for instance i were to create a patch with one slider object in it and then replicate the patch ten times in seperate sub patches, what would be the most effective way to handle 'wireless' communication with just one of the patches? what i want to do is have one universal sequencer object that sends its control information to multiple different places exclusively - so rather than call up the individual subpatch and input the pattern i would be doing it all on one object and then sending the toggle/slider/flonum etc etc etc. data to one particular, but indentical object wirelessly.. even better would be if anyone could suggest a way of doing it in which that the universal object acts as a virtual 'control', so any changes to data would be immedietly sent to a pre-chosen destination.. any ideas? thanks

    • Apr 02 2007 | 6:45 pm
      I haven't done much with this, but I believe (correct me if I'm wrong) that vade has some stuff for that. Unique identifiers for each instance, etc. I'm sure others have similar abstractions, externals, but I'm not at my computer right now.
      Keith
      On 4/2/07, ol wrote: > > hi - is there a way to share data between identical objects in seperate sub patches without using pattr? say for instance i were to create a patch with one slider object in it and then replicate the patch ten times in seperate sub patches, what would be the most effective way to handle 'wireless' communication with just one of the patches? > what i want to do is have one universal sequencer object that sends its control information to multiple different places exclusively - so rather than call up the individual subpatch and input the pattern i would be doing it all on one object and then sending the toggle/slider/flonum etc etc etc. data to one particular, but indentical object wirelessly.. even better would be if anyone could suggest a way of doing it in which that the universal object acts as a virtual 'control', so any changes to data would be immedietly sent to a pre-chosen destination.. > any ideas? thanks >
    • Apr 02 2007 | 7:02 pm
    • Apr 02 2007 | 7:28 pm
      On 2 Apr 2007, at 19:45, keith manlove wrote:
      > I haven't done much with this, but I believe (correct me if I'm wrong) > that vade has some stuff for that.
      I'm of the overkill school of thought when it comes to data sharing in Max: I use relational databases. But that's probably just me.
      -- N.
    • Apr 02 2007 | 8:11 pm
      Nick Rothwell wrote: > > On 2 Apr 2007, at 19:45, keith manlove wrote: >> I haven't done much with this, but I believe (correct me if I'm wrong) >> that vade has some stuff for that. > > I'm of the overkill school of thought when it comes to data sharing in > Max: I use relational databases. But that's probably just me.
      Relational databases are scrutinized and optimized to a degree that the average Max patch data-sharing scheme never will be. They'll scale predictably. External databases allow all sorts of interesting possibilities for data manipulation.
      Sure, it's overkill - no doubt about that. I don't store nearly as much data for Max as I do with my mail client, and credible arguments have been made against using databases there. (Only 15952 messages in my "music" archive folder.) But I've been very happy to follow your progress facilitating Max-database communication because I have a severe aversion to ugly surprises in the end stage of large projects.
      I am willing to trade some constant predictable overhead for the healthy future of my sanity.
    • Apr 02 2007 | 8:22 pm
      On 2 Apr 2007, at 21:11, dlurk wrote:
      > But I've been very happy to follow your progress facilitating Max- > database communication because I have a severe aversion to ugly > surprises in the end stage of large projects.
      There is that, yes.
      I also recently had a Max/MSP based installation which needed to run calculations from a block of legacy code written in C, and I didn't have time to refactor it into an external object, so I just dug out a spare Mini-ITX computer, bunged Linux on it, installed MySQL, and populated the database via a Perl file scraper. Max/MSP on a Mac mini then trawled the database via MXJ.
      OSC is for wussies. Real programmers network their applications with JDBC.
      Nick Rothwell / Cassiel.com Limited www.cassiel.com www.myspace.com/cassieldotcom www.loadbang.net
    • Apr 02 2007 | 8:37 pm
      On 2 Apr 2007, at 21:22, Nick Rothwell wrote:
      > OSC is for wussies. Real programmers network their applications > with JDBC. > Oh Nick, you are a devil!! ;-)>
      L
      Lawrence Casserley - lawrence@lcasserley.co.uk Lawrence Electronic Operations - www.lcasserley.co.uk Colourscape Music Festivals - www.colourscape.org.uk
    • Apr 02 2007 | 8:49 pm
      On 2 Apr 2007, at 21:37, lawrence casserley wrote:
      >> OSC is for wussies. Real programmers network their applications >> with JDBC. >> > Oh Nick, you are a devil!! ;-)>
      Actually, my current toy is DataSynapse GridServer. It's big in the financial institutions but I reckon an MXJ library for it would be possible, should one want to do realtime synthesis with several thousand servers at once.
      (I guess it would have to be additive.)
      Nick Rothwell / Cassiel.com Limited www.cassiel.com www.myspace.com/cassieldotcom www.loadbang.net
    • Apr 02 2007 | 9:57 pm
      you can share things using coll.... if you put a 1 as the second argument you won't get file not found errors. so make a
      coll shared 1
      and put that exact name in each coll in whichever subpatch joo need.
    • Apr 04 2007 | 2:17 pm
      Nick Rothwell schrieb: > OSC is for wussies. Real programmers network their applications with JDBC.
      Love that... I always had the feeling better to jump from Midi to something superior than OSC... ;-)
      Instead of twirking my head with object oriented confusion, get into data based programming... ;-)
      Your work is encouraging...
      All statements without knowing what we are talking about... ;-)
      Stefan
      -- Stefan Tiedje------------x------- --_____-----------|-------------- --(_|_ ----|-----|-----()------- -- _|_)----|-----()-------------- ----------()--------www.ccmix.com
    • Apr 04 2007 | 2:17 pm
      Nick Rothwell schrieb: > (I guess it would have to be additive.)
      like Hedge fonds...
      Stefan
      -- Stefan Tiedje------------x------- --_____-----------|-------------- --(_|_ ----|-----|-----()------- -- _|_)----|-----()-------------- ----------()--------www.ccmix.com
    • Apr 04 2007 | 2:41 pm
      ol schrieb: > hi - is there a way to share data between identical objects in > seperate sub patches without using pattr?
      Is there a specific reason why you would not want pattr? It would be the easiest and clearest...
      Else you could do this...
      -- Stefan Tiedje------------x------- --_____-----------|-------------- --(_|_ ----|-----|-----()------- -- _|_)----|-----()-------------- ----------()--------www.ccmix.com
    • Apr 04 2007 | 6:04 pm
      ol schrieb: > hi - is there a way to share data between identical objects in > seperate sub patches without using pattr?
      Forget the patch I just sent, this is better and easier... (need more sleep... ;-)
      -- Stefan Tiedje------------x------- --_____-----------|-------------- --(_|_ ----|-----|-----()------- -- _|_)----|-----()-------------- ----------()--------www.ccmix.com
    • Apr 06 2007 | 11:23 am
      On 4 Apr 2007, at 15:17, Stefan Tiedje wrote:
      > Instead of twirking my head with object oriented confusion, get > into data based programming... ;-)
      Try iBATIS or Hibernate: you get to do both at the same time...
      Nick Rothwell / Cassiel.com Limited www.cassiel.com www.myspace.com/cassieldotcom www.loadbang.net