4.5 & 4.6 on Macintel


    Oct 18 2006 | 2:55 pm
    Been off-list for a few months. Good to be back...
    Is there a way to have version 4.5 & 4.6 both setup on my new Macintel? Figured this would be a convenient way to use patches that are dependent on my old CFM externals until I get everything ported.
    I figured it would have something to do with moving the "Application Support" stuff into the application folder, right? Just wanted to see if anyone had concrete instructions before doing some testing myself.
    Searched the docs and forum. Could not come up with the right search times. Feel free to point them out to me.
    ----- Nathan Wolek nw@nathanwolek.com http://www.nathanwolek.com

    • Oct 18 2006 | 4:26 pm
    • Oct 18 2006 | 4:44 pm
      "Old" externals built for the PPC architecture can't work natively under Max 4.6 on Intel machines. I'm afraid you'll have to wait for an Universal Binary of it if there's not one already.
      This has been asked many time here. Don't forget to have a look at the old messages.
      Hope this helps.
      Best,
      Julien.
      Le 18 oct. 06 à 18:26, 曾 興魁 a écrit :
      > > error (path...) granola~: loading old CFM externals disabled. > do you have experience?
    • Oct 18 2006 | 5:36 pm
      On 18-Oct-2006, at 16:55, Nathan Wolek wrote: > Is there a way to have version 4.5 & 4.6 both setup on my new > Macintel?
      Sure. Just do it.
      Max 4.5.x installs itself into the folder /Applications/MaxMSP 4.5/ Max 4.6.x installs itself into the folder /Applications/MaxMSP 4.6/
      Application Support/ isn't an issue because the Cycling '74/ folder has been moved inside the main MaxMSP 4.6/ folder. (Discussed a few times on the list but I'd be hard put to find a query turning up those messages.)
      I've had no conflicts, I even run both at the same time. Despite the overhead for Rosetta on 4.5, it's much faster when testing PPC/CFM/ Carbon versions of my externals than the G3s at my disposal.
      I wouldn't recommend 4.5 as a production environment for audio/video on an Intel machine, but it's fine for testing and working on patches that require non-UB'd externals and even deals with MIDI reasonably adequately. I can't believe someone asked "why?" for this given the litany of laments regarding CFM-only 3P externs, of which there are still a fair few.
      Speakin' of which: there has been some discussion on the dev list touching on porting questions. That list is sufficiently focussed that I would recommend browsing through the entire archive of this year.
      Good luck, Peter
      -------------- http://www.bek.no/~pcastine/Litter/ ------------- Peter Castine +--> Litter Power & Litter Bundle for Jitter Universal Binaries on the way iCE: Sequencing, Recording & Interface Building for |home | chez nous| Max/MSP Extremely cool |bei uns | i nostri| http://www.dspaudio.com/ http://www.castine.de
    • Oct 18 2006 | 7:23 pm
      On Oct 18, 2006, at 1:36 PM, Peter Castine wrote: > Sure. Just do it.
      For those that may encounter this thread in the future, Peter is right. Installed 4.5 and now I can run old stuff just fine. Works right next to my newer 4.6 install.
      ----- Nathan Wolek nw@nathanwolek.com http://www.nathanwolek.com
    • Oct 18 2006 | 7:23 pm
      On 18 oct. 06, at 19:36, Peter Castine wrote on the maxmsp list:
      > Despite the overhead for Rosetta on 4.5, it's much faster when > testing PPC/CFM/Carbon versions of my externals than the G3s at my > disposal. Yes Peter, this is much handier than working with 2 machines but I wondered: Is it reliable to test PPC version externals only with Rosetta on Intel Macs? I mean, is it sure they will always behave the same (run fine) on PPC machines?
      Roby
    • Oct 18 2006 | 8:27 pm
      On 18 Oct 2006, at 18:36, Peter Castine wrote:
      > I've had no conflicts, I even run both at the same time.
      I've had 4.6.x throw all its toys out of the pram because of a MaxAudioAPI.framework belonging to some earlier incarnation, so the PPC and UB installations are capable of lobbing projectiles at each other...
      -- N.
      nick rothwell -- composition, systems, performance -- http:// www.cassiel.com
    • Oct 19 2006 | 8:54 am
      On 18-Oct-2006, at 22:27, Nick Rothwell wrote: > On 18 Oct 2006, at 18:36, Peter Castine wrote: > >> I've had no conflicts, I even run both at the same time. > > I've had 4.6.x throw all its toys out of the pram because of a > MaxAudioAPI.framework belonging to some earlier incarnation, so the > PPC and UB installations are capable of lobbing projectiles at each > other...
      Hmmm. I installed 4.5.6 first, then 4.6. Don't know if that order is significant. Also, I didn't test all the betas that were out there, is it possible there were some changes to the framework along the way?
      It's good to know that the 4.5/4.6 combination can go wrong, but it can also go right. Nathan's main concern (the location of the 'Cycling 74 folders) isn't the problem.
      -- P.
      -------------- http://www.bek.no/~pcastine/Litter/ ------------- Peter Castine +--> Litter Power & Litter Bundle for Jitter Universal Binaries on the way iCE: Sequencing, Recording & Interface Building for |home | chez nous| Max/MSP Extremely cool |bei uns | i nostri| http://www.dspaudio.com/ http://www.castine.de
    • Oct 19 2006 | 9:07 am
      On 18-Oct-2006, at 21:23, Roby Steinmetzer wrote: > Is it reliable to test PPC version externals only with Rosetta on > Intel Macs? > I mean, is it sure they will always behave the same (run fine) on > PPC machines?
      You're only going to get 100% reliability with extensive end-user beta-testing covering every possible hardware configuration on Planet Earth.
      Rosetta seems to be sufficiently identical to PPC-based chips for testing purposes in the context of the one-[wo]man development team building externs. My experience with beta-testing Litter Power and porting another dozen or so externs is that I have had zero issues due to Rosetta vs. real PPC.
      But I take no responsibility for what happens to you.
      -------------- http://www.bek.no/~pcastine/Litter/ ------------- Peter Castine +--> Litter Power & Litter Bundle for Jitter Universal Binaries on the way iCE: Sequencing, Recording & Interface Building for |home | chez nous| Max/MSP Extremely cool |bei uns | i nostri| http://www.dspaudio.com/ http://www.castine.de
    • Oct 19 2006 | 9:45 am
      On 19 oct. 06, at 11:07, Peter Castine wrote:
      > On 18-Oct-2006, at 21:23, Roby Steinmetzer wrote: >> Is it reliable to test PPC version externals only with Rosetta on >> Intel Macs? >> I mean, is it sure they will always behave the same (run fine) on PPC >> machines? > > You're only going to get 100% reliability with extensive end-user > beta-testing covering every possible hardware configuration on Planet > Earth. > > Rosetta seems to be sufficiently identical to PPC-based chips for > testing purposes in the context of the one-[wo]man development team > building externs. My experience with beta-testing Litter Power and > porting another dozen or so externs is that I have had zero issues due > to Rosetta vs. real PPC.
      I just read on macfixit.com that there may be some maths problems with Rosetta introduced with X.4.8. -> http://www.macfixit.com/article.php?story=20061017090342511
      p
    • Oct 19 2006 | 1:56 pm
      On 19 Oct 2006, at 10:07, Peter Castine wrote:
      > My experience with beta-testing Litter Power and porting another > dozen or so externs is that I have had zero issues due to Rosetta > vs. real PPC.
      As I recall, Rosetta doesn't emulate Altivec.
      -- N.
      nick rothwell -- composition, systems, performance -- http:// www.cassiel.com
    • Oct 19 2006 | 9:17 pm
      On 19-Oct-2006, at 11:45, Patrick Delges wrote: > -> http://www.macfixit.com/article.php?story=20061017090342511
      Do you have a reference that doesn't require a paid subscription?
      If the problems are as obscure as the Pentium math problems of yesteryear (ca. 1995), you probably won't notice them in cs10format.
      Like I said, no guarantees. But I've not had problems.
      On 19-Oct-2006, at 15:56, Nick Rothwell wrote: > As I recall, Rosetta doesn't emulate Altivec.
      Possibly. I didn't notice any Altivec in the source for registry:-)
      --
      As a starting point, may I suggest the following (freely available) link?
      Yes, there is software that doesn't run properly on Rosetta. But look at what that software is: mostly stuff coding directly to hardware or very low-level system functions: DiskWarrior, drivers, games (low- level graphics). Look at what percentage of the software has problems. Suggest also looking at the sort of issues named. If those issues effect your external, then be warned.
      Patrick: if you want to have a real PPC Mac for testing, I've got a Lombard that will be going up for sale soon.
      Cheers
      -------------- http://www.bek.no/~pcastine/Litter/ ------------- Peter Castine +--> Litter Power & Litter Bundle for Jitter Universal Binaries on the way iCE: Sequencing, Recording & Interface Building for |home | chez nous| Max/MSP Extremely cool |bei uns | i nostri| http://www.dspaudio.com/ http://www.castine.de
    • Oct 20 2006 | 8:01 am
      On 19 oct. 06, at 23:17, Peter Castine wrote:
      > On 19-Oct-2006, at 11:45, Patrick Delges wrote: >> -> http://www.macfixit.com/article.php?story=20061017090342511 > > Do you have a reference that doesn't require a paid subscription?
      Ooopps. No. I don't pay neither, there is free access to these articles only during a couple of days after publishing. There weren't many details, only the fact that 4D users were asked not to upgrade to X48 because of some rounding issues... If you google with "rosetta" and "4d", you'll find some articles.
      > If the problems are as obscure as the Pentium math problems of > yesteryear (ca. 1995), you probably won't notice them in cs10format.
      ;-) Don't tell me you use cs10format!
      > Patrick: if you want to have a real PPC Mac for testing, I've got a > Lombard that will be going up for sale soon.
      We have enough PPC Macs and clones (and even pre PPC) here, but only 1 intel iMac, the administrator's one, that I used with Max4.6beta demo early in the morning to port sfmarkers~ to UB. As the demo expired, I can't port my amazing cs10 externals family to UB.
      p
      _____________________________ Patrick Delges
      Centre de Recherches et de Formation Musicales de Wallonie asbl http://users.skynet.be/crfmw/max
    • Oct 20 2006 | 1:07 pm
      On 19 Oct 2006, at 22:17, Peter Castine wrote:
      > Possibly. I didn't notice any Altivec in the source for registry:-)
      It didn't seem worth trying to look up lists and symbols using the DSP chain. One can be *too* performance-obsessed...
      -- N.
      nick rothwell -- composition, systems, performance -- http:// www.cassiel.com