2008/8/10 Piero B. Contezini

The lastest version support asio or what you want is someone to code support for it? If you just want to build a version that already has support, it is quite easy! I dont have a windows box but perhaps i can build on my virtual machine.

Well, ChucK uses RTAudio and RTAudio can be switched to talk ASIO instead of Windows media/ DX/thingy drivers. As I understand it flipping this switch, then building is quite straightforward. That's what I'd like. This would get us a "chuck.exe" identical to the latest canonical one except for doing ASIO. This has worked fine for the past three versions of ChucK but somebody needs to build one, the official distribution still doesn't. Here's Philippe's note on he process from the dev-list; *BTW, i am running Chuck on Asio since a few days. It's on a Motu 828mkII. It's quite fast on the real time latency, but a little bit harder on the cpu. There is no code change for the compile. Just add an Asio folder with the Asio The asio files can be downloaded from the rtaudio website: http://www.music.mcgill.ca/~gary/rtaudio/index.html#downloadhttp://www.music.mcgill.ca/%7Egary/rtaudio/index.html#download Please know that I am not sure at all if this create bugs or not, but it's a start I guess. files as described on the RTaudio page, add __WINDOWS_ASIO__ to the preprocessor definition and you are set. * I would do it myself but my XP install is tweaked so far that I don't dare install larger MS aplications like their free version of their C++ compiler, which will in turn need SP2 and heaven knows what. Actually, I gather the newer versions of RAudio should also support building applications that could use both types of audio drivers but clearly that would need some infrastructure to select which one to use. Yours, Kas.

Moudi, Sorry to hear about your lost HD, but great to hear that you had backups and a ASIO build! Looking forward to hearing from you and thanks for your reply. Yours, Kas.

i already "rolled" an ASIO version with the new chuck code. but then my hard disk crashed. fortunately i was able to backup the whole thing in advance to another disk, but i have no working desktop atm and can't acess this data. will try to set up the machine again asap and bring the new ASIO release to public.

Hi. Steve, I'm sorry but I'm having trouble with it. Attempting to run this from the prompt (or probing, or sending it files) gives me a error saying "the system cannot execute the specified program". Double clicking gives me "This application has failed to start because the application configuration is incorrect. Reinstalling the application may fix this problem." My soundcard does both ASIO and DX, at the same time if need be, so it may be in the ASIO, it could also be me being on XP and you on Vista? Yours, Kas. 2008/8/11 Stephen Sinclair

Hey, I posted an ASIO build to the forum. Let me know if it works properly since I couldn't really test it.

Steve

Hey, Steve, Do you have IE7 installed? Just out of curiosity...

That's a good question. I looked and it turns out I had kept IE6 around, probably in case I ran into HTML help files, but definitely no 7.

I've had that error before with programs compiled on VS 2005. I ran "depends" on the executable and it shows that there's a flagged dependency on IEFRAME.DLL. I've determined on some previous projects that sometimes compiling with IE7 installed makes programs not work on computers that don't have IE7 installed. For reasons, of course, that have absolutely nothing to do with web browsing.

In other words, I think it might be another instance of this little dung nugget:

http://www.music.mcgill.ca/~sinclair/content/blog/fun_times_in_microsoft-lan...http://www.music.mcgill.ca/%7Esinclair/content/blog/fun_times_in_microsoft-l...

At this point I think I can't even express my disdain for MS developer tools.

I agree with your choice to keep things polite and civilised here and completely identify with how hard this is in this case.

Anyways, I guess it'd be best to get it compiling with MingW. In the mean time I'll try compiling it on an XP computer, since I'm at work and have access to one.

I just tried the new XP version that you posted on the forum but get the same error. As I understand the link you gave above some programs compiled under XP in this way on a computer that has IE7 (which I imagine your XP box does?) can then only run on Vista, is that correct? In that case this is at least a predictable error, which I suppose is at least of some help. It's a good thing I'm a polite guy because otherwise I might be tempted to make some medical diagnosis about who-ever thought of this, something I'm not qualified to do, then sugest a method of treatment that I believe would be both unconventional and illegal. I'd like to sugest we document these findings in the docs for compiling the Win version, hopefully this would avoid somebody else running into this in the future. This sounds quite ugly to me and I feel it would be better if people that are on Windows exclusively, yet not as well-versed in these matters as you are, could have a go at editing the source as well. To me large sections (the STK comes to mind) look quite readable but issues like this would take all the joy out of poking around. Lots of thanks for taking the time to take this this far already, Yours, Kas.

On Mon, Aug 11, 2008 at 9:28 PM, Stephen Sinclair wrote:

On Mon, Aug 11, 2008 at 7:22 PM, Kassen wrote:

...

I just tried the new XP version that you posted on the forum but get the same error. As I understand the link you gave above some programs compiled under XP in this way on a computer that has IE7 (which I imagine your XP box does?) can then only run on Vista, is that correct? In that case this is at least a predictable error, which I suppose is at least of some help.

Yup. I probably won't get a chance tomorrow, but I have a Windows box at school without IE7 installed so later this week I can try on that one. Hm, come to think of it I may even be able to do it over ssh. (cygwin yay)

And it worked! I managed to build it using the VS2003 environment from the command line. Posted on the forum. Let me know if it works. Steve

On Tue, Aug 12, 2008 at 8:09 AM, Kassen wrote:

Steve;

...

And it worked! I managed to build it using the VS2003 environment from the command line. Posted on the forum. Let me know if it works.

Yes! bingo!

It runs the otf examples, including the ones that use samples and --probe gives me all my in and outputs so that means ASIO (ChucK on DX drivers will only do stereo in&out). Lovely.

Great! Good to hear it.

I did run into the same issue I had with Moudi's last ASIO build, being that upon shutdown of the program using ctrl-c it would get stuck at "cleaning up..." for a few seconds, then display a error. Switching "compatibility mode" (found in "properties" if you right click on the exe) to "Windows NT 4.0 (service pack 5)" or "Windows 2000" fixes this. That could just be due to me tweaking, disabling and deleting so much. Perhaps somebody could confirm?

Hm, maybe it's having trouble shutting down the ASIO audio system? I haven't used ASIO much so I might not be much help here. Steve

Steve, Okay, good to know. Very silly, it shouldn't even matter.

I agree, but that's the situation we're in. I'd be all in favour of Windows disappearing from the planet but right now we are kinda stuck with it for some things so let's make the best of it.

Heheh.. it seriously blew my mind when I realized they'd introduced a bug that means that if you have IE7 installed when you compile completely unrelated software, you must have IE7 or Vista in order to run it. Totally insane. You can't make this stuff up.

I think to some degree people tend to assume others are "like them". For MS people it's probably self evident that everybody has the latest version of IE. Some time ago I noticed that my Gnome interface on my Ubuntu box depends on a CD burning program... which is quite annoying if you're a bit of a minimalist and don't have a CD burner on that box.

Documentation is good, and in fact the mail archives even work quite well for that. However, I think the real answer would be to get it going under MingW. MSYS can be a bit of a pain to get set up, but I've found it quite reliable once you've got it working. One idea that came to mind is to fork msysGit:

http://code.google.com/p/msysgit/

I'm a fan of this project because in addition to being a great source code management program, it provides a single installer that unpacks itself wherever you want with a complete MSYS/MingW environment ready for hacking. Tweaking it to include everything needed to compile ChucK wouldn't be all that hard. I'll try to look into this when I get a chance.

That does sound good, yes.

Another thing I'd like to do sometime is create a tutorial on how to make a ugen. Looking at the source, it seems really non-obvious, so I think having some explicit help in this direction would probably get more people interested.

I think a fair amount of people are already interested in that , yes, but indeed it does look hard. I can edit existing Ugens and expect it to work in a few tries but new ones would hook into the system in so many places that it gets a bit scary. I think there are others that think about it that way. For example; on the forum there has been a lot of talk about overdrives and there are some nice designs, partially homebrew, partially inspired by SC ones so that could be a nice place to start.

No problem. I'd really like to see more people collaborating here, so I'm all for it.

Me too. I have long held that docs on how to build Ugens would be good for the comunity as that would enable more people to add to ChucK, the same would go for includes so we could have libaries. ChucK is working great as a intro for programming and the syntax is close enough to C to understand -for example- what goes on inside of a Ugen but ChucK does not prepare you for the full complexity of something the size of the ChucK code itself so now there's a group of people "stuck" there. This is a area where I think we can learn a lot from our friends over at the SC community. SC is of course larger and older so that's a factor int heir user-contributed Ugens but I also think the process they have for adding them is a big factor. If you would take the time to write this documentation I'd be delighted to test it and give feedback. Yours, Kas.