CAL - the word that dare not speak its name. Noel, what's the story here, please?

It was announced (I think when Sonar 8 was released) that CAL is no longer supported/deprecated - though it continues to work. So I'd guess you're right when you say

Cake would prefer that CAL simply fade away

.
 
Documentation of CAL has always been thin - I think there was a downloadable chapter in the Sonar Power 7 that covered some basics but basically suggested you learn by analysing existing CALs

Well, thanks Noel. I have to say I am always impressed that you respond personally to so many of these posts. It's a clear commitment to an excellent product.

Having said that, I can't help but feel somewhat disappointed here. I was going to make a very serious effort - as a professional software developer - to see if I could find a solution to the problem I've posted elsewhere - deriving a tempo reference automatically for 'fit improvisation' - by analysing the MIDI events in a track and algorithmically determining where the bar lines are likely to lie. (and I would have donated the result, if successful, to the community, I'm just doing this in the context of my own interest as an amateur musician).

In other words instead of having to laboriously tap in a tempo reference on a new track and do 'fit to improvisation', the script would have created the tempo reference events based on an analysis of the MIDI events in the improvised free-tempo track. For anyone like myself who doesn't like the constraints of metronomes, this would be extremely useful, I would think.

I have a couple of algorithms which I believe will work for the 4/4 case, which is by far the most useful - and from what I have seen of CAL it is, I think, capable of doing what I want, in that you can programmatically insert notes into tracks. All I have to do is determine exactly what event time to give each inserted event.

So I *think* CAL would have sufficed, but as I said, I'm very reluctant to expend effort on something which may just vanish in the near future, or where functions I use are not available in a reasonable range of historical releases, say, S6-8.

So, yes, I am disappointed. I think this decision regarding CAL is very regrettable, because CAL at least offers some third-party extendability to Sonar.

I could of course alternatively consider whether an MFX plugin COULD be made to do the job - but going back and looking, this seems to be bound up with the DXi SDK and I think DXi is really a dead technology as well now, so is this necessarily a good decision either - at some stage, you're presumably going to drop DXi support and focus on VSTi, since, as far as I know, you're now pretty much the only vendor supporting DXi. Or Microsoft will drop support since the technology never gained any traction. In any event the number of DXi developers must be vanishingly small these days, so if I go down this path and get stuck, I'm not likely to be able to find any technical assistance. And based on my previous experiences developing a control surface plugin, the documentation is fairly spartan for these SDKs, though I do appreciate that you make them available to developers in the first place.

I urge you to reconsider the historical decision (which I imagine perhaps you may not even have been involved in) to sunset CAL. I appreciate that it places a small burden of extra testing and compatibility issues on new releases, but absent any other kind of scripting support - unless you plan to add it in a future release - you are significantly limiting the extensibility of the product by third parties.

I don't consider Reaper to be a serious competitor to Sonar in the marketplace you mainly address, but it does have some remarkable extendability through both scripting support (in Perl and Python), and, equally remarkably, the ability to write new audio plugins (Jesusonic) that run in real time using a simple C-like language - you can create your own reverb, limiters, echo, pitch change and all sorts of things in a few lines of code without a complicated SDK to grapple with, which has resulted in a growing number of interesting third-party plugins written by people who have some programming skills but don't want to grapple with the complexities associated with a full-blown VST plugin.

This third-party ecosystem benefits all users and costs Cockos nothing, other than the presumably acceptable cost of supporting scripting in the first place.

Anyway, that's my 10 cents worth. I don't know what other forum members think. Many will never even have heard of CAL, I guess, some will still use the odd CAL script but whether the ability to extend Sonar's MIDI functionality via scripting is of interest to people other than myself, I can only conjecture. But it does seem an odd decision to support running CAL  scripts on the menu even in the latest Sonar versions, yet deprecate it without any obvious replacement. Normally, when a software manufacturer deprecates something, a replacement is already available. In Sonar's case, that's clearly not the situation. I think this is a darn shame.

I'm a long-time CAL fan, too. I'm still using scripts that I wrote back in 1989. I would be tickled to have CAL make a comeback, be properly documented and even enhanced.

Unfortunately, I don't think that's going to happen, and frankly if it were my decision I wouldn't be putting development time into it either. It's just not the kind of feature that sells DAWs nowadays. Today it's all about canned shortcuts - assembling loops, pasting pre-made MIDI patterns, stuff like that. We tweakers are a diminishing fraction of the market.

Eh Bit, I think your right about this. If we included split notes to tracks on the process menu we could probably dispense with about 90% of the cal requirement these days. It was mostly about doing clever things with midi wasn't it? Short of running a few cals from time to time I havent written anything in it for years. Back in the day maybe Greg say it as a quick way to provide all the midi manipulation features that people asked for.

Amazed


Eh Bit, I think your right about this. If we included split notes to tracks on the process menu we could probably dispense with about 90% of the cal requirement these days. It was mostly about doing clever things with midi wasn't it? Short of running a few cals from time to time I havent written anything in it for years. Back in the day maybe Greg say it as a quick way to provide all the midi manipulation features that people asked for.

There was a time when, as far as I know, every feature in Cakewalk Pro Audio could be included in a CAL script. Creating a CAL script was as easy as recording your moves and saving that as a script.

Haaa the glory days of Cakewalk. ;)

UnderTow

bitflipper


I'm a long-time CAL fan, too. I'm still using scripts that I wrote back in 1989. I would be tickled to have CAL make a comeback, be properly documented and even enhanced.

Unfortunately, I don't think that's going to happen, and frankly if it were my decision I wouldn't be putting development time into it either. It's just not the kind of feature that sells DAWs nowadays. Today it's all about canned shortcuts - assembling loops, pasting pre-made MIDI patterns, stuff like that. We tweakers are a diminishing fraction of the market.

I fall into the canned group but I don't apologise. I really don't have a lot of midi needs that cakewalk doesn't cover. But there are some and using CAL scripts have been the answer more than once. With that said, I was a long time Logic user and just about everything I wanted for MIDI manipulation was there in Logic's menus. I have more than once in the past ported Sonar's midi files to Logic for MIDI editing. To wrapped this up - I need certain functions that CAL does easily and would like it very much if some of those were part of Sonar's workflow. All at the same time - I would like for Cakewalk to always support scripting of some sort. Many more possibilities will be available for us non-scripters when people like you create and share. So Cake - please keep scripting available!!!!!

It's just not the kind of feature that sells DAWs nowadays. Today it's all about canned shortcuts - assembling loops, pasting pre-made MIDI patterns, stuff like that. We tweakers are a diminishing fraction of the market.

True enough. But even the people who make music like stringing beads could use a good scripting/macro capability. Maybe they could use it more. It is fine to deep six CAL, but you can not possibly replace the flexibility of scripting/macros with a gazillion icon-linked commands and a matrix/step sequencer.

aj I believe you could do what you are trying to do by writing a generative MIDI out VSTi.

bitflipper
We tweakers are a diminishing fraction of the market. 

This is an over-generalization and I will have no part of it. :P

quote

aj I believe you could do what you are trying to do by writing a generative MIDI out VSTi.

unquote

Noel, thanks, that is an interesting idea. I hadn't thought of a VST that creates/modifies MIDI events, its a strange yet possibly plausible idea. It would be an odd beast, since it will presumably be inserted into one track and output MIDI events into a second track. I'll have to ponder this a bit, I think. But many thanks for the suggestion.

Meanwhile, I'm going to experiment with the idea on, er, another platform (sorry), temporarily, where I can write a script in Perl, a language I am comfortable with, hack around with various algorithms, and then when (if) I have the concept working I will look at turning it back into C code, I think. Much as I admire C/C++ it is not a language for quick and dirty experimentation, to say the least. Perl has much more powerful constructs such as associative arrays and inbuilt sorting functions that in C require you to call into complex libraries to achieve the same thing. And CPAN has the world's biggest software library of pre-written solutions for Perl. If you want to know what day Easter Sunday falls on, there's a CPAN library for that. And so on. A great time-saver. If you ever WERE going to replace CAL with another scripting language, Perl or Python would certainly be my preferred choice. I just feel odd about Python because the idea that whitespace has semantics is just, well, kinda weird. But I know many people love it to bits. Perl, well, it's ugly but boy does it get the job done. As someone once said, its the Swiss Army Chainsaw of programming languages.

An interesting VST that I've been watching, but never used is VstLua.  However, 1) it's beta, and 2) it says it might not work in SONAR (though that may have been with an older SONAR version).

Anyway the web page for it is: http://t-zec.org/vstlua

Perhaps some adventurous programmers out there will want to investigate for possible use in SONAR.

Dickie Fredericks


If we could move a few CAL scripts to the Process menu where we could choose which ones Id only need to put maybe 2-3 there. That would be cool.

I often use Split Notes to Tracks, Undupe and sometimes StrumIt

I don't know about adding them to the Process menu, but you can make keybinding shortcuts for them.  That might help a little.

I still use CAL on occasion.  If it isn't going to be enhanced, at least I hope that future versions of SONAR continue to support its use.

Willy you are a jewel.  I can't thank you enough for the stuff you do to help us out.

CAL is a great asset for us. It needs to be supported better and developed more.  Thank you.