LinuxMusicians

Hello,

I got recently interested on Algorscore: http://kymatica.com/Software/AlgoScore, seems an interesting SW.

I tried to build it without success (on Mint 18.3). After some tweaks to go on with the compilation, the major show-stopper I've seen (if I understood correctly from the build output) is that the csound API that the SW uses (2007-2008) to make the realtime audio is no more aligned with the latest API (csound 6).

Anyone in the list was able to build it recently from the source?

Also the mailing list, support etc refer to gna,org which seems down.

Alberto

Hi there! Interesting software. From where you got the code? All links seem dead to me.

CrocoDuck wrote:Hi there! Interesting software. From where you got the code? All links seem dead to me.

Good question

I downloaded some time ago, did the search today and the only useful source I'm able to find is (sic !):
http://download.cnet.com/AlgoScore-for- ... 18290.html

Apart from the huge load of advertisemnt that you'll get there you'll be able to download AlgoScore-081112.tar.bz2
I know, this is not the way to get open source sw.

I this I started from there...

The author has plan to develop algoscore 2 in Supercollider, but the project is dead (or, let's say, very slowly progressing) in his github repository.

Alberto

Uhm... I think you would be best contacting the author and query about the status of the project. Not a good idea to download random code from not trusted sources. Maybe he will give you the code and clarify what dependencies you need to track in order to build it, or even if it can work at all on up to date Linux environments.

Hi there, I had a full spare afternoon and I attacked the problem of building Algoscore again.
My driver is just "curiosity". I finally was been able to build it with some editing of Cmake stuff (read: nightmare).
In practise I disabled the CSound output (the program is strictly "linked" with CSound 5.X and
I was not able to build it on my system), placed some configuration files of GTK2 in appropriate places and few other
minor modifications.
I got the following nice screenshot (no idea of what it represents, anyway ).
I took some notes of the process, if someone is interested I'll share them.

This makes me seriously think about how weak is the SW portability in the GNU/Linux world. This program has all the necessary requirements
for portability (i.e. open-source, built against open source components), but, as of now, without intricate editing it is impossiible
to build and to enjoy it due to evolution of components themselves. Maybe something like Flatpak could save old software from being lost but up to now how many SW we lost? What do you think?

Alberto

This page has some useful info: http://kymatica.com/apps/algoscore. (Note that is different from the url in the OP) From that page:

NOTE: This is an abandoned project and included here only for historical reasons. I had lots of plans for AlgoScore2, which would be based on SuperCollider for sound synthesis (in real-time) instead of CSound, and probably written in SuperCollider itself. Then many years passed, and the original GNA.org project page has been lost, but I’m still hoping I’ll restart this project some day!

There is also a link to the source tar.bz2 on that page, which is probably better than cnet!

AlbertoZ wrote:Hi there, I had a full spare afternoon and I attacked the problem of building Algoscore again.
My driver is just "curiosity". I finally was been able to build it with some editing of Cmake stuff (read: nightmare).
In practise I disabled the CSound output (the program is strictly "linked" with CSound 5.X and
I was not able to build it on my system), placed some configuration files of GTK2 in appropriate places and few other
minor modifications.
I got the following nice screenshot (no idea of what it represents, anyway ).

Cool!

AlbertoZ wrote:This makes me seriously think about how weak is the SW portability in the GNU/Linux world. This program has all the necessary requirements for portability (i.e. open-source, built against open source components), but, as of now, without intricate editing it is impossiible to build and to enjoy it due to evolution of components themselves. Maybe something like Flatpak could save old software from being lost but up to now how many SW we lost? What do you think?

i think this has 2 sides: on the one hand, in the windows world you *may* have more luck successfully running ancient binaries. However, without the sources, it is impossible to assume ownership and adapt the software to new needs. While something like Flatpak can help keep old software running on newer systems, it cannot "keep it alive".

With the sources, however, if there is enough interest in a given piece of software, it can be brought back. It may require a non-trivial amount of work - "intricate editing" as you describe - but ideally that should be necessary only once, and then shared.

Indeed too often we do the 'intricate editing' without the 'sharing' - which is understandable (not blaming anyone!), but also a missed opportunity.

AlbertoZ wrote:I took some notes of the process, if someone is interested I'll share them.

I think that would be great! Perhaps we could put the project in version control somewhere and add your documentation and edits there, so the next person will have a better experience and the project can grow again. I'd be happy to put it up on my github if no-one else is interested. Let me know!

milo wrote:This page has some useful info: http://kymatica.com/apps/algoscore. (Note that is different from the url in the OP) From that page:

NOTE: This is an abandoned project and included here only for historical reasons. ...!

There is also a link to the source tar.bz2 on that page, which is probably better than cnet!

Hi milo, yes, I know about the page you linked and the fact that the project is dead.. But it looks terribly interesting at least from graphical, unconventional score notation POW! BTW, I started exactly from the same source archive.

raboof wrote:
i think this has 2 sides: on the one hand, in the windows world you *may* have more luck successfully running ancient binaries. However, without the sources, it is impossible to assume ownership and adapt the software to new needs. While something like Flatpak can help keep old software running on newer systems, it cannot "keep it alive".

With the sources, however, if there is enough interest in a given piece of software, it can be brought back. It may require a non-trivial amount of work - "intricate editing" as you describe - but ideally that should be necessary only once, and then shared.

Definitely agreed. The example of Surge is the first that comes to my mind..

raboof wrote: I think that would be great! Perhaps we could put the project in version control somewhere and add your documentation and edits there, so the next person will have a better experience and the project can grow again. I'd be happy to put it up on my github if no-one else is interested. Let me know!

Why not? Give me 1-2 days (to repeat the process and to organize better my notes), and I'll be back on this again.
Good to know that there is interest on it. Maybe we can put the Csound interface up and running again?

Alberto

Hello again, attached here is a small pdf describing the build process. I know that what I describe is all but optimal,
but bear in mind that I don't know the source nor the Build tool (CMake).
What I'd like to do is to add back the CSOUND support to hear how the scores sounds.
Ideas are welcome.
Good luck..

Alberto

[off topic] Why is is to hard to build an application/executable code from old source code? Why aren't compilers backwards compatible? It's weird that you have to replace "old" names of commands with new ones in certain text files without actually changing any real functionality. It'll keep people from developing because in a few years time you can throw your work in the bin. Why don't they keep the same name for the same procedures/commands from one C++/Cmake version to another?

Linuxmusician01 wrote:[off topic] Why is is to hard to build an application/executable code from old source code? ...

In one word? Evolution. This is where functions and interfaces may change: or you are able to recreate the same ecosystem in which the SW has been created or you will be in trouble. I anyway subscribe the comments of raboof:

raboof wrote: i think this has 2 sides: on the one hand, in the windows world you *may* have more luck successfully running ancient binaries. However, without the sources, it is impossible to assume ownership and adapt the software to new needs. While something like Flatpak can help keep old software running on newer systems, it cannot "keep it alive".With the sources, however, if there is enough interest in a given piece of software, it can be brought back. It may require a non-trivial amount of work - "intricate editing" as you describe - but ideally that should be necessary only once, and then shared.

In detail, here we have calls to functions of CSOUND 5.x that are no more present in the modern CSOUND 6.x: the issue is not the compiler or CMake, but evolution of libraries against which Algoscore is linked.

Alberto

Greetings !

I once wrote an article about AlgoScore:

https://www.linuxjournal.com/content/al ... ic-numbers

Very cool software. IMHO the better option would be to update the Algoscore codebase rather than struggle with outdated dependencies.

Best,

dp

I did some more digging. Seems AlgoScore at some point lived at sourceforge, then moved to Gna. Gna is discontinued, but I was able to recover some slightly newer sources and history from the dump at https://archive.org/details/Gna_code_hosting.

Converted them to git and shared them at https://github.com/raboof/AlgoScore . Docs are also back up at https://raboof.github.io/AlgoScore/algo ... anual.html

I was able to fix up the build, it even compiles against csound now and starts. I haven't tried actually doing anything yet, but I don't think it will produce output yet, because I had to remove a call to 'csoundSetOutputValueCallback' - this method is no longer available on csound 6 and I haven't figured out what its replacement is yet.

@AlbertoZ: thanks for the explanation.

davephillips wrote:Greetings !
I once wrote an article about AlgoScore:
https://www.linuxjournal.com/content/al ... ic-numbers
dp

Hi Dave, I got interested in AS because of your article

raboof wrote: I was able to fix up the build, it even compiles against csound now and starts. I haven't tried actually doing anything yet, but I don't think it will produce output yet, because I had to remove a call to 'csoundSetOutputValueCallback' - this method is no longer available on csound 6 and I haven't figured out what its replacement is yet.

Really good news! Great, Raboof.. I'll give a look ASAP..

Linuxmusician01 wrote:@AlbertoZ: thanks for the explanation.

You're welcome..