BlitzMax trans compiler, in theory

Monkey Archive Forums/Monkey Discussion/BlitzMax trans compiler, in theory

AdamRedwoods(Posted 2012) [#1]
Has anyone thought of hacking the monkey trans compiler to create an entirely new BMK-- BlitzMax to C trans compiler?

The implications would be that all BlitzMax code would then be compilable under the Monkey trans, which would then allow BlitzMax64? This may also eliminate BlitzMax.

The target, of course, would be limited to pc,mac,linux.

Too dangerous? A lawsuit waiting to happen?


Samah(Posted 2012) [#2]
It would be easier to just port your BlitzMax code to Monkey...


AdamRedwoods(Posted 2012) [#3]
It would be easier to just port your BlitzMax code to Monkey...

why do it the easy way when there's the hard way? :O)

some people were asking for bmax64. i figured this would be a push for them to all switch to monkey.


ziggy(Posted 2012) [#4]
I think it would be a much more interesting (and heavy) project to port the blitzmax BRL modules to Monkey code based on the C++ or java target. This would be enormous work but it could attrack most current blitzmax users and potential new ones. Remember than current BRL max modules are now public domain.


Armitage1982(Posted 2012) [#5]
I think it would be a much more interesting (and heavy) project to port the blitzmax BRL modules to Monkey code based on the C++ or java target. This would be enormous work but it could attrack most current blitzmax users and potential new ones. Remember than current BRL max modules are now public domain.

Best think that could ever happen to BlitzMax & Monkey IMO :)

But I doubt it will happen since most Modules are linked to specific platform technologies and many require fixes and updates.

Writing an extensive documentation on how to write port and support new target would be much more valuable.


ziggy(Posted 2012) [#6]
But I doubt it will happen since most Modules are linked to specific platform technologies and many require fixes and updates.
Yes, it would be a titanic task, and not sure it would be possible, but making the transition based on a Java target could help getting ride of the platform speciffic code. Obviously all the graphics part will have to be modified to new drivers designed for the Java VM, etc. If we were a bigger community, it would be an interesting community project, and work could be done in a "by module" basis so it would be very unit-testing prone. That said, I don't think it is viable unless a big team works lots of hours a day.


Armitage1982(Posted 2012) [#7]
I'm currently playing with Java and it's true that most packages could be very useful like RessourceBundle for localization, Swing + Substance or JavaFX for UI, JDOM for XML, plus lots of documentations, tools, tutorial, etc.

Add something performant, heavily cross-platform and games centric like LibGDX and I'm sure you get back most BlitzMax users in Monkey.

It's nearly the same as porting most BlitzMax modules to Monkey.

The real question you will have to answer before starting something that huge is:
Why should I wait something like this to happen in Monkey as well as learning this new language when I could start using Java right now ?

Java is in the making since decade, it's stable, professionally used, the Eclipse IDE is crossplatform plus if you are using Monkey to write Java code then you won't outmatch its performances (it may even be the opposite actually). The syntax is also very similar to Java.

Yeah, I think a good communication would be way better to attract new users than all this hard work.
I choose Monkey because I though it would be the next BlitzMax but in fact this is not its purpose. Everything will stay lightweight in order to write the same code and output to multiple target easily.

BlitzMax is still very good, fast and complete! So why not continue to improve it and keep monkey for small mobile devices? I would not say that BlitzMax is over but many left because they wanted a bit more or was only interested by the programming language in itself (and so, certainly switch to Monkey because of the nice new features and they don't mind if something is missing because they can write it again and again). Is it hard to believe that in 2012 people are still interested to develop for Windows/Linux/Mac only?