Hi Kunal,
as you probably have guessed, the answer is "it depends".
Usually it's not very hard to port things; just make a new module using gr_modtool, make
the same blocks as you had before, and copy over the core functionality from your existing
blocks – in many cases, that's just the content of the constructor and the "work"
function. Many blocks have zero custom code aside from that!
However, some do. Most blocks don't make extensive use of the finer parts of PMT (which
has gotten better API lately) etc. Some do!
Also, we don't know the quality of the code you'll be porting. Do you know the feeling
when you sit down with some code and are unsure whether you're just not understanding
things correctly or the code has errors, and you start to look for the test cases, so you
know whether things you try are still correct? And then there's no test cases and you need
to write them yourself, and then you realize the original author never bothered to
actually verify whether what they do is correct? That might be the point where you decide
to rewrite from scratch, anyways.
So, really, you'll need to sit down and try with the code you have. It **probably** is
pretty easy to get ported (and seriously, people will be helpful if you ask questions
here!). Most GNU Radio OOT modules aren't badly written and well worth porting!
Best,
Marcus
On 18.08.24 05:11, KunalPrabhakar Sangurmath wrote:
Hi everyone,
I'm currently working on an Out-of-Tree (OOT) module that was originally developed for GNU
Radio 3.8. I need some advice on the best approach to port this module to GNU Radio 3.10:
1. Should I port the existing 3.8 version directly to 3.10, if yes, how or
2. Would it be better to start from scratch and implement the module in 3.10?
I'd appreciate your thoughts on which option might be more efficient and
reliable.
Thanks,
Kunal.
