4 ms·

PyLucid – A Lucid Interpreter Written in Python

43 points by herodotus 7y ago

mindcrime 7y ago
Looks like this may be the corresponding Github repo? https://github.com/billwadge/pyLucid https://github.com/billwadge/pyLucid
alexchamberlain 7y ago
There are at least 2 other projects with the name PyLucid: - https://github.com/jedie/PyLucid https://github.com/jedie/PyLucid - https://github.com/yelantingfeng/pyLucid https://github.com/yelantingfeng/pyLucid You might want to consider other names
- FisDugthop 7y ago
  What would you call a reference interpreter for the Lucid programming language? I understand your point, but nomenclature is not somehow first-come-first-serve; there are such things as better and worse uses for names.
- johnisgood 7y ago
  Someone just named their library in Python or whatever that was "Assembly".
gnufx 7y ago
I don't remember how Lucid compares, but the SISAL heritage probably needs reviving too. (Dataflow work is happening again at Manchester -- one of the dataflow hardware and SISAL sites -- via OpenStream for parallel programming.) Concerning POP, I think it lives to some extent in Poplog, though I don't know how alive that currently is. Correction: I realize the article refers to POP-2, and POP-11 is in Poplog. I actually rescued an ancient POP-2 manual from the bin recently, but I don't know how POP-11 compares.
klik99 7y ago
Fascinating - after several years of designing and a year of implementation, I'm running a closed beta of an LLVM-compiled language (general purpose but made for audio and dsp) that was strongly influenced by Lucid. For performance reasons it's not strictly "push" dataflow like Lucid, rather it pulls the output. It's a mix of dataflow from Lucid and type system from Haskell, aimed towards being performant and practical. https://www.elasticaudio.com/ https://www.elasticaudio.com/ Outside of monitoring systems and hardware specifications, dataflow has been sorely overlooked, and Lucid is truly mind-bending. I could see dataflow becoming more relevant as Moore's law and performance/cost benefits start scaling horizontally, while functional languages are seeing more practical use. Potential uses include Audio, obviously, but UI, graphics, and anything that involves pipelines calling to outside environments. EDIT I see PyLucid implements what Wadge calls "demand-driven dataflow" which I think is equivalent to "pull dataflow", though the ideal Lucid implements "eager dataflow" (I believe equivalent to my term "push dataflow")

43 points by herodotus 7y ago

Looks like this may be the corresponding Github repo? https://github.com/billwadge/pyLucid https://github.com/billwadge/pyLucid

alexchamberlain 7y ago

There are at least 2 other projects with the name PyLucid: - https://github.com/jedie/PyLucid https://github.com/jedie/PyLucid - https://github.com/yelantingfeng/pyLucid https://github.com/yelantingfeng/pyLucid You might want to consider other names

FisDugthop 7y ago
What would you call a reference interpreter for the Lucid programming language? I understand your point, but nomenclature is not somehow first-come-first-serve; there are such things as better and worse uses for names.
johnisgood 7y ago
Someone just named their library in Python or whatever that was "Assembly".

gnufx 7y ago

I don't remember how Lucid compares, but the SISAL heritage probably needs reviving too. (Dataflow work is happening again at Manchester -- one of the dataflow hardware and SISAL sites -- via OpenStream for parallel programming.) Concerning POP, I think it lives to some extent in Poplog, though I don't know how alive that currently is. Correction: I realize the article refers to POP-2, and POP-11 is in Poplog. I actually rescued an ancient POP-2 manual from the bin recently, but I don't know how POP-11 compares.

klik99 7y ago

Fascinating - after several years of designing and a year of implementation, I'm running a closed beta of an LLVM-compiled language (general purpose but made for audio and dsp) that was strongly influenced by Lucid. For performance reasons it's not strictly "push" dataflow like Lucid, rather it pulls the output. It's a mix of dataflow from Lucid and type system from Haskell, aimed towards being performant and practical. https://www.elasticaudio.com/ https://www.elasticaudio.com/ Outside of monitoring systems and hardware specifications, dataflow has been sorely overlooked, and Lucid is truly mind-bending. I could see dataflow becoming more relevant as Moore's law and performance/cost benefits start scaling horizontally, while functional languages are seeing more practical use. Potential uses include Audio, obviously, but UI, graphics, and anything that involves pipelines calling to outside environments. EDIT I see PyLucid implements what Wadge calls "demand-driven dataflow" which I think is equivalent to "pull dataflow", though the ideal Lucid implements "eager dataflow" (I believe equivalent to my term "push dataflow")