Re:Pydis – Redis clone in 250 lines of Python, for performan...
by noname120 on Sunday November 15, 2020 @01:03PM
(#25103235)
Fair point, here is a pull request that addresses my criticism: https://github.com/boramalper/pydis/pull/15
Re:Pydis – Redis clone in 250 lines of Python, for performan...
by pastage on Sunday November 15, 2020 @03:49PM
(#25104679)
> as long as the heavy work is done by calling native non-Python code.
Do not merge that. I once realized that grep is really fast, and replaced our custom solution with bash script, but I still claim that app is written in idiomatic PHP/python/ruby. This project has a lot of problems that it depends on C code is only a symptom. Noting all those bad things in a project Readme is a trap I fall into way to often it never works.
Re:Pydis – Redis clone in 250 lines of Python, for performan...
by smoyer on Sunday November 15, 2020 @03:15PM
(#25104389)
I don't think the tone was hostile and if you ignore any hint of tone you might hear, you have to admit that the title is really click-baity (yes I understand that you didn't post it). Regardless of the criticisms of the project, it's great that you're experimenting and I would be curious to know what's causing the reduction in performance ... is it only part of the program or is that the performance of the Python code being interpreted.
Re:Pydis – Redis clone in 250 lines of Python, for performan...
by boramalper on Monday November 16, 2020 @02:34AM
(#25109105)
> I would be curious to know what's causing the reduction in performance ... is it only part of the program or is that the performance of the Python code being interpreted.
Mostly the latter but also a bit of the former: I could have optimised it further — at the expense of idiomaticity — but I believe gains would be marginal. Other than that, Python is not that fast so you pay some penalties there too.
I am not the one who posted here, why so aggressive?