Reptorian @ Reptorian @programming.dev

Posts

0

Comments

49

Joined

2 yr. ago

I used it only as last resort. I verify it before using it. I only had used it for like .11% of my project. I would not recommend AI.

I'll admit I did used AI for code before, but here's the thing. I already coded for years, and I usually try everything before last resort things. And I find that approach works well. I rarely needed to go to the AI route. I used it like for .11% of my coding work, and I verified it through stress testing.

I never had a issue with loops. If I feel something is off, I terminate, check code, and fix it.

I just write in a language few people really knows well. Not like I expect AI to do a great job of basing off my code.

I only used LLM-generated code once. I tested and made modifications to see if it was I want. It worked out.

In G'MIC, there's repeat(num_of_iters,_var_name,code();); on JIT code, and repeat code_block done outside of JIT. It has while, for, dowhile on JIT too. Other than repeat, there is only do while, and for which is while outside of JIT.

Note: _var_name, can be omitted. So, if you need to just repeat a code N times, that can be removed.

Eh, I rather write code by hand no matter how long it takes.

I use KDE Kate for my coding. Scripting more accurately to some users, but I don't find a meaningful distinction.

I actually use repeat(iterations,index_name(optionall),);. No need for i++ or ++i for loop in many cases.

I had this problem only once. Luckily, I do not need recursion. I know my way using iterative process.

Not what I'm saying. Just saying if I need speed, I'd use a naturally faster language.

Numpy is good for that. But, I gravitate toward naturally faster langs where I don't need to import a library for speed. I use Python for simple, dumb scripts.

Nah. I only used AI as a last resort, and in my case, it has worked out. I cannot see myself using AI for codes again.

Sometimes, I just rewrite my code until it is good enough. Other times, I leave it to my memory, so I can figure it out later. And others, I'm just not happy about it, like the times I did bigbin2dec and it would only work well with something like thread-ripper.

It means that I made changes to code, so that it can work with any multi-threading strategy. G'MIC is a interpretative language with JIT support, and you don't have any control over automated multi-threading strategy. It can be thread 0 to thread N linearly, or even interleaved. So, the workaround is to make a image of size equal to cpus count, and do the multithreading there.

Yeah, I love image processing, I worked on it for 8+ years now.

I use it for scripting too. I don't need Python as much as before nowaday.

As always, I do image processing. I'm a G'MIC filter developer. Recently, did some code changes to my combinatorics tools to be insensitive to multi-threaded strategy.

Uh, that would be infuriating to see. (Yes, I can see tabs in KDE Kate)

What if you use tabs for indentations and space for alignment?

What if you prefer a DSL? I use G'MIC for image processing tasks.