Products > Programming

Goto bad Spooktober tale: Counting goto in the Linux kernel

<< < (8/10) > >>

DC1MC:

--- Quote from: Siwastaja on October 31, 2023, 07:32:55 am ---...
We were taught in university how goto is bad because you can jump anywhere, including other functions but that obviously was a lie and tells a lot about the level of understanding in the academic world.

--- End quote ---

Why no one is discussing animore about  setjmp and longjmp in C  :-DD

Cheers,
DC1MC

DiTBho:

--- Quote from: Nominal Animal on October 30, 2023, 09:33:39 pm ---You'll understand how the discussion derailed.  Reply #3 was, intended or not, just yelling "Ha-ha! Look at that peg leg!"

--- End quote ---

Can you see how frustrating this forum is?

I said in order

* Linux is not a good reference. Is it? you confirmed it's not
* then I mentioned PPC because it's what I am working at the moment and it's happening right now that we have 5 months of regression, Linus himself is complaining
* so, Linux is not a good reference, not even for looking at how gotos are used
* at that point someone suggested that a big problem is documentation, and I confirmed it
* little anecdote about how the "cowboy to the console" approach to the wild west of Linux, can even create absurd situations where if nobody understands absolutely nothing about the code then it's a big problem for development. It was full of gotos, in many point it was abused, strange pointer arithmetic, inline assembly and all kinds of shenanigans, and to understand how the code worked, thousands of hours of work were needed, practically rewriting everything from scratch
* then I pointed out that if there are no precise regulations you end up having little projects that only stand up if there are a lot of people following them, just like Linux, or worse still, just like linux in my anecdote
* at that point I suggested looking elsewhere, pointing out that where there are regulations, things are much better both understandable and manageable, even by small teams
* I mentioned VxWorks, and the discussion derailed *HERE* when I pointed out Windiriver's approach, and note why it derailed: because some "troll" willfully misunderstood what I wrote in order to ridicule those who do competent work
And frankly it bothers me a lot that people who don't even have the competence to speak (for example about DO178, someone here called it "bullshit for trained monkeys"), and this is because they have never done anything mission critical, often allow themselves to start a reply with the "it's bullshit", as happened in the topic of cyclomatic complexity.

Even worse when what they call "bullshit" often has serious reasons behind it, sometimes reasons of an organizational and not just technical nature.

And this because "practice" means "organize people doing software" as important as "design software" as well as "find a resonable way to allocate resources to test software".

All points that have continued to be repeated for years, and you explain it and every damn time we always come back to the same points, over and over, which makes me realize that arguing in this forum is a total waste of my time.


So, sorry, but no.

Siwastaja:
Or, maybe you just take someone calling a practice "bullshit" (A) personally, (B) too seriously? I mean, it's not that strong of a word. It just means whoever said that does not believe the practice is good at all. More important is to look why they think the practice is bullshit. And don't get hurt on behalf of a process.

I don't remember seeing that "trained monkey" thing you refer to, and it's nasty language I admit that. I have never ever name-called professionals working on the regulated fields you mention, and have no reason to do that. If you read what I wrote, you will see it's the opposite; I'm happy I can create software more freely and still sleep well, not being on their shoes; and I'm happy I can still discuss what I think about safety critical software, but it seems you are not happy about me discussing it.

But whatever. I think your fundamental issue on this forum is in anger management and poor communication. You are not the only one, I can say from personal experience of having the exact same issue. But in reality, others are not as stupid as you may think. Give us a chance.

DiTBho:

--- Quote from: Siwastaja on November 01, 2023, 11:48:52 am ---But in reality, others are not as stupid as you may think. Give us a chance.

--- End quote ---

... well, about that, it's you who that, misunderstanding what I wrote, thought that we must be really *very stupid to accept*(1) - according to your interpretation -  that it is enough to make Chatgpt(1) write some random documentation in the code to pass a piece of code by the QA team, who is composed by professionists!

I didn't specify what it was "necessary condition, not sufficient", my fault there, I edited my post, but how many times have I repeated how things really work? How many? So, before starting with "I'm horrified", didn't you even have any doubts? no, because in your opinion we only do bullshit! It takes you very little to brand things as "bullshit" (like in the Cyclomatic topic, where I simply ignored some of your bad statements), and it costs me too much time to explain it to you, and then what? See, you didn't even get the concept!

Damn, this is the common pattern here in this forum(2), and it's really frustrating!

Among other things, if I have always underlined that developers and testers should never talk to each other except through specific documents, the Windriver's approach creates a shortcut that allows, through mandatory documentation, to better describe both the code and the various doubts about how to write and test it, which is brilliant "extension" because it reduces the numer of iteractions and helps people growing up "team building"!

It seems like something within everyone's reach, because everyone can start describing the code, obviously in a concise and targeted way; Windriver pays some professionists with skills in being concise and targeted way, and QA guys have the same skill, so here OpenSource guys have to learn to do it well, but it's a true benefit and it's feasible, if not mandatory, it should at least become a habit!


(1) precisely here I saw the offense to the entire category

Siwastaja:
I merely commented on your claim that AI cannot be used to cheat "because it's in natural language". Don't take it out of context. I totally agree with what you now write; that competency in management is what prevents such cheating. You(!!) are the one who mentioned cheating with AI, as if it would happen if the document were not in natural language! You were the one whose descriptions about these regulated fields made me horrified. You are the one giving bad picture about "your" field. You fail to communicate because you are arrogant and write quicker than you think when angry.

And while I trust a lot of safety critical software (otherwise, I would not board an airplane), I still have my doubts about how the arrogance you show affects your work.

Now can you do us all a favor and start actually ignoring people you claim to ignore and therefore stop derailing threads? Thank you in advance.

Navigation

[0] Message Index

[#] Next page

[*] Previous page

There was an error while thanking
Thanking...
Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod