Le 11-06-2024, candycanearter07 <candycanearter07@candycanearter07.nomail.afraid> a écrit :
> Lawrence D'Oliveiro <ldo@nz.invalid> wrote at 00:57 this Sunday (GMT):
>> On Sat, 08 Jun 2024 13:49:45 -0400, Joel wrote:
>>
>>> ... I just wonder what it would take to start from nothing.
>>
>> One way to find out: <https://www.linuxfromscratch.org/>.
>
>
> Isn't LFS somewhat difficult? You'd have to have some dedication to see
> it though to the end.

It's not that difficult. But you need plainty of time because it's very
long. And then, you'll have nothing usable (the reason behind BLFS). And
then it will take a lot of time for any upgrade.

Said differently, LFS if a good way to learn, not to manage your own
computer.

--
Si vous avez du temps à perdre :
https://scarpet42.gitlab.io

Le 08-06-2024, rbowman <bowman@montana.com> a écrit :
>
> So, what do you want to do? Of course you can do all of the above with C,
> C++, or another language but you may be writing more code and doing more
> debugging. If the Python prototype is too slow you can rewrite it.

Like one said:
- with C, you have a million ways to do one thing,
- with python, you have a million libraries to do one thing,
- with C++, you have a million ways to screw things up.

--
Si vous avez du temps à perdre :
https://scarpet42.gitlab.io

On Fri, 14 Jun 2024 09:45:29 -0400, DFS wrote:

> On 6/13/2024 6:16 PM, rbowman wrote:
>> On Thu, 13 Jun 2024 15:08:50 -0400, DFS wrote:
>>
>>> - malloc the md_count[] array to size 9 (because there
>>> are only 9 unique values in the size 10 dataset)
>>
>> What memory location is md_count[9]++ incrementing?
>
> 9 is the correct size for md_count[], but then you have to play games
> with index addressing and looping.
>
> So malloc() to size 10 and life's good.

Right. Did you play games with the index when you malloc'd 9?

>> For that matter what are the initial values of md_count[0] through
>> md_count[8] ?
>
> 0.

man malloc

The malloc() function allocates size bytes and returns a pointer to the
allocated memory. The memory is not initialized. If size is 0, then
malloc() returns either NULL, or a unique pointer value that can
later be successfully passed to free().

'The memory is not initialized.' is underlined on the man page. Some days
you get lucky.

Le 09-06-2024, Tyrone <none@none.none> a écrit :
> On Jun 8, 2024 at 7:42:16 PM EDT, "Joel" <joelcrump@gmail.com> wrote:
>
>> What purpose does coding that serve? I can code, I choose not to
>> because everything's already been coded, by someone else.
>
> What an extremely naive thing to say.

You seem surprised. Is it the first time you read a message from Joel?

--
Si vous avez du temps à perdre :
https://scarpet42.gitlab.io

Le 11-06-2024, candycanearter07 <candycanearter07@candycanearter07.nomail.afraid> a écrit :
> Tyrone <none@none.none> wrote at 02:32 this Sunday (GMT):
>>
>> Granted, all of the alleged "coding challenges" that Feeb presents here HAVE
>> already been coded. But that's what happens when you think that "All
>> programming is about solving trivial math problems".
>
> A lot of times, programming involves connecting "trivial problems in an
> advanced way".

Feeb is not DFS. Both are funny, but not in the same way. DFS keep his
pseudo. Feeb changes it a lot of times: from Fabian Russel to Diego
Garcia including Nux Vomica among others. DFS is a Windows fanboy who
knows about programming, Linux and Windows. Feeb knows nothing about
everything, he's a Windows fanboy pretending being a Linux fanboy who
tries to make believe every Linux fanboys are first class morons as him.

> Sorting a list of names would be needed for a dropdown,
> letter or word counts are usually used to limit how much text a user can
> input, and averages are useful for any kind of data summary.

And here, it's only an English forum. When you want to sort the names of
your CD albums, it can become tricky. For example how do you take care
of the letters with diacritics signs? How do you include Cyrillic
letters ? Or the Arabic or Jews letter which are written from right to
left? You can have a lot of fun with something which looks like a simple
problem.

>> In the real world, very few problems have "already been coded". Which is why
>> people who proclaim that "AI" is going to make programmers "obsolete" have no
>> GD clue what they are talking about.
>
> Yeah, AI is very very annoying and not a good replacement for coders.

No, AI is not boring, what's done with it is impressive. Of course, if
you try to do things for which it was not designed, it wont work. It's
like trying to cross the ocean with a car: it will drown. It doesn't
mean the cars are useless, it means you need to use them for what they
were designed.

It's the same with AI. For now, they can't replace programmers, it
doesn't mean they are useless. For things like translation, they are
impressive. For programmers, they can be used as help, for example if
you don't remember an algorithm. They can't correct a bug, they can
program some little scripts.

--
Si vous avez du temps à perdre :
https://scarpet42.gitlab.io

On 6/14/2024 12:46 PM, rbowman wrote:
> On Fri, 14 Jun 2024 09:45:29 -0400, DFS wrote:
>
>> On 6/13/2024 6:16 PM, rbowman wrote:
>>> On Thu, 13 Jun 2024 15:08:50 -0400, DFS wrote:
>>>
>>>> - malloc the md_count[] array to size 9 (because there
>>>> are only 9 unique values in the size 10 dataset)
>>>
>>> What memory location is md_count[9]++ incrementing?
>>
>> 9 is the correct size for md_count[], but then you have to play games
>> with index addressing and looping.
>>
>> So malloc() to size 10 and life's good.
>
> Right. Did you play games with the index when you malloc'd 9?

I did not. So I got a little lucky.

Kind of related: Here's how python checks array bound every time the
array is accessed:

int
PyList_SetItem(PyObject *op, Py_ssize_t i,
PyObject *newitem)
{ PyObject **p;
if (!PyList_Check(op)) {
Py_XDECREF(newitem);
PyErr_BadInternalCall();
return -1;
}
if (!valid_index(i, Py_SIZE(op))) {
Py_XDECREF(newitem);
PyErr_SetString(PyExc_IndexError,
"list assignment index out of range");
return -1;
}
p = ((PyListObject *)op) -> ob_item + i;
Py_XSETREF(*p, newitem);
return 0;
}

Python-3.11.0.tgz.2\Python-3.11.0\Objects\listobject.c
line 257

Which is another thing that contributes to its slowness vs C.

>>> For that matter what are the initial values of md_count[0] through
>>> md_count[8] ?
>>
>> 0.

Lucky.

> man malloc
>
> The malloc() function allocates size bytes and returns a pointer to the
> allocated memory. The memory is not initialized. If size is 0, then
> malloc() returns either NULL, or a unique pointer value that can
> later be successfully passed to free().
>
> 'The memory is not initialized.' is underlined on the man page. Some days
> you get lucky.

Yep. In the big stats program I posted I messed around with using a
malloced array of 20,000. It ran a bunch of times just fine, but all of
a sudden I got nonsense results.

So I filled it with 0s before using it, and no problem.

But calloc saves a line of code, so that's what I used.

On geeksforgeeks a calloc() example is:

int* ptr;
ptr = (int*)calloc(n, sizeof(int));

Isn't the 2nd int* redundant?

I just used:

int *ptr = calloc(n, sizeof(int));

and it seems fine.

Stéphane CARPENTIER <sc@fiat-linux.fr> wrote:
>Le 09-06-2024, Tyrone <none@none.none> a écrit :
>> On Jun 8, 2024 at 7:42:16?PM EDT, "Joel" <joelcrump@gmail.com> wrote:
>>
>>> What purpose does coding that serve? I can code, I choose not to
>>> because everything's already been coded, by someone else.
>>
>> What an extremely naive thing to say.
>
>You seem surprised. Is it the first time you read a message from Joel?

I may not have the IT skills you have, but I know my way around
computers, everything I need to do has been coded by someone, I never
find that I would need to code myself.

--
Joel W. Crump

Amendment XIV
Section 1.

[...] No state shall make or enforce any law which shall
abridge the privileges or immunities of citizens of the
United States; nor shall any state deprive any person of
life, liberty, or property, without due process of law;
nor deny to any person within its jurisdiction the equal
protection of the laws.

Dobbs rewrites this, it is invalid precedent. States are
liable for denying needed abortions, e.g. TX.

On 14 Jun 2024 17:27:00 GMT, Stéphane CARPENTIER wrote:

>
> And here, it's only an English forum. When you want to sort the names of
> your CD albums, it can become tricky. For example how do you take care
> of the letters with diacritics signs? How do you include Cyrillic
> letters ? Or the Arabic or Jews letter which are written from right to
> left? You can have a lot of fun with something which looks like a simple
> problem.
>

Ha, ha, ha, ha, ha, ha, ha, ha, ha, ha!

To every COMPETENT C PROGRAMMER (that's not you) it is a VERY SIMPLE
problem.

But just to demonstrate your idiocy, I shall not describe the routine
process. I'll allow you to "solve the problem."

IOW, I'll allow you to fail miserably at a task that is extraordinarily
simple to COMPETENT C PROGRAMMERS.

Ha, ha, ha, ha, ha, ha, ha, ha, ha, ha!

Of course, there are no competent C programmers on this group (aside from
me) that could assist you.

This group is plagued with pathetic C poseurs.

Le 14-06-2024, Joel <joelcrump@gmail.com> a écrit :
> Stéphane CARPENTIER <sc@fiat-linux.fr> wrote:
>>Le 09-06-2024, Tyrone <none@none.none> a écrit :
>>> On Jun 8, 2024 at 7:42:16?PM EDT, "Joel" <joelcrump@gmail.com> wrote:
>>>
>>>> What purpose does coding that serve? I can code, I choose not to
>>>> because everything's already been coded, by someone else.
>>>
>>> What an extremely naive thing to say.
>>
>>You seem surprised. Is it the first time you read a message from Joel?
>
>
> I may not have the IT skills you have, but I know my way around
> computers, everything I need to do has been coded by someone, I never
> find that I would need to code myself.

It's not about the technical skills. It's about the naive way you see
things. If everything had already been coded, there would be neither new
programs, nor improvements of old programs. Which is just plain wrong.

--
Si vous avez du temps à perdre :
https://scarpet42.gitlab.io

Le 14-06-2024, Farley Flud <ff@linux.rocks> a écrit :
> On 14 Jun 2024 17:27:00 GMT, Stéphane CARPENTIER wrote:
>
>>
>> And here, it's only an English forum. When you want to sort the names of
>> your CD albums, it can become tricky. For example how do you take care
>> of the letters with diacritics signs? How do you include Cyrillic
>> letters ? Or the Arabic or Jews letter which are written from right to
>> left? You can have a lot of fun with something which looks like a simple
>> problem.
>
> To every COMPETENT C PROGRAMMER (that's not you)

That's not you neither. We are even on this point.

> it is a VERY SIMPLE problem.

Why? Is there a library like in python? So, either you can do it all by
yourself and it's far from that simple. Or, you are using a library to do
it for you and you are just like any other python programmer. It's your
choice. I'll have fun with any choice you could make. i can't believe
you could come with a third possibility.

Make me laugh, I'm waiting for your answer.

> But just to demonstrate your idiocy, I shall not describe the routine
> process. I'll allow you to "solve the problem."

Of course. As I wrote just before, you are not a competent programmer, so
you have no clue for a way to do it. And most importantly, you neither
found it with google (which is your only programming skill).

> IOW, I'll allow you to fail miserably at a task that is extraordinarily
> simple to COMPETENT C PROGRAMMERS.

OK, I failed. So now, if you dare, provide me the answer. I'm pretty
sure you can't and we'll both fail on this.

> Of course, there are no competent C programmers on this group (aside from
> me) that could assist you.

In fact, there are competent C programmers here. I'm not one of them
and I never claimed to be. You are not one of them unlike your claims.

> This group is plagued with pathetic C poseurs.

You are the only one C poseur here, but you are not pathetic, you are
really funny. And the competent C programmers don't need to give me the
answer because :
- I don't need it, I know the difficulties and that's enough for me,
- the issue is the same, disregarding the programming language,
- they have better things to do,
- if they give me the answer, you would copy/paste it, claiming it your
own and they don't want that,

--
Si vous avez du temps à perdre :
https://scarpet42.gitlab.io

On 6/14/2024 4:01 PM, Lying Larry Pietraskiewicz spewed his usual bullshit:

> This group is plagued with pathetic C poseurs.

There is NOBODY here that's nearly the liar and programming/IT fraud you
are:

"C Programmer Extraordinaire"

"Then I'd write it in C, using a reentrant lexical parser
and parallel processing."

"The GIMP people want me to implement netpbm HDR formats. The
ImageMagick folks want me to implement AI. Give this poor boy
a fucking break!"

"I am heavily involved in virtually ALL AREAS of contemporary
computing"

LMFAO! That last one takes the cake.

As deluded and dishonest as you are, you still can't hold a candle to
Rex Ballard.

On 14 Jun 2024 20:55:27 GMT, Stéphane CARPENTIER wrote:

>
> OK, I failed. So now, if you dare, provide me the answer.
>

Good. You admit that you are a failure. I will reward you,
like a dog, for your admission with the following treat:

Unicode normalization:

http://www.unicode.org/reports/tr15/

GNU/Linux provides the incomparable "libunistring" that every
competent C programmer should have in his toolbox:

https://www.gnu.org/software/libunistring/

>
> In fact, there are competent C programmers here.
>

There are *no* competent C programmers here (except little ol'
me).

Example: no one here (except me) could ever use C to do the simple
task of error-free floating point summation.

Another example: no one here (except me) could ever exploit the
hardware "instruction pipeline" of Intel and AMD processors using
C code.

Other examples are endless.

Nothing here (aside from me) except C poseurs.

On 6/14/2024 5:20 PM, Farley Flud wrote:
> On 14 Jun 2024 20:55:27 GMT, Stéphane CARPENTIER wrote:

>> In fact, there are competent C programmers here.
>>
>
> There are *no* competent C programmers here (except little ol'
> me).

You're a joke.

> Example: no one here (except me) could ever use C to do the simple
> task of error-free floating point summation.

You can't do it, either. ALL you could EVER deliver is a plagiarized
example from the GMP documentation.

> Another example: no one here (except me) could ever exploit the
> hardware "instruction pipeline" of Intel and AMD processors using
> C code.

Another vague claim you'll NEVER deliver on.

Melzzzzz is gone, but he whipped your silly ass up one side and down the
other when it came to assembler.

> Other examples are endless.

So are your ridiculous lies and boasts.

> Nothing here (aside from me) except C poseurs.

You are NOT a competent programmer, in ANY language.

* you ran away screaming from many simple string challenges in C.

* you ran away mewling from a simple directory reading challenge in C.

* you ran away wimpering from a challenge to summarize .csv files in C.

* you ran away crying from a challenge to download the latest
['tar.gz', 'tar.xz', 'tar.bz2', 'tar.lz', '.zip'] files from each
directory at http://mirror.rit.edu/gnu/

You're a lying, blustering fraud.

STFU and pay Mom the rent.

Le 14-06-2024, Farley Flud <ff@linux.rocks> a écrit :
> On 14 Jun 2024 20:55:27 GMT, Stéphane CARPENTIER wrote:
>
>>
>> OK, I failed. So now, if you dare, provide me the answer.
>>
>
> Good. You admit that you are a failure. I will reward you,
> like a dog, for your admission with the following treat:
>
> Unicode normalization:
>
> http://www.unicode.org/reports/tr15/
>
> GNU/Linux provides the incomparable "libunistring" that every
> competent C programmer should have in his toolbox:
>
> https://www.gnu.org/software/libunistring/

You granted me with the best of the both answers available to you. You
still can dive even when you are at the bottom. It's really impressive.
So like any python programmer you despise, you rely heavily on
libraries. But unlike them, you are unable to understand their purpose.
I'm speaking about sorting characters, not about the way to recognise
them. For example is "Γ" before or after "G" which is before or after
"g"? And then, you'll have the Chinese characters to sort, how do you do
it?

>> In fact, there are competent C programmers here.
>>
>
> There are *no* competent C programmers here (except little ol'
> me).

Your sentence is easy to correct in a logical syntax: you just put a
"not" before and it's done. The issue is not only about technical skills
but about understanding the issues to solve. And you lack both of the
required qualities.

> Example: no one here (except me) could ever use C to do the simple
> task of error-free floating point summation.

Mouarf.

> Another example: no one here (except me) could ever exploit the
> hardware "instruction pipeline" of Intel and AMD processors using
> C code.

Mouarf. Again.

> Other examples are endless.

You have examples, when you dream you provided a solution, but for this
specific case, you provided nothing.

> Nothing here (aside from me) except C poseurs.

You provided me with a link toward an unrelated library. I'm still
waiting for the answer you will never been able to give: how do you sort
characters in mixed languages?

So for me, you are only a C poseur. It's not enough to have a link
toward a library, it's important to be able to use it.

--
Si vous avez du temps à perdre :
https://scarpet42.gitlab.io

On Fri, 14 Jun 2024 14:14:33 -0400, DFS wrote:

> Yep. In the big stats program I posted I messed around with using a
> malloced array of 20,000. It ran a bunch of times just fine, but all of
> a sudden I got nonsense results.
>
> So I filled it with 0s before using it, and no problem.
>
> But calloc saves a line of code, so that's what I used.

Yeah, that's the joy of undefined behavior. In the original program with
'int mode' not initialized it segfaulted pretty reliably on
'md_count[mode]' if 'mode' happened to be a garbage number. It *might*
have been 0 for a dew runs. 'double mean' was uninitialized but on a few
runs it was 0.0. Maybe the compiler was feeling nice but I wouldn't bet on
it. If it wasn't 0.0, 'mean += nums[len];' wouldn't blow up but it would
be bogus.

'int md_count[COUNT_CNT];' was another uninitialized one. I printed it out
and some elements were 0, some were bogus which wiped out the mode logic
even with mode initialized.

Luckily, it did have 10 elements although the 0th would never be used. If
it were md_count[COUNT_CNT-1] it could potentially get interesting. It's
on the stack so md_count[9] possibly could corrupt something else on the
stack like 'len'. That's a real crap shoot that depends on how the stack
frame was built and what is adjacent.

> On geeksforgeeks a calloc() example is:
>
> int* ptr;
> ptr = (int*)calloc(n, sizeof(int));
>
> Isn't the 2nd int* redundant?

Most C compilers let it slide but malloc, calloc, and realloc return void
pointers. C++ doesn't.

$ g++ -o mean mean.c
mean.c: In function ‘int main()’:
mean.c:15:21: error: invalid conversion from ‘void*’ to ‘int*’ [-
fpermissive]
15 | ptr = calloc(COUNT_CNT, sizeof(int));
| ~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~
| |
| void*

If you work with both languages and/or different compilers you get into
the habit of explicitly casting the return. A C++ purist would frown on
using calloc or malloc although you certainly can. 'new' probably uses one
or the other under the hood but you don't want to mix and match free and
delete.

Going the other way C++ allowed variable declarations outside of blocks. C
finally caught up, I forget which version, but I would follow the C rule
in C++ although I really prefer declaring a variable close to where you're
going to use it.

Stéphane CARPENTIER <sc@fiat-linux.fr> wrote at 20:38 this Friday (GMT):
> Le 14-06-2024, Joel <joelcrump@gmail.com> a écrit :
>> Stéphane CARPENTIER <sc@fiat-linux.fr> wrote:
>>>Le 09-06-2024, Tyrone <none@none.none> a écrit :
>>>> On Jun 8, 2024 at 7:42:16?PM EDT, "Joel" <joelcrump@gmail.com> wrote:
>>>>
>>>>> What purpose does coding that serve? I can code, I choose not to
>>>>> because everything's already been coded, by someone else.
>>>>
>>>> What an extremely naive thing to say.
>>>
>>>You seem surprised. Is it the first time you read a message from Joel?
>>
>>
>> I may not have the IT skills you have, but I know my way around
>> computers, everything I need to do has been coded by someone, I never
>> find that I would need to code myself.
>
> It's not about the technical skills. It's about the naive way you see
> things. If everything had already been coded, there would be neither new
> programs, nor improvements of old programs. Which is just plain wrong.

Also, there's always going to be a small niche that you need filled. I
know I have a bunch of stuff I made after being frustrated I couldn't
find anything online (or by just cobbling together existing programs in
a bash script :P)
--
user <candycane> is generated from /dev/urandom

Stéphane CARPENTIER <sc@fiat-linux.fr> wrote at 16:42 this Friday (GMT):
> Le 08-06-2024, rbowman <bowman@montana.com> a écrit :
>>
>> So, what do you want to do? Of course you can do all of the above with C,
>> C++, or another language but you may be writing more code and doing more
>> debugging. If the Python prototype is too slow you can rewrite it.
>
> Like one said:
> - with C, you have a million ways to do one thing,

I think it's more that you just have the basic blocks and have to
abstract everything yourself more?

> - with python, you have a million libraries to do one thing,

Yeah, pypip hasa ton of seemingly redundant stuff, I won't argue there.
For most common functions, there's usually some kind of standard tho.

> - with C++, you have a million ways to screw things up.

YEAH. I still get confused between cstrings, std::string, and
std::string_view. And std::vector vs std::array vs c arrays. And all the
different kind of pointers. And
--
user <candycane> is generated from /dev/urandom

On 14 Jun 2024 23:46:05 GMT, Stéphane CARPENTIER wrote:

> For example is "Γ" before or after "G" which is before or after
> "g"? And then, you'll have the Chinese characters to sort, how do you do
> it?
>

?????????

Is 10 before or after 3 which is before or after 5?

You are dumb! Characters are not sorted; numbers are sorted --
and this is the whole point of normalization.

Of course, Chinese chars are way down the list with American
ASCII at the very top. I suppose that, in time, the Chinese
will scrap Unicode and replace it with an encoding that has
their characters at the very top.

candycanearter07 <candycanearter07@candycanearter07.nomail.afraid>
wrote:
>Stéphane CARPENTIER <sc@fiat-linux.fr> wrote at 20:38 this Friday (GMT):
>> Le 14-06-2024, Joel <joelcrump@gmail.com> a écrit :
>>> Stéphane CARPENTIER <sc@fiat-linux.fr> wrote:
>>>>Le 09-06-2024, Tyrone <none@none.none> a écrit :
>>>>> On Jun 8, 2024 at 7:42:16?PM EDT, "Joel" <joelcrump@gmail.com> wrote:
>>>>>
>>>>>> What purpose does coding that serve? I can code, I choose not to
>>>>>> because everything's already been coded, by someone else.
>>>>>
>>>>> What an extremely naive thing to say.
>>>>
>>>>You seem surprised. Is it the first time you read a message from Joel?
>>>
>>> I may not have the IT skills you have, but I know my way around
>>> computers, everything I need to do has been coded by someone, I never
>>> find that I would need to code myself.
>>
>> It's not about the technical skills. It's about the naive way you see
>> things. If everything had already been coded, there would be neither new
>> programs, nor improvements of old programs. Which is just plain wrong.
>
>Also, there's always going to be a small niche that you need filled. I
>know I have a bunch of stuff I made after being frustrated I couldn't
>find anything online (or by just cobbling together existing programs in
>a bash script :P)

I do understand that people have needs like that, and it's conceivable
I would too, it's not like I couldn't code, I just haven't encountered
that need apart from mIRC scripts in a long time.

--
Joel W. Crump

Amendment XIV
Section 1.

[...] No state shall make or enforce any law which shall
abridge the privileges or immunities of citizens of the
United States; nor shall any state deprive any person of
life, liberty, or property, without due process of law;
nor deny to any person within its jurisdiction the equal
protection of the laws.

Dobbs rewrites this, it is invalid precedent. States are
liable for denying needed abortions, e.g. TX.

On 2024-06-14 5:05 p.m., DFS wrote:

> As deluded and dishonest as you are, you still can't hold a candle to
> Rex Ballard.

You deadnamed the faggot! Transphobe! It's Debbie Ballard now and "she"
is stunning and brave.

--
Andrzej (Andre) Matuch
TG: @AndrzejMatuch
Catholic, paleoconservative, Christ is king.

On 14 Jun 2024 23:46:05 GMT, Stéphane CARPENTIER wrote:

>
>>> In fact, there are competent C programmers here.
>>>
>>
>> There are *no* competent C programmers here (except little ol'
>> me).
>

C is a deceptively very simple language. It has very simple constructs
but to apply them properly requires profound technical insight.

Just one example:

The poseurs on this NG have actually posted C code similar to this:

===============
Begin shit code
===============

type1 function_1(...) {
....
}

type2 function_2(...) {
....
}

....
while(...) // loop over
{ ....
var1 = function_1(...);
....
var2 = function_2(...);
....
}

=============
End shit code
=============

Then the poseurs will fawn all over themselves for
a great coding job when their code, like the above,
is pure shit.

They can't see the glaringly obvious shitness.

Le 15-06-2024, Farley Flud <ff@linux.rocks> a écrit :
> On 14 Jun 2024 23:46:05 GMT, Stéphane CARPENTIER wrote:
>
>> For example is "Γ" before or after "G" which is before or after
>> "g"? And then, you'll have the Chinese characters to sort, how do you do
>> it?
>>
>
> ?????????

I'm not surprised you didn't understood.

> Is 10 before or after 3 which is before or after 5?

The 3 comes always before 5. Now, for 10, it can comes before or after
3 depending on the way you are sorting. If you are sorting strings, it's
before, if you are sorting numbers, it's after. Now, I'm able to answer
your question. Are you able to answer mine?

> You are dumb!

Yes, I now that but it doesn't answer my question.

> Characters are not sorted; numbers are sorted --
> and this is the whole point of normalization.

I know you shit in your basement, so I believe everything's a mess in
your basement. But normal people in real life need to sort things to be
able to find them after. So, when they have books at home or when they
are going to a library to find books, they hope the books are sorted. If
they aren't sorted, they'll have to look at every book one by one to
find the one they want and it would take forever.

And it's easier to find books sorted by authors or by name than by
number of pages. I'm not sure you'd be able to understand why, but I
won't explain more the reason: you could try to find out by yourself,
but you would need to learn how to read first which would take forever.

So, your inability to find a way to sort characters doesn't prove there
is no need for it. It's something which can be solved, but not by you.

> Of course, Chinese chars are way down the list with American
> ASCII at the very top. I suppose that, in time, the Chinese
> will scrap Unicode and replace it with an encoding that has
> their characters at the very top.

Once again, you wrote something only to show you don't know anything
about it. Sorting Chinese characters has nothing in common with sorting
English characters, so one needs to find a way to sort them together.
It's not that obvious.

--
Si vous avez du temps à perdre :
https://scarpet42.gitlab.io

Le 15-06-2024, Farley Flud <ff@linux.rocks> a écrit :
> On 14 Jun 2024 23:46:05 GMT, Stéphane CARPENTIER wrote:
>
>>
>>>> In fact, there are competent C programmers here.
>>>>
>>>
>>> There are *no* competent C programmers here (except little ol'
>>> me).
>>
>
> C is a deceptively very simple language. It has very simple constructs
> but to apply them properly requires profound technical insight.

And in this message you managed, once again, to show you don't have the
basic technical insight (to say nothing about the profound ones).

> Just one example:

I didn't found it. My ISP doesn't show me every message, but it doesn't
cut the messages it shows me. So the issue is you.

> The poseurs on this NG have actually posted C code similar to this:

I don't find the code.

>===============
> Begin shit code
>===============
>
> type1 function_1(...) {
> ...
> }
>
> type2 function_2(...) {
> ...
> }
>
> ...
> while(...) // loop over
> {
> ...
> var1 = function_1(...);
> ...
> var2 = function_2(...);
> ...
> }
>
>=============
> End shit code
>=============

There is no code, there is only declarations.

> Then the poseurs will fawn all over themselves for
> a great coding job when their code, like the above,
> is pure shit.

First, there is no code. Second you are unable to explain why it's shit.
Third, you are unable to improve it.

--
Si vous avez du temps à perdre :
https://scarpet42.gitlab.io

On Sat, 15 Jun 2024 06:30:03 -0000 (UTC), candycanearter07 wrote:

> Stéphane CARPENTIER <sc@fiat-linux.fr> wrote at 16:42 this Friday (GMT):
>>
>> - with C++, you have a million ways to screw things up.
>
> YEAH. I still get confused between cstrings, std::string, and
> std::string_view.

Here’s another one I found (used in Blender):
<https://github.com/imageworks/pystring>

On Fri, 14 Jun 2024 14:14:33 -0400, DFS wrote:

> Here's how python checks array bound every time the
> array is accessed:
>
>
> int PyList_SetItem(PyObject *op, Py_ssize_t i,
> PyObject *newitem)
> {
> PyObject **p;
> if (!PyList_Check(op)) {
> Py_XDECREF(newitem); PyErr_BadInternalCall();
> return -1;
> }
> if (!valid_index(i, Py_SIZE(op))) {
> Py_XDECREF(newitem); PyErr_SetString(PyExc_IndexError,
> "list assignment index out of range");
> return -1;
> }
> p = ((PyListObject *)op) -> ob_item + i;
> Py_XSETREF(*p, newitem);
> return 0;
> }
>
>
> Python-3.11.0.tgz.2\Python-3.11.0\Objects\listobject.c line 257
>
>
> Which is another thing that contributes to its slowness vs C.

From the same source file:

static inline int
valid_index(Py_ssize_t i, Py_ssize_t limit)
{
/* The cast to size_t lets us use just a single comparison
to check whether i is in the range: 0 <= i < limit.

See: Section 14.2 "Bounds Checking" in the Agner Fog
optimization manual found at:
https://www.agner.org/optimize/optimizing_cpp.pdf
*/
return (size_t) i < (size_t) limit;
}

Somehow I don’t think that’s the bottleneck in PyList_SetItem.

Also have a look at the list_extend routine, since that is commonly used
for appending lots of items.