Programming thread

[Creative Username]

Yes and no, there's a lot of extraneous horseshit in the spec, and you run into this behavior regularly, which is, in part, why JSON is replacing it in a lot of data serialization contexts.

it might have shit like namespaces and there might be a bunch of weird shit around it like xslt, but if you look at only the core document format specification, it's not as horrible as a lot of programming circlejerkers say it is

This is a place where I have felt significant corporate pain. The MS Office and MS Visual Studio groups had their own, non-interoperable XML specifications (was working with Sharepoint 2013 at the time; Microsoft specified data queries in Sharepoint using an XML dialect called CAML) and it required weird monkey patching to make them work together. Boss hated when I'd spend time to make my life easier in the long run, which required doing things like this. He'd been basically a nocoder for a decade or two. Horrid workplace unless you know how he likes his cock sucked.

this isn't xml's fault really, this is niggerlicious niggers defining different dialects of a format designed to be interoperable and not have a bunch of stupid fucking dialects
don't blame xml for microsoft's stupid mistakes because if you applied that rule for everything microsoft has fucked up badly you would quickly conclude that everything is terrible (which is true for other reasons, just not because microsoft is really good at fucking everything up)

XML seems straightforward because most programmers worth a hoot has tried their hands at doing HTML, usually when young.

html is the real nightmare because it follows the "be generous in what you accept" rule to its logical conclusion and shows why following that rule to its logical conclusion can be a bad idea

 

[Polyarmory]

html is the real nightmare because it follows the "be generous in what you accept" rule to its logical conclusion and shows why following that rule to its logical conclusion can be a bad idea

Agreed. I think that rule should only apply when you're not influencing the specification, and web browsers are basically defining it.

HTML was also intended to be ergonomic to write, so the standard has some irregularities for "quality of life" like being allowed to not close certain elements.

 

[Creative Username]

Agreed. I think that rule should only apply when you're not influencing the specification, and web browsers are basically defining it.

then you get all the browsers having to informally accept broken shit in a specific way when some nigger writes a shitty page during some browser monopoly period, and now everybody has to accept broken shit just like how internet explorer did or google chrome does
i'm telling you: bullshit heuristics are not worth it
if it was 199x slightly before browsers became mainstream and i had a say in all of it, browsers would have the spine to say "this page is irreparably malformed, and it cannot be displayed. you may want to contact the site's administrator to inform them of this issue. [view source]" and they would make errors like that whenever the syntax is not perfectly well-formed

 

[WielkiWapiesz2137]

The so called John Von Neumann machine seemed to have its start on German WWII tech.

(((Von Neumann))) was as german as (((von Mises))). The fact is that the glorious creation of G-d such as LISP(the language in which universe was written) are revealed to His chosen people while gentile cattle tries to poorly imitate His creations.

 

[Creative Username]

The fact is that the glorious creation of G-d such as LISP(the language in which universe was written) are revealed to His chosen people while gentile cattle tries to poorly imitate His creations.

don't forget gerald sussman

 

[Rango Dango]

don't forget gerald sussman

i think gerald is sus, man

 

[Anti Snigger]

You guys don’t get it. COBOL is a fucked language, nobody should ever use it under any circumstance.
However, we didn’t realize this soon enough. There are hundreds of millions of lines of COBOL running on machines from the 1960s that are integral to the US government and many financial institutions. There’s already a well documented problem with the retirement drain in these sectors. Learning COBOL automatically will make me more employable in these sectors than anyone else on earth

 

[Private Tag Reporter]

There are hundreds of millions of lines of COBOL running on machines from the 1960s that are integral to the US government and many financial institutions.

Have they tried re-writing it in Rust?

 

[Anti Snigger]

Have they tried re-writing it in Rust?

Personally based on what I know, I’d rewrite cobol into Fortran first and foremost, if nothing else to stage it for further refactoring. I also think AWK would be a solid middle ground given the way files are parsed

 

[ChudJack009]

You guys don’t get it. COBOL is a fucked language, nobody should ever use it under any circumstance.
However, we didn’t realize this soon enough. There are hundreds of millions of lines of COBOL running on machines from the 1960s that are integral to the US government and many financial institutions. There’s already a well documented problem with the retirement drain in these sectors. Learning COBOL automatically will make me more employable in these sectors than anyone else on earth

dont most banks and other such financial institutions also run on COBOL?

 

[Creative Username]

COBOL is a fucked language, nobody should ever use it under any circumstance.

while nobody should use cobol today, it was leagues better than the alternatives for doing financial calculations on flat databases in the 60s

There’s already a well documented problem with the retirement drain in these sectors. Learning COBOL automatically will make me more employable in these sectors than anyone else on earth

i think the main problem with cobol is that you need to know the codebase and the only guy who really knows the codebase is the boomer who wrote it and maybe you too but only if you were raised from birth by the boomer who wrote it so you can understand the codebase

Have they tried re-writing it in Rust?

i'd love to see a satire where a rustlet rewrites a million lines of cobol in rust only to have done it all with single precision floats for the blazing fast performance
then the resulting rounding errors bankrupt everybody on the planet somehow

dont most banks and other such financial institutions also run on COBOL?

he said "governments and many financial institutions"

 

[jONALD]

I finally got my hands on an introduction to programming that isn’t some retarded “here is this function, and this function and when you combine them you get this output!”
THATS NOT A (((FROM 0))) INTRODUCTION YOU STUPID NIGGERS!

I need my learning to be broken down further than that to grasp the concepts and how such things work, it’s how I learn best practices from those that came before me. I am happy to report to no one that I have finally found such a book!
Common Lisp: A gentle Introduction To Symbolic Computation
It all just clicks for me better. sorry C bros, I really tried.

 

[analrapist]

dont most banks and other such financial institutions also run on COBOL?

Everyone will tell you that they're modernizing, but there is a substantial amount of technical debt.

Here's sample COBOL code from IBM: https://www.ibm.com/docs/en/develop...c=mode-example-sample-cobol-program-debugging

If you've wrangled ASM and C, especially together, you should have a good notion of what this code is doing without knowing the language at all. I've never learned a word of COBOL and I could translate this into C/ASM nearly blindly.

Naturally, practice makes perfect.

If I'm being honest, the way COBOL explicitly defines memory spaces in the program code speaks to me. I've had a notion of writing a language that can enable that kind of behaviour. C abstracts it away too much for my taste.

I need my learning to be broken down further than that to grasp the concepts and how such things work, it’s how I learn best practices from those that came before me. I am happy to report to no one that I have finally found such a book!
Common Lisp: A gentle Introduction To Symbolic Computation
It all just clicks for me better. sorry C bros, I really tried.

Bro, I love me some C, and I'm no big fan of Lisp beyond its (high) merits, but go with what clicks. Prolog "clicked" for me, and I've loved it since. Ruby "clicked" for me. C "clicked". That click is how I read "personal compatibility".

 

[Belisarius Cawl]

"So, there was a nigger, who came up with this idea...)"
If Terry had used "kike" instead, would people think he was namecalling Stallman instead? Because I feel like that would improve the comparisons (White men doing good things, and Jews doing overly complicated things, and the "niggerlicious" being more akin to Indian tech). The so called John Von Neumann machine seemed to have its start on German WWII tech.

If you're talking about Konrad Zuse and his work, that was done completely independently of largely Anglo-American efforts during and after WWII

 

[Outer Space Traveller]

(((Von Neumann))) was as german as (((von Mises))). The fact is that the glorious creation of G-d such as LISP(the language in which universe was written) are revealed to His chosen people while gentile cattle tries to poorly imitate His creations.

Sorry, I meant teach developed by WWII side and "it just so happened" that John Von Neumann(-steinberg) happened to copy the homework, and today's machine have the architecture named "thanks to him" (but not really).

I am aware Neumann was """(((german-dutch)))""".

If you're talking about Konrad Zuse and his work, that was done completely independently of largely Anglo-American efforts during and after WWII

Yes, that was what I meant to say.

 

[Belisarius Cawl]

Sorry, I meant teach developed by WWII side and "it just so happened" that John Von Neumann(-steinberg) happened to copy the homework, and today's machine have the architecture named "thanks to him" (but not really).

I am aware Neumann was """(((german-dutch)))""".

Yes, that was what I meant to say.

I am not aware that von Neumann had any idea of what Konrad Zuse was doing. Zuse's work was at the time quite obscure and his workshop was obliterated by an Allied bombing raid which meant he had to rebuild later. von Neumann was born in Hungary and of course spoke Hungarian (as well as a number of other languages) but like other Eastern European Jews also knew German and Yiddish. I am not aware of any Dutch connection or knowledge.

 

[Creative Username]

I am not aware that von Neumann had any idea of what Konrad Zuse was doing. Zuse's work was at the time quite obscure and his workshop was obliterated by an Allied bombing raid which meant he had to rebuild later. von Neumann was born in Hungary and of course spoke Hungarian (as well as a number of other languages) but like other Eastern European Jews also knew German and Yiddish. I am not aware of any Dutch connection or knowledge.

a lot of the time when some nigger invents some niggerlicious contraption or theory there are like 3 other niggers creating the same shit at the same time
best example: newton's series of gay callout google docs against leibniz (and vice versa)
"see i made a tweet about infinitesimals a year before you released your stolen theory and the screenshot shows YOU LIKED IT"

 

[Belisarius Cawl]

a lot of the time when some nigger invents some niggerlicious contraption or theory there are like 3 other niggers creating the same shit at the same time
best example: newton's series of gay callout google docs against leibniz (and vice versa)
"see i made a tweet about infinitesimals a year before you released your stolen theory and the screenshot shows YOU LIKED IT"

A closer comparison might be linear programming, devised around the WWII era to solve governmental optimization problems by American George Dantzig and Soviet Leonid Kantorovich (who also both (((echoed))))

 

[Unimaginative Username]

The so called John Von Neumann machine seemed to have its start on German WWII tech.

It's earlier than that: when Babbage's work on his analytical engine was reviewed in 1982 by IEEE Annals of the History of Computing, the reviewer proclaimed "this computer is too much like a modern computer", and it's not surprising because before we had machines, "computer" was a job performed by human beings, who would carry out calculations using lists of simple, unambiguous instructions. It's the natural way to do things, and it's not surprising that Zuse, Von Neumann, and others would adopt a similar strategy even though they were unaware of Babbage's work.

then you get all the browsers having to informally accept broken shit in a specific way when some nigger writes a shitty page during some browser monopoly period, and now everybody has to accept broken shit just like how internet explorer did or google chrome does
i'm telling you: bullshit heuristics are not worth it
if it was 199x slightly before browsers became mainstream and i had a say in all of it, browsers would have the spine to say "this page is irreparably malformed, and it cannot be displayed. you may want to contact the site's administrator to inform them of this issue. [view source]" and they would make errors like that whenever the syntax is not perfectly well-formed

The problem wasn't with HTML, which was itself an "application" of the nightmarishly-complex SGML, it was with the various implementations. SGML's syntax allows closing tags to be omitted when unambiguous (i.e., when encountering a start tag for an element that can't be nested within the currently open element, close elements until it can be legally nested). For example, since paragraphs in HTML cannot be nested, <p>aaa<p>bbb... would be unambiguously parsed as <p>aaa</p><p>bbb.... The problem is that early browsers and HTML editors didn't read the spec (not surprising, since the SGML standard was and presumably still is outrageously expensive to purchase) and just did things on an ad hoc basis, especially when introducing their own novelty elements with ill-defined nesting rules.

 

[Belisarius Cawl]

It's earlier than that: when Babbage's work on his analytical engine was reviewed in 1982 by IEEE Annals of the History of Computing, the reviewer proclaimed "this computer is too much like a modern computer", and it's not surprising because before we had machines, "computer" was a job performed by human beings, who would carry out calculations using lists of simple, unambiguous instructions. It's the natural way to do things, and it's not surprising that Zuse, Von Neumann, and others would adopt a similar strategy even though they were unaware of Babbage's work.

Any chance the Analytical Engine will ever actually be built?