New IDTech 4 port

[lowenz]

.....in ADA! :| Maybe with a Vulkan renderer.

 

http://www.phoronix.com/scan.php?page=news_item&px=Ada-Doom-3

[nbohr1more]

Interesting. Hard to say how useful since the whole structure of the code was rearranged during the ADA port process.

 

Also seems to be contrary to the original intention of the project.

 

ADA was chosen because it's very strict and robust. It's for long-term unbreakable code for military and industrial clients.

 

Vulkan, on the other hand, is brand new, experimental, and not well defined in terms of specification. Right now, AMD and Nvidia

are probably letting Developers code Vulkan renderers any way they want to and then defining the specification based

on their use-case and feedback. So a sloppy, vendor driven specification that is currently "work in progress" being chosen

for a stable "military grade" code project?

 

What the hell is going on?

 

Maybe they think Vulkan is so unstable they need a super stable underlying backend so they can expedite bug tickets.

Eg. "See, even if the code was created by John Carmack and converted to ADA, your stinky Vulkan specification causes

problems... Make the specification more strict!!!"

[lowenz]

Good interpretation!

[HMart]

Intriguing project, i don't know anything about that language, looking at it looks very similar to the language Jonathan blow (witness developer) is making called Jai!

 

Btw nbobr1more how can a language be "military grade"? Any computer language in good hands is very stable any "rock solid military grade" language in bad (noob) hands is unstable, IMO is the programmer that counts not the language, on the other end speed and readability of the code does depend on the language.

[jaxa]

SELinux is written in C. Is 'C' military grade?

[nbohr1more]

ADA is more strict about the scope of code that can compile. You need to pay attention to the local and global scope of variables and functions. Even perfectly logical code won't compile if there is some universal principle that would lead to a risky code condition if used on other code. I suppose you can externally achieve a similar level of risk avoidance in C++ with things like LLVM code analysis.