User talk:EdK: Difference between revisions

2,163 bytes removed ,  9 years ago
Clipper Harbour XBase
(Clipper Harbour XBase)
 
(2 intermediate revisions by 2 users not shown)
Line 25:
 
== Language and library organization ==
(Moved to [[Rosetta Code:Village Pump/Semantic MediaWiki/Semantics]] (--[[User:Short Circuit|Michael Mol]] 18:05, 29 August 2010 (UTC)))
 
== Yorick ==
I really appreciate all the work you're doing with language, implementation and library association and documentation. I've got some ideas about associating those pages within the context of Semantic MediaWiki. Would you be interested in hashing this out and helping get it implemented? (It should only involve making changes to wiki pages, much like what you're already doing.)
 
Your language list included "Yoric", but the language's actual name is "Yorick". I corrected it in your listing so that I could correct the relevant categories. -- [[User:Sekoia|Sekoia]] 22:38, 10 October 2010 (UTC)
My thoughts are:
* Distinguish between library, API and bindings between libraries and bindings to a language. That way:
** An example said to use, e.g. PyGTK could be programattically assumed to use GTK, and an example said to use Xcb could be assumed to use X11.
** An example using GLib could be assumed to be using OpenGL
** GLUT is both an API and a library, but the original has long lapsed. There are replacements, though, and they could all be assumed to be implementations of GLUT.
** The ISO89 C Runtime has an API spec, and any library/C language binding that correctly implements it could be reasoned by SMW to provide that runtime, and so any C example that uses an API spec could be reasoned to be an example use of those libraries.
 
== Clipper / Harbour / XBase ==
That's the gist of the idea, for now. Does this strike your interest? --[[User:Short Circuit|Michael Mol]] 17:28, 13 August 2010 (UTC)
Do you have any objection to Clipper and Harbour being stored under an XBase heading in task solutions? It would seem to me better than duplicating code under separate Clipper and Harbour headings. [[User:Axtens|Axtens]] ([[User talk:Axtens|talk]]) 13:00, 8 June 2014 (UTC)
 
I think I understand.
 
For example:
* pyGame is a binding of SDL to Python. SDL in turn uses OpenGL. This information should be captured somehow.
* There are several differant impementations of MPI, some of them have several language bindings. This should be captured.
 
but I am not sure why it matters what implementation of ISO89 is being used. They are generaly closly tied to the compiler, so specifing the compiler would generaly specify the library. Or am I misssing something?
 
I'm not exactly sure how to do this in this contect. [[User:EdK|-EdK]] 20:40, 19 August 2010 (UTC)
:* I think the pyGame->SDL->OpenGL can be captured. [[User:Ce]] and [[User:Coderjoe]] have been doing a lot of work in sussing out the specifics of how SMW operates.
:* I agree about the MPI considerations. It's rather similar to the variations between BASIC, FORTH, SQL or LISP implementations, when you look at it in a particular way. Each is similar, but significantly vary between implementations to the point where some implementations are considered to be their own languages. I'd ''love'' to find a way to decently capture that!
:* I don't know that the particular version of the ISO89 runtime has practical value, but it ''is'' a distinction in how code may operate between compilers that purport to implement it, due to compiler-specific bugs and omissions. (If it were possible to break down the standard/implementation relationship by implemented/unimplemented/buggy components, I'd certainly prefer ''that!'') --[[User:Short Circuit|Michael Mol]] 15:55, 20 August 2010 (UTC)
Anonymous user