[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Enigma-devel] Port of r99 to Lua 5.1

From: Ronald Lamprecht
Subject: Re: [Enigma-devel] Port of r99 to Lua 5.1
Date: Fri, 31 Mar 2006 00:00:18 +0200
User-agent: Mozilla Thunderbird 1.0.7 (Windows/20050923)


Tacvek wrote:
Suggested action: Delete all Lua 4 files, and all tolua++ files (including those in ./tools). Then re-add files from tolua++ and lua 5.1 as requested by the Makefile.ac's.

Well for subversion's sake I did patch all files that continue to exist and did "svn add", "svn delete" the others.

If after doing that and applying the partial patch it does not compile without error, please let me know. If all else fails I can create a Berlios account, somebody can add me to the project, and I can upload the changes to a new svn branch.

I created a feature branch at "branches/lua5" as a copy of "trunk" r104 and added Lua 5.1, tolua++ 1.0.91 with your Enigma patches to the branch (r106).

For development on this branch either checkout this branch or switch an existing working copy to this branch.

Tacvek, if you tell me your Berlios account I can add you to the project.

Please let me now how progress with the patch comes. The main thing needed is updating the lua scripts. Once you get to the point where a level can load we will see if my code related to userdata magicvally works. If it does not I suspect that lua.cc will note an error, and/or a segfault.

First we have to preload the library compatibility scipt and recode the %upvalues - and then let us hope your magic works.

I'm not really a Lua programmer, so I can't work on updating the Scripts, but I try to debug any userdata problems. My problem before was trying to see if there would be problems by reading the code, as I was not yet able to test that code. If push came to shove, I could recode lua.cc to never touch userdata directly, but to interact with userdata only via tolua++. Obviously tolua++ will not have any problem with compatability with itself.
In the long run, that is probably a good idea anyway.

As I understood Daniel, he would prefer to skip the tolua access for world objects in the short term.

If you succeed to proof that the lua.cc interface functions can be substituted by tolua++ ones you may convince him for a long run switch.


reply via email to

[Prev in Thread] Current Thread [Next in Thread]