[Date Prev][Date Next][Thread Prev][Thread Next]
[Date Index]
[Thread Index]
- Subject: LuaCheia Summary for week ending March 3
- From: "Thomas Wrensch" <twrensch@...>
- Date: Wed, 05 Mar 2003 15:51:39 -0800
The following is a list of the decisions being tossed around on the
LuaCheia mailing list. This will also be posted on the Wiki on the
LuaCheiaSummaries page.
-------------
Source file naming:
Make all source file names lowercase only. Agreed to by several people.
Seems to be a concensus.
---
Modules polluting namespace issue:
Modules may add functions to the global namespace. Suggested way to
clean this up by:
lib.libname.functionname = functionname
functionname = nil
This suggestion was not discussed in detail. No decisions made.
---
USES() function to implicitly load modules:
Suggestion to specify a standard API which can be implemented by several
modules. The USES() load specify the API and load an appropriate model.
---
New CVS Structure
New CVS structure implemented that separates the base Lua
implementation, specific project files, and the LuaCheia code.
No complaints (no reason, it's a good setup), so it is a defacto group
decision.
---
LuaCheia directory access module:
Discussion on several important issues regarding this module.
Names for access to working directory and changing the working
directory. Uncertain about names for the function to change the working
directory (cwd() versus chdir()). Also some discussion around merging
the two functions.
Seems to be leaning toward pwd() and chdir(), but no
concensus.
Use of 'lib' namespace rather than luacheia namespace for this module.
Leaning toward agreement to this.
Discussion of the way information is returned by the listdir() function.
Adding a trailing '/' to the names of subdirectories. Optional approach
is to return a table whose keys are the file names and values are
information about the files.
Definitely leaning toward trailing '/'. No decision
on the "name as key" table structure.
---
Using existing loadmodule interface
Should we use the existing loadmodule interface or continue with our own
version.
Leaning toward supporting both.
---
Updates to position statement on web page:
Suggetions for modifications and fixes to the position statement on the
LuaCheia web page: Remove planning info, eventually add one-line
summaries for modules, eliminate some controversial wording.
Seems to be general agreement to idea of changes,
details still up in the air.
---
Supporing static and dynamic modules:
Suggested approach for adjusting modules for both static and dynamic
loading when using glua:
#ifdef GLUA_STATIC // Okay for static & dynamic link
#define GLUA_MODULE_NAME GluaModule_sys
#endif
For dynamic link name is always GluaModule, in static link all names
have to be different.
Insufficient discussion yet to know if this is
accepted, also need to generalize to other loaders.
---