Darren Duncan wrote:The first module I'll attempt this with is SQL::Statement.I'm just in the middle of a spate of releases of this module (two in the last two weeks and another coming in a day or two) after an intense several months of development. If you fork today instead of next week, you'll be forking from a partial version especially in regard to the tests. The code is (knock on wood) complete, but I am still updating the tests and POD. If you can wait a few days, there will be a more complete release. You may want to work from the latest SVN development version at http://svn.perl.org/modules/SQL-Statement.
Thanks for the notice Jeff. I can see that SQL::Statement is receiving regular attention and care these days. I will wait for your release and/or start with your SVN and then resync with your release later.
I can't say as I fully understand what you're up to, though the word "Borg" comes to mind :-). Good luck!
You're half right. I'll try to explain myself better.
Part of what I'm offering to the community is a cross-distro code refactoring of sorts, taking redundant or overlapping functionality and combining it into an elegant shared code base whose improvements can be shared by all.
The experiment right now is a practical test of whether the code I already have in the form of SQL::Routine is suitable for use by the refactored code, and if not then to determine what functionality it still needs to gain in order to do that (very little I expect).
Note that, while I expect the ports to be otherwise completely backwards compatible, there are a few exceptions to this.
1. They will only work with Perls 5.008001 and newer, since that is what SQL::Routine itself requires. Or rather, it claims to require this; my module may still work with older Perls back to 5.006, but I'm not testing those. Therefore, my refactoring will also strip any code that is strictly for older Perls such as 5.005.
2. Any C/XS components will be stripped out, if they exist, so the result is Pure Perl; if a module requires those parts, then I will either work around it or not refactor that one.
Afterwards, if any module refactoring is successful, and the original author wishes to adapt the changes into the official/standard version that they maintain, then I am willing to disclaim any ownership interest in the modified code (requesting only credit), which keeps the legal matters simplest.
Even if a successful refactoring is not wanted by the original author, then I will continue to maintain it for awhile as, if nothing else, it and its test suite will effectively comprise a larger and more thorough test suite for my own module.
-- Darren Duncan
