[wp-hackers] Improving Plugin (and Theme) metadata

[Jacob Santos]

Hmm, I think we both agree that the new way is totally awesome. I just 
have to once again disagree, but let me explain.

I don't think the current way is going to be replaced. It would be crazy 
to go from hundreds, if not thousands of plugins to only a handful, as 
I've stated before. Therefore, it makes logical sense that the current 
method will be supported for the next few versions. Given WordPress 
awesome backwards compatibility record, I don't believe the current 
method is going anywhere within the next few versions. Perhaps you are 
going to deprecate it, but it seems illogical and unreasonable to 
completely remove it.

I don't believe it is going to be removed, but I have no authority on 
that matter, you do! However, I'm sure you'll see the reasoning. Right 
now there is a chance to fix it for WordPress 2.5, and let me explain.

1. The patch is simple, should not cause any huge problems, once the 
issues are addressed and trust me by this weekend they will be. The 
counter to this is that by adding this feature, WordPress will then have 
to support it and supply fixes for something that is going to be 
eventually deprecated. That I can understand. However, I would suggest 
that any major defects be marked as "won't fix" until the next solution 
is implemented.
2. The patch will get plugin developers ready for the eventual improved 
solution, which also uses gettext. So you have a fix for WordPress 2.5 
that uses gettext and to the user, it will appear to be a bump up in an 
already established feature.
3. We don't have an implementation right now that replaces the current 
behavior. My interest is almost diminished to be honest, so I'm unsure 
if I will devote time to a project that won't see the light of day until 
probably June or July. That is only if it is completed by April. If an 
implementation is not completed by April, then it is unlikely that it 
will get into 2.6, so it will be skipped to 2.7. By then interest will 
be completely diminished and as there probably won't be a huge interest 
in this feature, it may completely die before 2.8 comes out.
4. The current method isn't going away, so it might as well match 
functionality for internationalization for developers as much as possible.

I'll say, that for myself, I would rather have a fix in 2.5, in the 
event that the worst case comes to pass. Not that I don't have faith in 
your skill, but that I don't have faith in anything that is so unsure. 
Plus, there have been several times that interest on a project has 
peaked and then died shortly thereafter with barely a whimper (of 
course, I say that after several months of finally doing something about 
an issue that has peaked and died many times).

You know, if you are planning on getting this awesome feature into 2.6, 
then I'll totally bow to you and will build an idol in your image. 
However, do consider my arguments on why something should be done in 
WordPress 2.5.



Peter Westwood wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Jacob Santos wrote:
> | I say, I think after Matt's announcement, we are going about it 
> slightly
> | off.
> |
> | I think we should fix the current localization issues with the current
> | base, since both issues have patches which with a few slight
> | modifications can fix the issues for 2.5. The current discussion can
> | make it into 2.6, and by the time that happens, probably will have an
> | awesome solution.
> |
> | The patch for #3089, only needs to have all __() functions switched to
> | translate(), so that it doesn't pollute the WordPress catalog. The
> | current patch by Hodgdon works, except for that one issue. Once that 
> fix
> | is committed, it will remove that ticket from our minds for 2.5 and we
> | can move on implementing a better way in 2.6.
>
> I am not comfortable with introducing one solution for localisation now
> when we are already talking about replacing it with something better.
>
> I think we would do better to push the localisation of metadata for
> plugins (and themes) out to 2.6 so that we can develop a good all-round
> solution.
>
> Introducing localisation for this metadata is going to cause work for
> plugin and theme authors to make use of it.  We do not want to give them
> more work by changing how it works in the next version - we do better to
> do it once in a way that works best and we are going to support forever.
>
> | The patch for #5651, only needs to be completed and should reduce the
> | memory load on big single file plugins. Add in the fix for #3141, and
> | you have the recipe for something that fixes three central issues, two
> | with memory usage and one with localization of plugins. I think it 
> would
> | make 2.5 even better than it already is and will mean a lot to those
> | that use the features.
>
> Optimising the current code for less memory usage and better performance
> is good for 2.5.
>
> westi
> - --
> Peter Westwood
> http://blog.ftwr.co.uk | http://westi.wordpress.com


-- 

Jacob Santos

http://www.santosj.name - blog
http://funcdoc.wordpress.com - WordPress Documentation Blog/Guide Licensed under GPLv2

Also known as darkdragon and santosj on WP trac.