I have had several ideas to add to WordPress that I want to finish before I move completely to some other projects. I’m thinking if I get most of them done before 2.6 is started and 2.5 is out, they’ll be able to get in WordPress. That is a dream, and I don’t have commit access, so it is totally up to the commit team.
HTTP Request and Post API
My idea for HTTP Request and Post API is finalized, except that it has not been included into WordPress. I’m unsure why that is, since I have included, both documentation and unit tests that covers the full extent of the library. I also stripped out most of the extraneous material.
I’m about ready to do a complete and major overhaul of the system, once again and see about instead creating a set of functions which is pluggable, instead of classes, which do the job. It would be less flexible, but I think the change will be more the way that the devs would like it. It stands to be as long as wp_mail() however.
The last test I can do, is profiling the code, to make sure that it is quick enough now to work in WordPress. If I do that and I prove that it works and doesn’t create too much overhead, I’ll be seriously pissed if it doesn’t get in.
WordPress Logging API
I did some initial work on a WordPress Logging API which will capture all notices, warnings, and errors, and have them logged to an array for a plugin to later access and do what the plugin will with the array. This will first disable displaying of PHP errors and notices and will secondly, allow for better debugging, both on a live system and on a debug system.
The API is complete and I’m unlikely to do any major overhauls on the system. I think the next step is to focus on major areas that could use the API and convert them over to using the new API. I was hoping to get the API in 2.5, but I think major areas could also be switched over to using it.
I’m thinking I’ll focus more on getting the API in 2.6 and do a lot more work on converting major sections and other logging sections over to the system. I also will need to do profiling to make sure that calls don’t cause too much overhead.
Controller API
WordPress already has and uses the controller model, but it is not easy and is convoluted in many places. I would rather create a system which the WP_Query and Rewrite is combined in areas and allow for an easier model for adding pages. I would very much like this idea, but I don’t see myself creating it for 2.6 and I doubt it will make it in 2.7. I’m thinking that with the amount of time I can spend on it, it might see the light of day in WordPress 2.8, if the commit team is fine with replacing the rewrite and query model.
Single Use Form Nonce Tokens
The current nonce form token system is lacking in some areas and I would rather see some additions that allow for replacing CAPTCHAs, but are safe enough to be deployed in public systems. I would rather this also be in WordPress, so that each plugin doesn’t have to create their own system.
I have the code, however, it is not licensed for use in GPL, so I’m going to have to rewrite the code for WordPress. I am also going to have to write some unit tests also. It will require going to a session based setup instead of a cookie based one. I believe, my plans for having it cookie based also has it use sessions.
I should note that it isn’t completely secure, but it should halt most of the lesser intelligent bots. I person would be able to defeat the nonce easily, but would require a lot of work. Spammers would not be interested, hackers probably will be.
New Plugin Architecture
There has already been discussion on this and I think I’m over not being able to get it in 2.5. I think having it in 2.6 or 2.7 is fine and I’ll be willing to do work on it for 2.6 during any downtime. However, I do think that this is probably medium priority on my list. It would be fun to write, but I want to get some of the easier and more complete components finished first.
Possibly Related Posts:
- Bullet: E-Book Library Management and Content Server
- Using ZendFramework 2 beta1 For Directory Project
- The Way of Kings and Cosmere Theory
- “In Time” Movie Premise Flawed
- Completing HTTP Library For PHP
nice ideas.
PLease tell me for the last idea; what do you think about the WordPress Plugin Framework – WPF: http://wordpress.org/extend/plugins/wordpress-plugin-framework/
I think it is pretty cool. I think it would be better to take certain pieces of it and apply them to the core rather than the whole WPF.
My idea has to do with how apply_filters and add_filters and the Plugin API works rather not with administration.
Thx!