The Future of XML Plugins: The Planet Plugin for Nucleus
Monday, 13 December 2004
Version 0.14 will probably the last release of XML Plugins. I am working on a new plugin called Planet which will eventually replace the separate RSS/Atom Reader and RSS/Atom Aggregator plugin.
The Planet plugin, which is still in development, will be able to do the same thing as the current plugins and much more. This is what I have in mind so far:
- A proper administration interface to handle all your feed subscriptions, with reports about the type and version of the feed, parsing and retrieval errors, cache status and more,
- The ability to import and export OPML files,
- Database based caching of individual entries,
- The ability to manually force an update of the feed,
- Set the interval of cache updates for each individual feed,
- By default honour the interval settings if they are specified in a feed,
- Retrieve feed updates through a cron job,
- Retrieve feed updates in the background through website requests,
- The ability to output headlines only, or the whole feed including contents,
- Multiple aggregators and single feeds,
- The ability to choose from separate templates for each aggregator or feed.
The Planet plugin should basically include everything you need to either simply place headlines on your blog to a complete planet such as Planet Mozilla.
The Planet plugin will be released in about a week or two. I completely rewrote the parsing engine which now should more compliant with strange feeds and should be much faster. The basic database cache handling also works and should evolve quickly to the feature set I outlined above. The engine is able to retrieve feeds in many different encodings and in all RSS flavours and Atom. The engine also supports the most common extensions.
Once I get the administrator interface working I’ll post some screenshot. In the mean time, I welcome any suggestion you have.
Sounds very interesting plug-in. Does it suport utf-8 properly as well?
This just sounds great. :-)
I even haven’t been able to put the original plugins to good use yet (I’ll find a way to integrate it into my design somehow), but am already looking forward to the Planet plugin
Mehran: Yes it supports many different encodings, including UTF-8.
A little demo of the capabilities of the engine:
http://www.rakaz.nl/project…
That is starting to look very nice, rakaz. I wouldn’t mind test driving it ;)
Brilliant work .. Count me in for the test driving team :)
A suggestion : Can you make all the settings (especially the chosen feeds) per member ? If done correctly, then maybe the planet plugin can be used to make a bloglines clone, is it not ?
How is this plugtin coming along?
I’m ready to test it, if you are? =)
As you probably noticed, this plugin is a bit delayed, because I wanted to make sure that RSS or Atom feeds can never cause problems on your weblog by providing broken, dangerous or even malicious code. So I needed a HTML parser that can filter out those unwanted elements and make sure the HTML conforms to one of the (X)HTML standards.
That parser was coming along nicely, but I put it aside for a moment to focus on some other things, such as a rewrite of the SkinFiles plugin, a new plugin which will be announced later this weekend, and bringing proper UTF-8 support to Nucleus itself.
I do expect to pick up work on this in a while. Once the HTML parser is finished, the Planet plugin will be also.
I’m very excited about this plugin and it’s preview. I built a very simple news aggregator myself without touching the Nucleus, but since I run Nucleus on the site, it might be reasonable to wait for your plugin.
I’m worried about the support of various codepages (like aggregating 5 different RSS feeds with Win-1251, ISO-8859-1, ISO-8859-2, UTF-8 and …) and specific FORCING of codepages for separate BROKEN (I mean having incorrect information – some siteops are too lazy to correct some parameters) RSS feeds.
Could you please update us on the status of this project? Does it support categories? Any integration with Nucleus users? I see it’s very neat and stable and I’m willing to betatest it on various codepages. If we’ll need to wait another month, I’m probably going to continue with my own Nucleus-independent project.
I have mail "emulynx" at the google mail. Thanks.