[wp-hackers] A thought about wp-rss2.php

[Roy Schestowitz]

Quoting Luc Saint-Elie <lstelie at gmail.com>:

> Hello,
>
> It's just a thought, not a proposition of some code.
>
> wp-rss2.php conception relates to an area when RSS feed were just a
> new, high tech way to access the weblog content.
>
> Today things are rather quickly evolving. I'm pretty sure a poll about
> the amount of RSS access among weblogs would be surprising
>
> In my case current situation is shown here :
> http://blog.saint-elie.com/images/viedusite/rss_access.jpg (as you can
> see 50% of my visitors arrive by a RSS way and I doubt to be the only
> one)
>
> The day when the first "browser" will be a RSS reader is not that far.
>
> The point of this message is that RSS is a higly delicate area, far
> more that HTML/XHTML.
> A very light error breaks the readers and so on.
>
> WP is a dam good HTML/XHTML machine, probably the best one... but a
> very average RSS machine.
>
> Hacking wp-rss2.php is a very risky task... and for an average wp
> install there is a lot to hack (to add creative commons namespace and
> tags, now to add iTunes domain name and tags for podcasters and so on)
>
> It would perhaps be a not so bad solution to have a system that output
> a static xml file updated each time a new post is isued, in the spirit
> of GoogleMaps plug ins.
>
> This way, correcting a badly tagged output would be easier and it
> would be possible more easily for users to adapt their system to new
> propositions. For example the recent iTunes Podcasting story as shown
> two things : the first one is that a huge lot of podcaster use
> WordPress (I'm one of those), the second is that these guys have to
> struggle a lot to suddenly adapt their install to the new situation.
>
> My two cents
>
> Luc

You are right in suggesting that wp-rss2.php is very fragile. However, if you
choose to muck about with it, then you might as well save it under a different
filename, syndicate it for a couple of days and if all goes well, 
overwrite the
old wp-rss2.php with your experimental revision.

Moreover, I suggest you look into the possibility of dumping 
'off-the-fly' site
summaries onto text files. Some systems use this approach, which also allows
people who are unfamiliar with RSS to syndicate sites. Text files can 
sometimes
be more manageable, although less semantically proper. If you then wish to
syndicate static HTML pages, you might as well have a look at code that 
I wrote
recently: http://www.schestowitz.com/Software/Syndication/

Regards,

Roy

-- 
Roy S. Schestowitz
http://Schestowitz.com