[ewiki] Re: Custom link formats

> If that plugin is missing from your installation, then it probably
> was added short after the last major release (R1.02a). There were
> a few +dev versions since then, but in case those also don't contain
> said plugin, just get the CVS tarball (made a fresh one five minutes
> ago) from  http://erfurtwiki.sf.net/downloads/CVS/

Thanks. Targets looks like it will work perfectly for me. The Syntax is a bit 
verbose for my tastes, but I don't think I'll be doing enough of these links 
to justify hacking up a shortcut. I'm still curious as to why my code didn't 
work, though. I understand the http part, but I'm baffled by the duplication. 
NOt a big deal since I won't be using it, but if anyone can explain what's 
going on, I'd be grateful.

> Thanks for the hints on the _SELF target. And if you could contribute
> parts of your customized site as sample layout wrapper back (or post
> it on our site), that would also be very great!

Certainly. I'm happy to contribute anything I can. I've also been working on 
an improved mpi_include that allows collapsing of the sub pages. It's occured 
to me, though, that the same thing can be done in the browser using CSS. I 
don't know much CSS & no javascript, but I should be able to hack this 
together & it seems like it will be a better solution then doing it all in 
HTML.

> > The favicon is indeed a bit more complex, and the reasons you mention
> > make it even more so. My plan is to download each icon once, then cache
> > them locally. This should greatly reduce the performance problems that
> > you mention. I'll probably have the server refresh the icons every 90
> > days or so in the event that they change, but this shouldn't be a
> > significant problem.  But the added complexity is why I tackled the
> > targets one first.
>
> The longer I think about it, the more I come to the conclusion that
> this feature should be implemented outside of ewiki. A cron script
> and a separate SQL table was probably the best solution. You then
> would only need a small wiki plugin, that either retrieves the
> favicon image url from the database, or simply tags it for later
> retrieval if it hasn't been there yet. No need to actually cache
> the image files IMO.
>
> If it is implemented more separately, such a feature and database
> table could be used by more applications, because there is no need
> to duplicate the effort in each application. What I could think of
> is something like a table named 'favicons' shared by ewiki, blogs
> and other software; and a simple cron script taking care of filling
> it and keeping it up to date. (emmh, sounds like yet another sub
> project...)

That's not a bad idea. I had assumed that the cache would be refreshed via 
cron, but you're right that just making a standalone php script to fetch & 
store the favicons, and that can be called from cron to refresh the database 
(or possibly a perl or shell script, since not everyone has php installed so 
it can be used from the shell) would probably be a useful tool. It should 
also be fairly simple to put together. I'll see what I can do...

Thanks,
Mike

Other related posts: