[openbeos] Re: new website, part 2

On 8/26/06, Charlie Clark <charlie@xxxxxxxxxxxxxx> wrote: 
Am 26.08.2006 um 14:44 schrieb Waldemar Kornewald:
> Right. CMSes like Bricolage have a dynamic backend which creates
> static HTML pages for the frontend. I guess they are using something
> like that, too.

While I share Jonas' scepticism regarding "Web 2.0" (which we all
know is a badger!) having static files does not mean not having a CMS
for managing the content if not serving it.


Are you replying to me? I never said this.

Human readable URLs can
be achieved through mod-rewrite.


They can even be achieved with a plugin, but I have two little
problems with Drupal:

1. Will it slow the system considerably down? Every lookup will take
longer and the official Drupal site doesn't use the plugin, either.
Maybe it's a minor issue.

2. Users should still use the node ID when linking to other pages.
Unfortunately, Drupal doesn't make this easy because you have to enter
URLs manually. How can I make sure they use the node ID instead of the
human-readable URL?
Why is the node ID so important? When you rename your article the URL
will change, so all links to your article will get invalid. That is a
*huge* mess and is MUST NOT ever happen on our site. In order to make
this 100% clear it might be better to not introduce human-readable
URLs with Drupal, at all.

If *everyone* promises (unrealistic!) to use the ID I can enable the
scheme "<title>-<id>" and everyone will have to use the ID at the end
of each article's URL when creating links. But is it realistic to
expect that people always use IDs (at least for the most important
articles)?

It's a limitation of Drupal and it would take *a lot* of work to fix this.

But I also wouldn't agree that
FreeBSD is a good example of a website - the online handbook apart -
some of the most useful content is spun off onto other sites such as
freshports or the freebsddiary (this might serve as a model for some
content for Haiku). Like the PostgreSQL website it suffers a bit from
legacy content but both projects are worth looking at for the way
they structure content and host their own mailing list archives.


The FreeBSD Diary comes much closer to what I had in mind for our
articles than a simple blog, but it's still too different from a
newsletter. I wanted to have something in-between a newsletter and the
FreeBSD Diary.

I don't like the PostgreSQL website. The most important pages
(Documentation, for example) are populated with so many links that you
can hardly find what you search for (e.g.: the FAQ).

I think that the real problem is that the main pages on our website
are not yet populated with enough content and links to related pages.
For example, if you want to download the development tools you will
probably click on "Downloads" link. I've fixed that and a few other
minor issues. IMHO, if the inter-linking of important pages is done
correctly the current scheme should work very well. Please continue to
add suggestions.

Bye,
Waldemar Kornewald

Other related posts: