[maemo-developers] maemo wiki survey

[From: Henning Sprang]

Quim Gil wrote:
> Dear maemo contributors, can you please answer these questions?
> 
> - What do you think about the current wiki tool at
> http://maemo.org/community/wiki (leaving content aside) ?

I'll answer this in the section good things/bad things below :)

> 
> - Are you using it? Did you use the previous one?

I used the previous one and invested some time to document the Java
support (which was poor at that time - it still is, but getting better
with the jalimo project).
Then I was busy for quite a while, and came back to see the wiki system
being changed.

As the site search function seems not to include the wiki, I di not find
any information about Java anymore, and thougth my page was deleted in
the conversion.

You can imagine how happy that made me :)

After looking a while, I found that the page was stored under another
name than I thought (JavaLanguage instead of just Java), and because the
search didn't get that, the page was taken into the new wiki, I just
couldn't find it.

Having found it, I realized, that it was not correctly converted to the
new syntax and looked quite bad. It seems that lists are limited to two
levels, for example.

All this took me ages, which had better been used updating the java
information in the wiki, or trying how Jalimo works, and report that on
the Wiki.

> 
> - Are you thinking of using the community wiki?

If the wiki will not get more usable, and other things I will mention in
bad things, I will not use it a lot anymore. I will document the stuff I
find out about Maemo in my own personal wiki.

> 
> - Mention good things / bad things of the current wiki (and what would
> you do).

Good things:
- garage login and Wiki login are the same (this is very minor, because
I have so many logins and passwords, one more wouldn't matter. plus,
this could have been done with any other wiki, too)

Bad things:
- it takes nearly a day(at least quite some time, mor ethan 5 minutes,
and eventually I did stop checking and waited until next day) until
login data is synchroized from garage to wiki.

- the wiki has no search function anymore

- the floating menu is not easy to use and intuitive

- I muss a contents entry of the headlines in a page automatically
generated for long pages with many headlines.

- the Wiki Markup is limited(e.g. why is there a limit on the level of
cascading lists?), and different from the ones most people already know
(MediaWiki and MoinMoin). Why force people learn another markup, why try
to reinvent the wheel by implementing even another wiki in midgard
instead of using  existing technologies? You see how it ends: you have some


Finally:
When I first heard of the midgard system and treid it, now nearly 10
years ago, I liked it a lot. It's a nice application development
platform (only I could never really use it in production, as it involved
having a rootserver and patch apache, which I had not - so in the
meantime, I rather work with java stuff and avoid php in general).

But I don't see the reasons why the decision had been made to move not
only the Website CMS(which was goof to move), but also the whole Wiki
system to Midgard.
If you have a Wiki running, that is used by a lot of people, you cannot
just switch it so easily. The internal CMS of a website, used by a
handful of people is something different.

The only upside it _could_ have had to integrate CMS and wiki in one
midgard system, would be better data integrity and unification.

But exactly this is not happening! Otherwise, the search function you
see on each page in the system (also on wiki pages,) would actually find
 wiki pages, but this does not happen.