Re: PostgreSQL website redesign

Le 20/11/2011 12:05, Thom Brown a écrit :
> On 19 November 2011 14:29, damien clochard <damien@dalibo.info> wrote:
>> Le 17/11/2011 01:10, Thom Brown a écrit :
>>> Hi all,
>>>
>>> As you will no doubt be aware, the PostgreSQL website is overdue for a
>>> fresh lick of paint,
>>
>> First of all, thanks to the people who handle this. This website needs
>> love that's for sure :)
>>
>> Here's a few ideas :
>>
>>  * I think target audience is new users. The website should focus on
>> providing simple answers to basic questions :
>>
>>   - What is PostgreSQL ?
>>   - Where can I get help ?
>>   - Where's the download link ?
>>   - Who uses PostgreSQL ?
>>   - How can I contribute ?
>>
>>  * Regular and advanced users don't need to go on the pg.org website
>> everyday. For them the website should only be a gateway to the various
>> community websites : the wiki, the planet, the pugs, pgxn, the mailing
>> lists, etc.
>>
>>  * Be more international, connect with local communities. If the main
>> website is not translated, at least it should make it clear **on every
>> page** that there are local websites in other languages (pg.fr, pg.jp,
>> etc. ). New users that are not comfortable with English should have a
>> chance to read content in their own language (when possible of course).
>
> We can certainly bear that in mind.  However, that's more of a website
> content issue rather than design.  This is an issue that needs
> addressing though.  While the project's default language is English,
> we have large German, French, Japanese and Portugese-speaking
> communities, so sign-posting users to sources of information in their
> own language would be useful. But the designer won't want to hit a
> moving target, so we will be freezing structural changes to the main
> website at least until the design has been implemented.  They can,
> however, provide their idea of how that would be presented though.
>

Pointing to non-english communities on every pages is not a big
structural changes :) It just requires some links in header or footer.
Any international corporate website has either langage-switching buttons
or links to the local offices.

However i understand the fact that the structure should be freeze before
the designer starts. It's common sense. But in the same time, if we
don't change the content itself, the new design will have a limited effect.

This was the biggest problem with the 2009 webdesign contest : people
were not allowed to modify the existing structure and content (except
for the front page). This was a mistake. It made the work very hard for
contestants and eventually the contest failed.

I really think that the website needs much more than just CSS magic. I
mean you can change the paint and add shiny stuff, the website will
still look archaic if the content remains untouched.

The goods news is that the structure and content just need some
simplification. It's not a lot of work. There's no content to add...
We just no removed and transfer pages elsewhere because the website is
too deep. There's way to much information. Like Joshua said many pages
should be transfered to the wiki : The "proganda" page, the history,
almost all the developer section, etc. All of this belongs in the wiki.

Reduce the complexity of the website will have many advantages : that's
less work the webmaster team, the designer will have more possibilities
and less content means there's lower risk of outdated content, etc.

So don't make the same mistake again. This is not just a CSS problem. In
fact, i'd rather keep the current CSS than leaving the structure
unchanged...

>>  * If think the message should be that PostgreSQL is free, innovative
>> and used by very big companies. A slideshow on the front page could
>> illustrate these 3 points : talk about a new feature in 9.1 , display a
>> use case, advertize an event such as pgcon, etc.
>
> I like that idea.
>

Yep at the same time : almost everyone does that kind of slideshow right
now, it's almost a web design standard... which means it may seems
"old-fashioned" in a few years ;-)

But a front page slideshow has lots of advantages... It keeps the
visitors a little longer on the front page and gives the feeling that
the website is dynamic (provided the information in the slideshow are
not out-dated of course... )


>>  *  I don't think the website should be optimized for tablets and mobile
>> phones...
>
> Agreed.  The site isn't for casual browsing, so no extra effort should
> go into catering for touch-friendly interfaces.
>

I also suggest we stop supporting IE6 and it'll save the designer a lot
of time...


>> c/ Sorry but I'm not a huge fan of the Postgres Open website :-/ It's
>> full of HTML and CSS errors everywhere. The meta tags are pretty poor.
>> Basic files like sitemap.xml or robots.txt are missing. There's no
>> favicon either....  On what basis is this an "excellent job" ?
>
> Problematic meta tags, dodgy HTML and CSS and sitemap.xml and
> robots.txt aren't a designer's job.

Seriously ? Are you saying that the web designer will not deliver
HTML/CSS code ? that he won't be responsible for the new website SEO  ?



--
damien clochard
dalibo.com | dalibo.org