Gentoo Foundation Website Redesign Contest -- the vote

[jspr]

Aaron Shi's design is by far my favorite one, BUT! if there's one thing I really, REALLY hate about it, it's that it puts too much emphasis on "extreme performance". C'mon now people, by now everyone knows the performance gains of Gentoo's build-from-source system are insignificant. The REAL advantages of Gentoo are the flexibility, the community and the just so darn powerful tools.

Take a look at the end post in this thread: http://forums.gentoo.org/viewtopic.php? ... &start=150

AFAIK, the top banner will be chosen randomly at each visit.

http://aaronshi.com/gentoo/gentoo-front1.png
http://aaronshi.com/gentoo/gentoo-front2.png
http://aaronshi.com/gentoo/gentoo-front3.png

[Liberty]

It doesn't matter how beautiful the site is as long it is functional. Although the designs are all in there little excellent i doubt that you will render a fast loading site. Much of them, seems to me, are based on slices design ( Adobe, Fireworks ) which generates to much traffic at the act of loading. Then you should put more accent on the w3c standards, and especially the wai rules, which right now lacks heavily. Although the site is html compliant the WAI rules are missing totally. Have you ever had the curiosity to run it through a screen reader ? THIS are things to bother about. The Linux spirit is respect of the standards otherwise we are not diferent from M$

[andreask]

Aaron Shi's design is by far my favorite one, BUT! if there's one thing I really, REALLY hate about it, it's that it puts too much emphasis on "extreme performance". C'mon now people, by now everyone knows the performance gains of Gentoo's build-from-source system are insignificant. The REAL advantages of Gentoo are the flexibility, the community and the just so darn powerful tools.

As it is now, this design makes the enitre Gentoo community look like "ricers". And there are enough people who see Gentoo as one big hellhole of mindless ricers (just read slashdot!). We REALLY don't need to reinforce that belief!

You are absolutely right! I think one of Aaron Shi's alternatives will be better.

[wendall911]

I agree with a previous poster. While all the designs look nice, it is a HUGE disappointment to see absolutely no consideration for xhtml or wai compliance. No design for multilingual support. Nothing. These are poor designs from a technical standpoint.

Want to see kickass xhtml + css designs that degrade well in lynx, etc.
Take a look at http://www.mezzoblue.com/zengarden/alldesigns/
These designs not only kick ass, but follow basic design guidelines and are xhtml, css and wai compliant, and most are under a cc license or equivalent.

We should expect more than a bunch of M$ / Macromedia / Adobe site clones for top entries. And we should expect more for a main site page for Gentoo.

Wendall

[Bungopolis]

I was very disappointed with all of the designs. They look far too corporate and are designed to appeal to consumers via traditional marketing. Gentoo is not trying to sell anything. I would prefer to see a clean, well structured, and accessible site.

I especially dislike the use of stock corporate 'workplace' imagery.

As has been discussed, the most important feature of the site is accessibility via good XHTML and the use of CSS. If it is a well constructed site, the design should be interchangeable.

[STEDevil]

I agree with a previous poster. While all the designs look nice, it is a HUGE disappointment to see absolutely no consideration for xhtml or wai compliance. No design for multilingual support. Nothing. These are poor designs from a technical standpoint.

???

What makes you draw such conclusions from looking at those SCREENSHOTS?!

In fact, looking at the Aaron site I can see myself easily implenting that design using nothing but "correct" (x)html & CSS and A++ WAI compliance. In fact looking at the page I'd be very surprized if Aaron is not himself very well versed in these technologies and I'd presume he already has a "working version" of the design not just some mockup images made in Gimp.

Also, people here seems obsessed about the site having to be done with XHTML. There is literally 0 difference in compability & features provided by HTML 4.01 vs XHTML 1.0 so which one to use of those basicly comes down to the preferance of the coder(s). Further, using XHTML of 1.1 or higher BREAKS the site COMPLEATLY in 90% of the browser used (eg IE, all versions) unless the first thing you do is break the very spec you are trying to follow by sending a bogous text/html doctype with your incompatible XHTM1.1 (or higher) document. That is certainly NOT following standards...

[wendall911]

What makes you draw such conclusions from looking at those SCREENSHOTS?!

My assumption based on experience with these style of designs. You are correct in saying that they could be implemented with compliant code, but not likely be wai compliant.

There is literally 0 difference in compability & features provided by HTML 4.01 vs XHTML 1.0

Now you are just completely wrong.
http://www.w3.org/TR/xhtml1/#diffs

And here is a FAQ on why you should use XHTML.
http://www.w3.org/MarkUp/2004/xhtml-faq

If you disagree with the W3C position on the matter, that's fine, but I assure that you are in the minority. Broken issues in IE aren't a reason to not use current standards. There are plenty of sites around that can teach you how to implement proper workarounds:
http://www.positioniseverything.net/

There are many more, that's just one I happen to use.

[DJ_Max]

There is literally 0 difference in compability & features provided by HTML 4.01 vs XHTML 1.0

I have to jump in on this. If you coded correctly in both, you will notice why XHTML is better, and will soon replace HTML.
XHTML is faster, cleaner, lowers page size, etc..

The list goes on.

Further, using XHTML of 1.1 or higher BREAKS the site COMPLEATLY in 90% of the browser used (eg IE, all versions) unless the first thing you do is break the very spec you are trying to follow by sending a bogous text/html doctype with your incompatible XHTM1.1 (or higher) document. That is certainly NOT following standards...

Older versions of IE break in XHTML(5). Newer IE6, Mozilla, etc work fine.

[WeirDave]

...darn I missed the voting.

[STEDevil]

What makes you draw such conclusions from looking at those SCREENSHOTS?!

My assumption based on experience with these style of designs. You are correct in saying that they could be implemented with compliant code, but not likely be wai compliant.

Lets make a deal. You list some actual WAI-conserns you have with Aarons design and I'll explain to you how that is easily solved using half a brain and some HTML & CSS knowledge.

There is literally 0 difference in compability & features provided by HTML 4.01 vs XHTML 1.0

Now you are just completely wrong.
http://www.w3.org/TR/xhtml1/#diffs

No I'm not. What you are linking to is the list of semantical differences between the two, not feature differences. Also, I am of cource assuming XHTML 1.0 will be used according to the compability guidelines (to enable it to be sent as mimetype text/html) which in fact makes XHTML slightly more limited then it's twin HTML 4.01. If you ignore the comp guidelines you are also losing the "right" to send XHTML 1.0 as text/html (which in turn will make the page barf in the majority of currently used browsers).

If you disagree with the W3C position on the matter, that's fine, but I assure that you are in the minority.

Eh.. the W3C is definitly NOT disagreeing with me.

the second line of the entire spec

A Reformulation of HTML 4 in XML 1.0

The whole idea of XHTML 1.0 is to be as close a carbon copy/dropin replacement of HTML 4.01 as is even possible under the added restrictions of XML.

And here is a FAQ on why you should use XHTML.
http://www.w3.org/MarkUp/2004/xhtml-faq

There are also FAQs about why you should NOT use XHTML, that is why I said this basicly comes down to developer preferance.
For the record, I myself use exclusively XHTML and have done so for the last 4 years. That still doesn't mean I think less of developers that prefer sticking with HTML.

Way too many people think that you have to use XHTML or the webpage is compleat crap. It's a lot more meaningfull to judge a site on actual Accessibility & "did the coder actually implement the specs correctly according to the spirit of the specs or did he/they just hack themselfs along until it validated using compleatly inapropriate tags".

After all, you CAN code a 100% valid & 100% accessible webpage using tablebased layout, but you are still breaking the "spirit of the spec" by abusing TABLE for a purpous it's not supposed to be used for. In comparrison XHTML vs HTML is just a "my willy is longer then yours" contest.

Broken issues in IE aren't a reason to not use current standards. There are plenty of sites around that can teach you how to implement proper workarounds:
http://www.positioniseverything.net/

#1 HTML 4.01 IS a CURRENT standard.
#2 I don't think you really understand just how broken IE is with X(HT)ML pages sent with a correct mimetype. You don't get a slightly misaligned or funny looking page, you get NO PAGE AT ALL. As in COMPLEATLY INACCESSIBLE. Nothing, nada, zilch. Blocking 90% of the websurfers on the net from even being able to visit the gentoo site is a very stupid thing to do.

XHTML is ...lowers page size..

XHTML have many things that might give it a slight edge over HTML in some circumstances. However lower page size is NOT one of them. An XHTML 1.0 page will ALWAYS be LARGER then the corresponding, correctly coded, HTML page. Many (me included) think that the net gain when all things are concerned by using XHTML 1.0 outweighs the costs, but it sure isn't a landslide victory like many incorrectly seems to assume. In fact, in most cases it's basicly down to the "flip a coin" level.

Older versions of IE break in XHTML(5). Newer IE6, Mozilla, etc work fine.

I havn't tested IE 6 with XP SP2, but at least before that IE 6 was compleatly incapable of showing a XHTML page sent with the correct
application/xhtml+xml mimetype ( http://www.w3.org/TR/xhtml-media-types/ ... -xhtml-xml )
If you have new facts on the issue please provide a link to a XHTML webpage where IE6 is fed the correct mimetype by the server and doesn't compleatly choke.

[calande]

I agree with a previous poster. While all the designs look nice, it is a HUGE disappointment to see absolutely no consideration for xhtml or wai compliance. No design for multilingual support. Nothing. These are poor designs from a technical standpoint.

We work only with XHTML & CSS. No WAI compliance for now, but we can get there, I agree. Multilingual support is easily implemented using either PHP, or JavaScript, and people who disable JS (few people) aren't reject from the web site - they would have to select the own language manually, which would be acceptable. We have many references in valid XHTML code (Auriance, OSResources, etc...). I can send you the XHTML version of the proposals for Gentoo's web site if you want to make sure.

We should expect more than a bunch of M$ / Macromedia / Adobe site clones for top entries. And we should expect more for a main site page for Gentoo. Wendall

Iris onMirror is IMHO in no way similar to an international IT corporation. What was your proposal by the way?

I have to jump in on this. If you coded correctly in both, you will notice why XHTML is better, and will soon replace HTML. XHTML is faster, cleaner, lowers page size, etc..

This is just not true. If you worked in web design, you would know what the benefits of XHTML are. At least, if we use XHTML (and it's not by fun), it's not for the same reasons as yours.

This was a web design contest, and the contest guidelines clearly said that it was pure design, noting about coding or such. The Gentoo team will do it, and they use their own technology, basically Python + XML. If you don't know what Python and XML is, google for it.

Best of luck,

Charles.

[Liberty]

And why not the WAI rules. I have a lot of friends that are visual impaired or even blind that tells me that accessing the site is quite difficult, because there is no support for them. What message do we send if the site can't be accessed by everybody. And i mean everybody. As it stands now the site will need minor corrections which would take no more than an hour. Is this so difficult to be done ?

[Proteus]

I think it is a good idea to have the site accessible to everyone as much as possible.

But - If the corrections are only minor ones and can be easily done why don't you do them and submit them to the website-team?

I can't think of a reason such input wouldn't be welcomed.

[calande]

I think it is a good idea to have the site accessible to everyone as much as possible.

But - If the corrections are only minor ones and can be easily done why don't you do them and submit them to the website-team?

I can't think of a reason such input wouldn't be welcomed.

That's what I was gonna say. That would be beneficial to list the modifications both for the new web site, and also for our personal benefit.
That'd be interesting.

[STEDevil]

We work only with XHTML & CSS. No WAI compliance for now, but we can get there, I agree.

Actually, the by far easiest way to get WAI compliance is to design for it from the ground up. It's MUCH harder to take a "finished" site and try to remold it into an accessible site later.

It's a very unfortunate missconseption that making sites with high accessibility is very difficoult. It's not, and frankly many times it's right out simple, but it does require reading though the W3C guidelines so you always have that knowledge in the back of your mind when you make designdesisions.

If you start from scratch, using proper (x)HTML (eg using STRONG, EM, CODE, DL, etc where apropriate), WAI almost comes at no extra effort at all. One just have to fully utilize the HTML spec by using the proper tags for the job togheter with a logical structure and you are almost there.

Thus the "proper" procedure is to first make a page containg the content only, no extra formating. Once you have a logical structure that works well in eg Lynx, then you can start adding CSS magic to move things around and make it pretty in graphical browsers. Any other way to do things is going to end up a lot harder to create and often also to maintain. Additionally, "CSS-ing" a site is a often a lot easier if you have a good solid (x)HTML structure to build on.

I don't have time to put in a real effort on this myself, but I'd be most happy to offer advice & suggestions of how to start and what to think on to get you (the devteam) started in the right direction. I can even throw togheter some quick mockupcode of a page that shows how it could be done (but in 2-3 weeks, since I'm compleatly overloaded with work presently).

[mcrawfor]

I think this choice was a mistake. Iris on Mirror seemed the best to me, simple, to the point, yet professional and pretty.

Purple color be damned. If we use the frontrunner Gentoo is going to look corporate and flashy. Not a good first impression for a serious-minded open source project.

-miles

[Proteus]

Well, it may be a mistake to you. - But to me it's not. (I voted for Aaron Shi.)

Just because some decision does not suit your mind it is not a mistake.
That would also be true for me if Iris on mirror won.
(And I understand that it is easy to say for me because my favorite won...)

Also, I still stand by my opinion that a website with a flexible, choosable design would have been the best decision.

[Liberty]

I think it is a good idea to have the site accessible to everyone as much as possible.

But - If the corrections are only minor ones and can be easily done why don't you do them and submit them to the website-team?

I can't think of a reason such input wouldn't be welcomed.

That's what I was gonna say. That would be beneficial to list the modifications both for the new web site, and also for our personal benefit.
That'd be interesting.

Molto bene !
I'm gonna try to find an hour to do it, and after that i'll send it to you.

[gilkyboy]

Although I missed the voting, I do think there's something to be considered when choosing an appropriate candidate, and that's usability. Most of these seem usable, to us, the gentoo devout, but we already know all about gentoo and that we can simply access the forums by typing forums.gentoo.org, or the packages lists with packages.gentoo.org. The homepage should instead cater towards those who will be considering. In light of that, I think it's definitly most important that useability comes first, and I think if I were just visiting to check it out, I'd be turned away by all the flash on most of these sites, they don't seem as homely as our community. I think first and foremost, we need to consider those needs rather then what looks good. I personally think the one that caters best to those needs is Derek Gerstmann's. It has a homely feel and is warm and inviting rather than flashing everyone away - it seems too big business.

[scharlach]

I missed the voting, too. But apart from issues concerning compatibility/scalability/performance ... for me the most annoying thing is the fact, that the "winnig" theme is just another suse.com clone.

[Lucero del Alba]

The winner of the redesign contest has just been announced on the Gentoo homepage.

I just wanted to thank the Gentoo developers and the community itself for giving us the opportunity of competing in this contest.

Gentoo is going to have quite an awesome design on their sites soon, congratulations Aaron

[Apreche]

Yes, this is definitely excellent. While it wasn't my first choice it is definitely better than the design we have now. But I believe that other people agree with me that the infinity sign oo at the end of Gentoo has to go. There can still be small changes right? We aren't going to just use the design as-is?

[jezebel]

the newb would like to make a few points:

1- the winner's design looks almost exactly like a design made for a company i worked at 3 years ago. which looked almost exactly like a lot of other designs made for a lot of other companies.

2- WAI and other standards need to be followed, or what the hell are we doing here, anyhow?

3- completely. not compleatly.

[scharlach]

[...] 1- the winner's design looks almost exactly like a design made for a company i worked at 3 years ago. which looked almost exactly like a lot of other designs made for a lot of other companies. [...]

Yes. And again: since it is not just looking like "any company's" design, but like the one another (inferior) linux distribution is currently using, i think it's extraordinary unlucky. But things go better with coke and a million flies can't be wrong ...

[STEDevil]

the newb would like to make a few points:

1- the winner's design looks almost exactly like a design made for a company i worked at 3 years ago. which looked almost exactly like a lot of other designs made for a lot of other companies.

The reason a lot of companies use similar designs is becuse they work very well in exactly the area where the current design (and iris, which is a simple carbon copy of the current design with just a few color & fontchanges)suck the most... INFORMATION OVERLOAD/Navigation.

A first time visitor to the current gentoo page will be compleatly lost in about 10 seconds becuse there are a myriad of unorganized links on the front page and even worse NO WAY BACK once you click on a link. Absolutely horrible and the very first NO NO in any basic "how to design a website" course.

The only reason many people here prefer the old (or Iris, which is the same) is becuse they've "learned" where to find stuff after visiting the site perhaps weekly for a few years...
You should NOT have to LEARN a website to find your way around it...

again: since it is not just looking like "any company's" design, but like the one another (inferior) linux distribution is currently using, i think it's extraordinary unlucky.

The design doesn't look the same as Suse's, like many ignorantly keep posting. Some of the LAYOUT does (design is the compleate package, not just the layout) ie the logical grouping of links on the front page.
In fact I think the Suse site look horribly ugly while the new Gentoo looks very cool indeed IMO.
I also keep wondering why people automatically dismiss greate site navigation just becuse some other Linuxdist happen to use something similar. Are you guys thinking straight? Why should the site deliberatly be made aqward and a navigation mess just to be unique? At least in my book, userfriendly beats unique on a website anyday... please suggest uniqueness in areas that doesn't go straight against usablility instead.