Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 7 Oct 2005 22:48:21 -0500
From:      Eric Ryan Harrison <erh@theimpossible.net>
To:        freebsd-www@freebsd.org
Subject:   Re: New website
Message-ID:  <20051008034820.GA30150@theimpossible.net>
In-Reply-To: <20051008013442.GX72352@cirb503493.alcatel.com.au>
References:  <20051008013442.GX72352@cirb503493.alcatel.com.au>

next in thread | previous in thread | raw e-mail | index | archive | help
I'm replying to Peter instead of crafting my own message
because I think he's touched on a few good points and I'd
love to just kinda throw my opinion into the queue as well.

Peter Jeremy (2005-10-08, 11:34):
> Whilst the new website looks OK on my laptop at 800x600, switching to
> my desktop reveals a number of serious shortcomings (as noted by other
> people as well).

I _personally_ think the new website looks freaking awesome.  But then
again, I use an average browser at an average resolution.  

> 
> 1) If your browser is not in a window that is roughly 800px wide, you
>    either need to scroll horizonally or have whitespace on the edges.
>    The old website resized automatically so this is a serious regression.
>    Looking at Emily's proposal PNGs that were posted recently, this is
>    far more obvious there.

CSS/XHTML is definitely a step forward.  However, it has been said numerous
times already, but web design is fluid.  The only thing that is _EVER_ certain
in web design is that none of your users will EVER render the same thing
the same way.  Your best option is to GIVE the user options and let it run
from there.  Percentages are the way to go here.  You can still achieve the
exact same look and feel with a percentage value that you do with all of
your fixed sizes, but this will make the page accessible to everyone from
the dude who's in a 80 character wide console to Mr.Expensive-Pants who
is browsing on his 5 foot HDTV display (or whatever you rich fools are 
buying these days).

> 2) The default text size is too small.  As other people have stated,
	-snip-

Text sizes should NEVER be set by a website.  Sure, it's nice to have
a 12px font offset your beautiful 14 px improperly used H3 elements
if you get off on that sort of thing.  But FreeBSD's website purpose
should always be to provide information.  Look at the first four letters
of your name.  'Free' (for those of you not good with substr()).  Sure, 
we can argue semantics all day about the definition of the word Free,
but last time I checked, FreeBSD didn't cost 300$ per license.  So
give out information about what you're doing and why you're doing it.
Be 'Free' in all things.  This is the ultimate form of Evangelism or
Advocacy or whatever snappy word you use for "Getting People To Use
The Stuff I Made."  You give away the OS.  You give away the documentation.
You just make things VERY difficult to get to for all users, and therein
lies the problem.

> 3) Visited links are no longer displayed as visited.  Another regression.
> 

This is a dead horse.  In 1999 when I first started monkeying with some
CSS stuff, I too was awed by the way I could do all kinds of stupid,
illogical things with links.  I could change the text to whatever I wanted
when I hover over it and the WORLD WAS AMAZED!!! (emphasis mine) at how
utterly irrelevant I was in the scheme of things.  FreeBSD is the BEST 
operating system I've ever had the pleasure of using and I want the 
website to reflect this.  

I have never read a single usability report that stated "In order to
make your site EASY to navigate, please make it so that your users
have no clue which link out of the hundreds of thousands of pages
on your site they have visited or clicked on."  And the sub-heading
to that fictional usability document would probably say, "Bonus points
for someone who DOESN'T change the color of links based on history AND
ALSO changes the text FOR a link on different pages THROUGHOUT the site
so that the users will never REALLY know where they are or where they
are going."

> I also think that dedicating about 50% of the home page to news,
> events, press releases etc is overkill.  More space should be
> allocated explaining what FreeBSD is and why you should read more
> about it.  
Whoa!  Hold the ship here!  I completely and utterly disagree with
you Peter.  Not because you're wrong.  It's a difference of opinion,
but thanks for giving your opinion clearly.  That's good form.

For me, I _really_ love having the home page be more than just a 
series of static documents that never change.  It's important to 
have all your necessary links, but I also think that it's essential
to website outreach efforts to also make the website's frontpage
do more instead of less.  This issue is a 50/50 split it seems between
usability experts, but I'd be willing to place money on more eventually
winning out.  I agree that the news section is poorly designed and
doesn't really work in the most efficient manner, but I fall under the
'more' category of web design.  The oldschool users usually are the ones
who dislike the 'more' idea, because they're used to the old page.

But after a while, in websites, your power users generally will 
start to bypass the main page anyway, so they are not the best
audience to listen to regarding what content the front page should
have.  I've seen some great things being posted about things that
were missing on the page (Handbook link, etc) and those were good
catches.  But after a few months, the power users will be typing
in the URLs to the pages they want to go to, and their opinions
on the frontpage design become moot.  
 
> Based on the new front page, I don't see anything that
> would make me bother to click through to the "learn more" - the
> existing text does not distinguish it from any other *BSD.

FreeBSD is a specific niche product.  A techy nerd type (which 
most of us probably are) already KNOW what FreeBSD is and 
what differences it has to other *BSD's.  We're not the target
for outreach.  We're already customers to the grand vision of
FreeBSD.  But completely clueless users won't even know what
a *BSD is or why they should use it.  They want to know about
how to get on 'The Internets' if you don't give them an Internet
Explorer link RIGHT on their desktop. 

The grand question that the webdesign team needs to ask themselves
is:  "What the heck ARE we trying to do?"  

It's great to say, "Hey, we're FreeBSD's website and we need to 
have the best website on the whole planet."  That is a very 
noble goal, but that never works out (unless you're MySpace.com
(but I seriously believe that's just a combination of strange
solar movement patterns affecting the brainwaves of American's
as a whole and will later be cured with a new additive to tap
water yet to be invented)).  

You need to decide who it is you are targeting with your website.
If you are targetting Systems Administrators already running
FreeBSD, then maybe http://freebsd.org should be a link to
the -CURRENT,-STABLE,-RELEASE kernels and a table of contents
for the Handbook.  If you're trying to get Microsoft Windows
users to give FreeBSD a shot on the desktop, maybe you need to
have a whole bunch of step by step walkthroughs that explain
from the get-go on how to do each and every task a regular
user would need.  Of course, if that's your goal, I think 
maybe you need to redesign the OS itself. ;)

I've always thought of FreeBSD as the fast, stable, and infinitely
configurable operating system for people who needed an 
operating system they could depend on.  I've used FreeBSD in
almost every environment that I've ever had to depend
on any sort of server for any sort of mission critical function.
FreeBSD has never let me down yet.  But if your sole product 
(the OS) is targetted towards systems administrators who 
already are capable with Unix systems and theories, then
having a website geared towards home users is hurtful to your
home users (who will be sold by the website and be kicked in
the junk by the product) AND the unix-heads (who will love
the product but get screwed on a crappy, featureless website).

The main topic behind this block of information has been
this idea:  Please, for the love of uid0, decide what it
is you are trying to accomplish and do that and only that
until your ability to accomplish your mission is so well
handled that branching off into wild ideas that may or
may not actually work won't do anything but good.

You're not at that point yet, and while I love the new 
look of the website in comparison with the drab, 1994 
looking version that made me think of "Netscape 4.0!",
I think it needs to be refined.
 
> The 'Donate' and 'Contact' links at the top need to be better
> distinguished from the 'text size' links.  My suggestion would be to
> move the 'Donate' and 'Contact' links into the main menu bar (and
> get rid of the need for the text size links by not breaking the
> text size to start with).

I agree with the breaking text thing.  Text sizes are absolutely
not supposed to be in the realm of something the web team should
want to control.  Use your <p> tags to denote paragraphical text.
Use your <li> tags to denote lists.  <h#>'s for your content headings
and then let the client decide what it is that he/she needs 
to read the website comfortably.  This is one of the only things
that all browsers actually do well.  MSIE might not handle CSS 
padding syntax, Lynx will display boring flat text, your Firefox
users are using Greasemonkey to integrate your webpage into a scrolling
RSS feed that they're dumping to Gmail and integrating into their
filesystem using a GoogleFS port, so the only sane option is to
say "screw" it and don't set your text sizes anywhere. 

Just s/font-style:\n//g and be done with the whole she-bang.  
Your website will be better for it. 

> There has been some suggestions that having a smaller home page means
> that you don't need to scroll.  Unfortunately, this isn't true unless
> you have a very tall screen.  Whilst I can how see most of the headings
> in the bottom section of the screen, none of them are links so I still
> need to scroll down.  How about converting all these headings to links?
> 
> Overall, I'm disappointed at the regressions in the adaptability of
> the website - just because lots of other websites state "best viewed
> with MSIE at 800x600 resolution" doesn't make it right.  It's a pity
> that there appears to have been no effort to provide a trial version
> of the new website to get some feedback.
> 
> -- 
> Peter Jeremy

Again, thanks for the nice email that I could expand on instead
of restating the same things over and over and over again.  I just
want to touch on a few things of my own before I close this email.

A) I get it.  FreeBSD's mascot is a daemon and he's cute and red
and cuddly so you want to have your base color be #990000 for
all of your links.  How adorable.  No, seriously, I get it.

However, if you'll go here:

http://www.internettg.org/newsletter/mar99/color_challenged_applet.html

(Sorry that it's Java based (there are other offerings available
elsewhere if you want to look))

... you'll see what happens to red/green color blind users when
they see anything on your website in #990000.  It shows up as
a dull grey/green.  Not a big deal, you suggest, and yet you've
chosen to implement that as your primary link color and have it
change to a grey on hover.  I love that FreeBSD is devilly and
all that stuff, but a website (with usability in mind) should 
leave the 990000 to the mascot's skin tone and leave the links
at the default blue/purple combination that has become the default
color in browsers for a reason.  If you don't think that blue's fit
where your links all seem to be, then perhaps a tweaking of
those colors are in order.  It's a usability issue more than anything
and comes into play quite heavily when in comparison to the Heading
links (or lack of actual links) that share the same color and
yet aren't actually links like one would expect.  Colors (in 
the way that is being used here) are very important, and you
have violated just about every usability theory on the planet
in one fell swoop.  The designer will hate it because you'll be
"destroying her vision", but at the end of the day, why bother
creating a website at all if you're not going to make it as
useable for the users as you possibly can.

B) I LOVE the fact that we (the collective FreeBSD users) now
have a new website to call home.  A lot of the rants that I've
seen on the lists over the last little bit of time have been
(for the most part) just culture shock.  I, for one, have been
VERY happy to FINALLY have a new layout that's a little more
modernized.  Despite all the things that have been said here,
it's far easier at the point we're at now to flexibly move
into a place that's better for the users and the maintainers
than our old webpage ever was.  There are some deficiencies
and I just want to make sure they're being addressed properly
instead of as a series of rants by users who woke up to find
a new website they didn't know was coming.

All of that said, the webdesigner did a GREAT job in the 
Summer of Code for us and she deserves our deepest thanks
for doing a thankless job with some VERY difficult restrictions
(like no PHP?  wtf is up with that... ;) ) to having to 
deal with 200,000 users around the world who love an OS that
fills the niche of the most brutally unpleasing (to the eye)
design out of all other OS's that fit in mainstream markets.
FreeBSD fills this role EXCELLENTLY, since it's the only OS
that I can count on to work on any platform I can imagine and
do it without the fluff I've come to expect from the Linux
distro's and (gasp) Sun and Microsoft (although their installer
still sucks too, so meh!).  With that in mind, FreeBSD does
this very well, but designing a site to please users who
FreeBSD appeals to seems like a frightening task. 

Designer: Thanks for putting in the hours this summer
and doing great things for our website.

Finally: I've read some negative things from the list 
recently and heard things like "Well, submit a patch"
or something like that.  That's expected from driver
software or in ports, but I didn't know FreeBSD's website
was "patchable" in this regards too.  I am a webdeveloper
for a living and would be more than happy to share my
knowledge in any way that I can (FINALLY! Something I can
contribute to FreeBSD!) and a lot of the problems seem
like very easy fixes in the CSS (now aren't you glad 
you can change every page with one CSS tweak? ;)) But
I don't know enough of the politics behind how things
are done and I'm going to still have to go and read up
on the website patch submission process, and by then
it may be too late.  But, regardless of if it seems
like I've been very critical, I'd be MORE than happy
to help with patches,suggestions,a critical external
perspective,or even just as a filter for giving the
maintainers links to websites that discuss usability
and design and whatnot and help them strip out the crap.

I'm here for FreeBSD because FreeBSD has always been
here for me, so if you don't see any website patches
from 'erh@theimpossible.net' coming in, you can assume
that I've still not figured out how to submit a patch yet.

Anyway, thanks again Designer.  Thanks again Dev's.

The site change is going to be a lot of fun.

-E

> _______________________________________________
> freebsd-www@freebsd.org mailing list
> http://lists.freebsd.org/mailman/listinfo/freebsd-www
> To unsubscribe, send any mail to "freebsd-www-unsubscribe@freebsd.org"



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20051008034820.GA30150>