design question - what makes a page of options clear for the user?
Whitecrest
lkru...@geocities.com says...
> We (publicpen.com) are designing a new weblogging service, similar to
> Typepad. Our graphic designers (Misty Vredenburg and Peter Agelasto)
> are currently doing multiple design mock-ups of possible control
> panels. These would be the panels you arrive after you, the owner of
> the website, log into your secret, password-protected control-page. We
> are trying to figure out what visual arrangement makes the options the
> most clear.
> None of the links work yet, but if you can guess their meaning from
> their arrangement on the page, and the words that were choosen for
> them, then our designers must be doing something right.
> Comments welcome. One design can be seen here:
> http://www.publicdomainsoftware.org/index.php?pageId=389
> and here is another:
> http://www.monkeyclaus.org/index.php?pageId=662
Well I guess you could make them a little more plain if you really
tried. If this is the best your graphic designers can come up with, you
might want to look for different talent.
--
Whitecrest Entertainment
www.whitecrestent.com
Toby A Inkster
> http://www.publicdomainsoftware.org/index.php?pageId=389
> http://www.monkeyclaus.org/index.php?pageId=662
See this message (excluding the last paragraph).
http://groups.google.com/groups?selm=pan.2003.09.26.18.05.28.305678%40goddamn.co.uk
--
Toby A Inkster BSc (Hons) ARCS
Contact Me - http://www.goddamn.co.uk/tobyink/?id=132
Talc Ta Matt
that stuff onto the sub pages.
For instance, there's no reason to have all those management things on there.
Just say...
WEBLOG ENTRY
new entry, manage old entries
FILES
upload, manage
IMAGES
upload, manage
Then of course on the manage pages you'd have options for specific things.
lawrence
I appreciate all the responses on this thread. The design with a lot
of options is meant to test the theory that users infer meaning from
context. It's an idea that both Edward Tufte and Jakob Nielsen have
pushed. The other design was scrapped and the designer started over
again with something different:
PeterMcC
If the above is your theorising, my apologies for any offence in the
following; however, if it is the underlying principal that the designers are
claiming is driving their work, they're talking tendentious nonsense. There
is no "theory" being "tested" - that meaning is in part derived from context
has been established beyond any doubt for some considerable time.
I may be missing something but the designs that have been suggested look to
be largely artless and unattractive - sold as a good thing because they are
at the cutting edge of some supposedly radical concept about meaning and
context. I'd be inclined to get the designers to re-examine some of the less
radical notions - form and function looks like a good place to start.
> The other design was scrapped and the designer started over
> again with something different:
>
> http://www.publicdomainsoftware.org/index.php?pageId=398
--
PeterMcC
If you feel that any of the above is incorrect,
inappropriate or offensive in any way,
please ignore it and accept my apologies.
lawrence
> following; however, if it is the underlying principal that the designers are
> claiming is driving their work, they're talking tendentious nonsense. There
> is no "theory" being "tested" - that meaning is in part derived from context
> has been established beyond any doubt for some considerable time.
>
> I may be missing something but the designs that have been suggested look to
> be largely artless and unattractive - sold as a good thing because they are
> at the cutting edge of some supposedly radical concept about meaning and
> context. I'd be inclined to get the designers to re-examine some of the less
> radical notions - form and function looks like a good place to start.
Thanks much for your feedback. Your remark is similar to my own
concern, but I'm glad that I won't have to say that to them. I'll
forward your critique to them. My main concern is that neither design
has gone far down its road. For the user test to be much of a test, we
need two really different designs that clearly make different
assumptions about how users interact with a computer screen. I'd like
one design to be to super-heavy with options, and the other design to
have only, at most, 3 options on the screen at a time. And then we can
watch the users interact, and see if they prefer the design that hits
them with all the options at once, or the design that protects them
from that complexity. And so far, neither of these designs go very far
in the direction they are supposed to go.
lawrence
rather than here. Both pages have the comments function enabled.
However, I do also ask that you tone it down it a bit. Several of the
replies so far have been agressive and hostile. We're looking for
advice, not beligerence. Keep it useful and constructive or please,
please, please don't post.
We begin testing next week so then our opinions will meet reality. The
test will include these two designs, plus the design that TypePad is
using for their service (we've purchased an account on TypePad so our
users can login and post a real weblog entry to the web, using Ben and
Mena Trott's service).
Obviously we are hoping that, in the end, we will come up with a
design that is cleaner and clearer than TypePad. We will keep
listening to user feedback and modifying things accordingly until we
reach that level. By the way, has anyone here used TypePad, and if so,
what do you think of the service?
We also hope that people who've used PostNuke will look at our control
panel and consider our's better. Of course, this won't be very
difficult to acheive.
lawrence
> > their arrangement on the page, and the words that were choosen for
> > them, then our designers must be doing something right.
> > Comments welcome. One design can be seen here:
> > http://www.publicdomainsoftware.org/index.php?pageId=389
> > and here is another:
> > http://www.monkeyclaus.org/index.php?pageId=662
>
> Well I guess you could make them a little more plain if you really
> tried. If this is the best your graphic designers can come up with, you
> might want to look for different talent.
As to the plainness, I was personally inspired the extreme minimalism
of Phillip Greenspun's early ArsDigita system:
Nothing but a little black text on a white background. However, both
of the graphic designers on the team were horrified with the idea of
taking minimalism to such an extreme, and also they made the
reasonable point that the link structure in Greenspun's system is
damned confusing. However, it was another design dimension along which
the designers were supposed to split (and so far have done so to any
significant degree), and I hope they will. The blue/white design is
supposed to be quite minimalist, the other design, is supposed to, in
the end, have a lot of flash and javascript.
Of course, in the end, whatever the users like is what will go with.
The idea is simply to come up with two very different designs, so we
can see which way users lean. If we had the resources we would test 4
designs, and thus test each combination on the two dimensions we
mentioned so far (complexity verus simplicity of choices, and
minimalism versus flashiness).
The blue/white design is pretty much in final form. The other design
has morphed a great deal. I think we'll have another version out by
late Sunday night. It might be worth checking on it then.
Again, I appreciate the feedback.
lawrence
> lawrence wrote:
>
> > http://www.publicdomainsoftware.org/index.php?pageId=389
> > http://www.monkeyclaus.org/index.php?pageId=662
>
> See this message (excluding the last paragraph).
You wrote:
>>>>>>>>>>>>>>>>>>
This site suffers from the
i've-heard-that-the-font-tag-is-evil-and-div-and-
span-are-better-so-i'll-use-nothing-but-div-and-span syndrome.
>>>>>>>>>>>>>>>>>>
Actually, I've found that it is not necessary to use span tags. If you
give a div tag a class, and then in the style sheet you go like this
to the class:
display:inline;
Then the div will behave exactly like a span. So you can build
webpages with nothing by the A tag and creative use of the div tag. It
simplifies the toolset you have to work with, with the benefits that a
carpenter might understand, if the carpenter had to walk around all
day with 5 different hammers in his belt, but then one day discovered
a magic hammer that could take on any shape and do anything he wanted.
So then, instead of 5 hammers, he could use just one, and his life is
made that much more simple.
However, if its true that both designs are made of nothing but a,div,
and span tags, then I'm surprised, because one of our designers does
all his work in Dreamweaver, and I'm under the impression that
Dreamweaver uses the full range of HTML 4.0 tags in the markup that it
creates.
Toby A Inkster
>> This site suffers from the
>> i've-heard-that-the-font-tag-is-evil-and-div-and-
>> span-are-better-so-i'll-use-nothing-but-div-and-span syndrome.
>
> Actually, I've found that it is not necessary to use span tags.
Um. Me thinks you're missing the point.
> So you can build
> webpages with nothing by the A tag and creative use of the div tag. It
> simplifies the toolset you have to work with, with the benefits that a
> carpenter might understand, if the carpenter had to walk around all
> day with 5 different hammers in his belt, but then one day discovered
> a magic hammer that could take on any shape and do anything he wanted.
> So then, instead of 5 hammers, he could use just one, and his life is
> made that much more simple.
Yes, but <div> is *not* a magic hammer. It's an ordinary hammer.
Consider a Swiss watch maker, crafting delicate timepieces using a variety
of intricate tools.
Now take away his tools and give him a single hammer.
Is he going to build watches with the hammer? No, he's going to cave your
skull in to get his tools back!
> However, if its true that both designs are made of nothing but a,div,
> and span tags, then I'm surprised
**If**??? You mean you haven't even **looked** at them?
Isofarro
> Then the div will behave exactly like a span. So you can build
> webpages with nothing by the A tag and creative use of the div tag. It
> simplifies the toolset you have to work with,
At the considerable cost of having a non existant structure. There are good
reasons to identify headers, paragraphs, lists and sections using proper
HTML elements. Using just divs and anchors is pointless to the extreme.
> and I'm under the impression that
> Dreamweaver uses the full range of HTML 4.0 tags in the markup that it
> creates.
Who authors pages, the tool or the user? If the user isn't using the range
of HTML elements, the tool isn't at fault.
--
Iso.
FAQs: http://html-faq.com http://alt-html.org http://allmyfaqs.com/
Recommended Hosting: http://www.affordablehost.com/
Web Design Tutorial: http://www.sitepoint.com/article/1010
lawrence
> > and span tags, then I'm surprised
>
> **If**??? You mean you haven't even **looked** at them?
Sure, we've spent hours talking about them, in house, but I don't
think I've looked at the source code on them. They are changing fast
and it is pointless to try to keep up with day-to-day changes.
Besides, if you know a page was made with Dreamweaver, then what is
the point of looking at the source code? Isn't all Dreamweaver code
pretty much the same?
I thought your remarks about the Swiss Watch maker needing a wide
array of tools was very interesting, so I sent it to both designers.
Of course it'll be pure-Greek to the fellow who designs only in
Dreamweaver - he doesn't know how to hand code. The other woman might
find it interesting.
I really do appreciate your feedback, however, mostly, on this forum
I'm looking for feedback about the clarity of the options, not the way
the HTML tags are written.
lawrence
> lawrence wrote:
>
> > Then the div will behave exactly like a span. So you can build
> > webpages with nothing by the A tag and creative use of the div tag. It
> > simplifies the toolset you have to work with,
>
> At the considerable cost of having a non existant structure. There are good
> reasons to identify headers, paragraphs, lists and sections using proper
> HTML elements. Using just divs and anchors is pointless to the extreme.
>
> > and I'm under the impression that
> > Dreamweaver uses the full range of HTML 4.0 tags in the markup that it
> > creates.
>
> Who authors pages, the tool or the user? If the user isn't using the range
> of HTML elements, the tool isn't at fault.
I appreciate the feedback, but surely there are a lot of designers out
there who use Dreamweaver or Frontpage exclusively?
Anyway, we are under a very tight deadline, so there may not be time
for the person in question to learn a new set of skills. Though, of
course, every project allows a person to learn new skills.
But somehow this conversation has gone off track. We're now talking
about HTML markup. Does anyone have an insight to offer regarding the
clarity, or lack thereof, of the designs?
lawrence
>
> > Then the div will behave exactly like a span. So you can build
> > webpages with nothing by the A tag and creative use of the div tag. It
> > simplifies the toolset you have to work with,
>
> At the considerable cost of having a non existant structure. There are good
> reasons to identify headers, paragraphs, lists and sections using proper
> HTML elements. Using just divs and anchors is pointless to the extreme.
She wrote me back: "Yes... and what is the good reason? the structure
is obviously not nonexistant - the japge renders."
I think that about sums up my view. XML has semantics, HTML doesn't.
When dealing with HTML, as long as the page renders across all
platforms (we test IE and Netscape on PCs and Macs, and Netscape on
Linux) then that seems to be good enough, yes?
Mark Parnell
stated:
> Isofarro <spam...@spamdetector.co.uk> wrote in message news:<7rg8mb...@sidious.isolani.co.uk>...
>>
>> Who authors pages, the tool or the user? If the user isn't using the range
>> of HTML elements, the tool isn't at fault.
>
> I appreciate the feedback, but surely there are a lot of designers out
> there who use Dreamweaver or Frontpage exclusively?
No, that's *deezyners*. ;-) But you are missing the point. The tool is
irrelevant (to a large extent, anyway). It is the ability of the user that
matters. Someone who knows what they are doing can make a better site in
Notepad than someone who doesn't know what they are doing who happens to be
using DW, and vice-versa.
>
> But somehow this conversation has gone off track. We're now talking
> about HTML markup.
What do you expect, in alt.html? :-)
^^^^
--
Mark Parnell
http://www.clarkecomputers.com.au
Mark Parnell
stated:
>
Are we talking about the same HTML here? HTML is a _Markup Language_. It
describes the structure of the document. If text is a heading, why not
mark it up as a heading? If it is a list, why not mark it up as a list?
And so on, and so forth. If you aren't going to do that, then you may as
well just write it in plain text.
> When dealing with HTML, as long as the page renders across all
> platforms (we test IE and Netscape on PCs and Macs, and Netscape on
> Linux) then that seems to be good enough, yes?
Taking headings as a specific example, there are various browsers which
allow the user to get a kind of "document outline" from a page, using the
headings on the page.
Toby A Inkster
> XML has semantics, HTML doesn't.
Quite the opposite actually. How could anyone interpret the semantics of
the following well-formed XML snippet?
<flibble>
<foo f="John" x="Steve" />
<foo d="Chas" f="Dave" />
<bar zibble="zonk">
Potato<baz />Celery<baz />&russian;
</bar>
</flibble>
There is simply no way to make any sense of it, unless you know what the
elements flibble, foo, bar and baz mean, and what their attributes mean.
You might be able to infer some meaning if you had access to the DTD, but
that is certainly not guaranteed.
On the other hand...
<body>
<img src="John" alt="Steve" />
<img id="Chas" src="Dave" />
<p id="zonk">
Potato<br />Celery<br />Π
</p>
</body>
... has a well defined meaning: any browser worth ots salt could tell you
that "Dave" is the URL from which it may load an image; <br /> should
cause a line break; if "John" can't be displayed, that it should display
the text "Steve" instead.
HTML has semantics, XML sometimes does.
Toby A Inkster
> I really do appreciate your feedback, however, mostly, on this forum
> I'm looking for feedback about the clarity of the options, not the way
> the HTML tags are written.
But the way the HTML tags are written *effects* the clarity.
If you use <div style="font-size:200%;"> to mark up a heading, then it is
not clear where the headings are when viewed in any non-CSS browser.
Steve Pugh
>I think that about sums up my view. XML has semantics, HTML doesn't.
LOL
You're 100% wrong.
HTML is nothing but semantics.
>When dealing with HTML, as long as the page renders across all
>platforms (we test IE and Netscape on PCs and Macs, and Netscape on
>Linux) then that seems to be good enough, yes?
What about Google? Lynx? Jaws? Home Page Reader?
Steve
--
"My theories appal you, my heresies outrage you,
I never answer letters and you don't like my tie." - The Doctor
Steve Pugh <st...@pugh.net> <http://steve.pugh.net/>
Isofarro
> Isofarro <spam...@spamdetector.co.uk> wrote in message
> news:<7rg8mb...@sidious.isolani.co.uk>...
>> lawrence wrote:
>>
>> > Then the div will behave exactly like a span. So you can build
>> > webpages with nothing by the A tag and creative use of the div tag. It
>> > simplifies the toolset you have to work with,
>>
>> At the considerable cost of having a non existant structure. There are
>> good reasons to identify headers, paragraphs, lists and sections using
>> proper HTML elements. Using just divs and anchors is pointless to the
>> extreme.
>
> She wrote me back: "Yes... and what is the good reason? the structure
> is obviously not nonexistant - the japge renders."
Rendering in one browser with a certain configuration is only one tiny step
on the world wide web.
> I think that about sums up my view. XML has semantics, HTML doesn't.
Semantics isn't structure.
> When dealing with HTML, as long as the page renders across all
> platforms (we test IE and Netscape on PCs and Macs, and Netscape on
> Linux) then that seems to be good enough, yes?
Hardly. What do you consider "Netscape" as being? I see there's no speech
based browser, no text-browser, no search engine spider, no khtml based
browser, probably no thought of Mozilla, no Opera, no Dillo, no iCab. Platforms - no
PocketPC, no WinCE, no Lineo devices, no Risc based systems, no set-top
boxes, no mobile-phone technology.
Isofarro
> Isofarro <spam...@spamdetector.co.uk> wrote in message
> news:<7rg8mb...@sidious.isolani.co.uk>...
>> lawrence wrote:
>>
>> > Then the div will behave exactly like a span. So you can build
>> > webpages with nothing by the A tag and creative use of the div tag. It
>> > simplifies the toolset you have to work with,
>>
>> At the considerable cost of having a non existant structure. There are
>> good reasons to identify headers, paragraphs, lists and sections using
>> proper HTML elements. Using just divs and anchors is pointless to the
>> extreme.
>>
>> > and I'm under the impression that
>> > Dreamweaver uses the full range of HTML 4.0 tags in the markup that it
>> > creates.
>>
>> Who authors pages, the tool or the user? If the user isn't using the
>> range of HTML elements, the tool isn't at fault.
>
> I appreciate the feedback, but surely there are a lot of designers out
> there who use Dreamweaver or Frontpage exclusively?
Pity. Using Dreamweaver or Frontpage in conjunction to a brain would
certainly improve matters. Its like a toggle switch. Turn on Dreamweaver,
turn off brain.
> Anyway, we are under a very tight deadline, so there may not be time
> for the person in question to learn a new set of skills.
So that's your excuse for using:
<div class="bigRedSpottyHeader">News</div>
instead of
<h1>News</h1>
Interesting argument.
lawrence
I appreciate all the reponses and I'll forward them. As to supporting
PDA's and set-top boxes and such, we're working on some code that will
output a XHTML 1.1 valid print-friendly version of each page.
Does anyone have anything, positive or negative, to say about the
clarity of either design, or to point to admin or options pages or
control panels on the web that you think are good examples of clarity?
Has anyone here used TypePad and if so would you like to share your
reactions to it? Did you think it was clear?
lawrence
That's reasonable, of course. Could you recommend a good design forum?
I searched groups on google and found none with the word "design" in
it. Do you yourself have a favorite design site? We've all (in-house)
been studying Edward Tufte a lot, and www.edwardtufte.com is good, but
it is a general design site, and not focused on web-design issues.
Christina Wodtke is also very good, but again, not focused on visual
design.
lawrence
> > there who use Dreamweaver or Frontpage exclusively?
>
> Pity. Using Dreamweaver or Frontpage in conjunction to a brain would
> certainly improve matters. Its like a toggle switch. Turn on Dreamweaver,
> turn off brain.
But surely you can't mean that? Look around the web and you see a lot
of excellent work being done in Dreamweaver and Frontpage. I have the
impression, from talking with people at the local tech society, that
the majority of designers use one of these tools.
Writing a sentence like "Turn on Dreamweaver, turn off brain" reduces
your comment to the level of a troll looking to start a flamewar. I
assume you're more intelligent than that.
lawrence
> > platforms (we test IE and Netscape on PCs and Macs, and Netscape on
> > Linux) then that seems to be good enough, yes?
>
> Hardly. What do you consider "Netscape" as being? I see there's no speech
> based browser, no text-browser, no search engine spider, no khtml based
> browser, probably no thought of Mozilla, no Opera, no Dillo, no iCab. Platforms - no
> PocketPC, no WinCE, no Lineo devices, no Risc based systems, no set-top
> boxes, no mobile-phone technology.
I don't know the answer to your questions, but doesn't Dreamweaver
take care of a certain amount of this? I'm not going to tell my
co-workers that they have to start hand-coding so we can support
WinCE. It seems like these are the sort of issues that Dreamweaver and
Frontpage are supposed to protect a designer from.
lawrence
> lawrence wrote:
>
> > Isofarro <spam...@spamdetector.co.uk> wrote in message
> > news:<7rg8mb...@sidious.isolani.co.uk>...
> >> lawrence wrote:
> >>
> >> > Then the div will behave exactly like a span. So you can build
> >> > webpages with nothing by the A tag and creative use of the div tag. It
> >> > simplifies the toolset you have to work with,
> >>
> >> At the considerable cost of having a non existant structure. There are
> >> good reasons to identify headers, paragraphs, lists and sections using
> >> proper HTML elements. Using just divs and anchors is pointless to the
> >> extreme.
> >
> > She wrote me back: "Yes... and what is the good reason? the structure
> > is obviously not nonexistant - the japge renders."
>
> Rendering in one browser with a certain configuration is only one tiny step
> on the world wide web.
>
> > I think that about sums up my view. XML has semantics, HTML doesn't.
>
> Semantics isn't structure.
One thing I get from this whole discussion is that it is important to
use heading tags for the headings, because Google will look at those.
But how can Google evaluate <li> or <ul>? Even <blockquote> is
meaningless if it doesn't come with a linked citation. As near as I
know, you can't link a citation, or an url, to a blockquote in HTML,
though you could do so in XML. I suppose you could go:
<blockquote name="we_hold_certain_truths_to_be_self_evi"
value="http://www.declarationOfIndependence.org">
And then write some software to parse that. That might be interesting
to do.
But basically, what I'm saying is, other than heading tags, which the
search engines can use, I don't see that the other tags give enough
information that one can parse them in some meaningful way.
Drifting off the subject a bit, as we have been for several posts, my
basic resistance to using something like <li> in the navigation bar is
that it seems to make an assumption of a homogenous list of links,
when in fact we've designed our software to make it easy for people to
put anything they want into the nav bars. We feel this is one of our
basic strengths compared to MoveableType.
Mark Parnell
stated:
> Mark Parnell <webm...@clarkecomputers.com.au> wrote in message news:<1muad62djjeyr.1...@40tude.net>...
>> Sometime around 13 Oct 2003 20:58:02 -0700, lawrence is reported to have
>> stated:
>>
>>> But somehow this conversation has gone off track. We're now talking
>>> about HTML markup.
>>
>> What do you expect, in alt.html? :-)
>
> That's reasonable, of course. Could you recommend a good design forum?
news:comp.infosystems.www.authoring.site-design (though you will find many
of the regulars from alt.html there as well) or perhaps
news:alt.comp.programming.interface-design (don't subscribe there myself so
can't comment on the quality :-) )
Toby A Inkster
> As near as I
> know, you can't link a citation, or an url, to a blockquote in HTML,
> though you could do so in XML.
In HTML:
<blockquote cite="http://www.ingsoc.org/bb/">
<p>Big Brother is watching you.</p>
</blockquote>
Isofarro
> Isofarro <spam...@spamdetector.co.uk> wrote in message
>> > When dealing with HTML, as long as the page renders across all
>> > platforms (we test IE and Netscape on PCs and Macs, and Netscape on
>> > Linux) then that seems to be good enough, yes?
>>
>> Hardly. What do you consider "Netscape" as being? I see there's no speech
>> based browser, no text-browser, no search engine spider, no khtml based
>> browser, probably no thought of Mozilla, no Opera, no Dillo, no iCab.
>> Platforms - no PocketPC, no WinCE, no Lineo devices, no Risc based
>> systems, no set-top boxes, no mobile-phone technology.
>
> I don't know the answer to your questions,
Open up this browser you consider as being "Netscape". Click on "Help" in
the browser menu, then click "About..." (or something similar). Therein it
will mention the actual browser version you are using. Now copy and paste
that browser version into a reply to this post. That will sort question
number 1.
> but doesn't Dreamweaver
> take care of a certain amount of this?
Not according to the websites out there that are supposedly created in
Dreamweaver.
> I'm not going to tell my
> co-workers that they have to start hand-coding so we can support
> WinCE.
To support WinCE based browsers requires you to add nothing into your pages.
To stop your website being usable on a WinCE device just use a tool and
don't bother learning about the medium you are working in (It worked for
everyone else).
> It seems like these are the sort of issues that Dreamweaver and
> Frontpage are supposed to protect a designer from.
Hardly. A good webdesigner uses his knowledge of HTML and his knowledge of
the Web to produce websites that work in a large variety of circumstances.
A software tool cannot help you here. Dreamweaver is like a hammer. It can
put a nail into a wall, but its no good at telling you you've just hammered
it into the wrong place.
Isofarro
> One thing I get from this whole discussion is that it is important to
> use heading tags for the headings, because Google will look at those.
> But how can Google evaluate <li> or <ul>?
Items in lists are related. Otherwise they wouldn't be listed together. Now
have a look at http://labs.google.com/ and look at Google Sets.
> Even <blockquote> is
> meaningless if it doesn't come with a linked citation.
http://www.diveintomark.com/ uses cite and quote as the basis of a citation
search.
> As near as I
> know, you can't link a citation,
Fragments of markup can be (and are) extracted and displayed as part of a
list of results. (Properly structured HTML fragments).
> But basically, what I'm saying is, other than heading tags, which the
> search engines can use, I don't see that the other tags give enough
> information that one can parse them in some meaningful way.
So your knowledge is limited, damn the rest of us? Nice.
> Drifting off the subject a bit, as we have been for several posts, my
> basic resistance to using something like <li> in the navigation bar is
> that it seems to make an assumption of a homogenous list of links,
You claim it is not a list? Then what is it?
> when in fact we've designed our software to make it easy for people to
> put anything they want into the nav bars. We feel this is one of our
> basic strengths compared to MoveableType.
No. MoveableType is written by a team (Ben and Mena Trott, plus Anil Dash)
that truely grok the web. They grok the fundamental concepts of structure,
community and interoperability - the essence of the web. Its no big
surprise that the most influential people on the web use MoveableType as
their blog software (apart from the guys who roll their own blogs).
Isofarro
> Isofarro <spam...@spamdetector.co.uk> wrote in message
>> > I appreciate the feedback, but surely there are a lot of designers out
>> > there who use Dreamweaver or Frontpage exclusively?
>>
>> Pity. Using Dreamweaver or Frontpage in conjunction to a brain would
>> certainly improve matters. Its like a toggle switch. Turn on Dreamweaver,
>> turn off brain.
>
> But surely you can't mean that?
What portion of these designers you talk about have a website that validates
and contains properly structured HTML? Go on, have a look. Now, when you
find out the vast majority of them have websites that don't validate and
have no sensible structure - then you need to ask yourself why that is that
these pages are not valid and not well structured. Is it the tool, or the
web designer that is causing the problem?
"Well that's what Dreamweaver / Frontpage generates" is a symptomatic excuse
that the tool is switched on, and the brain is switched off. Web designers
create websites, they use tools to do so - that can only really be done
well with the brain engaged. So markup that is broken coming out the tool
chosen for the job is fixed in one of two ways:
* Fix it by hand (requires a brain and knowledge of HTML - no problem for
someone developing websites)
* Use a tool that isn't broken.
> Look around the web and you see a lot
> of excellent work being done in Dreamweaver and Frontpage.
I see 0.7% of pages on the web that actually validate. Looking for websites
that are properly structured in this is needle and haystack time. How many
of these 0.7% of sites are done by webdesigners exclusively relying on
Frontpage or Dreamweaver? How many websites in the 99.3% of websites are
created using Frontpage and Dreamweaver? You will find the vast majority of
professional web designers in the 99.3% "haven't got the basics right"
category.
> I have the
> impression, from talking with people at the local tech society, that
> the majority of designers use one of these tools.
Numbers does not equate to a compelling argument. Sheep mentality is a known
disorder on the web. Rather evaluate the tools on merit and the ability to
aid you in getting the job done properly.
Toby A Inkster
> Its no big
> surprise that the most influential people on the web use MoveableType as
> their blog software (apart from the guys who roll their own blogs).
Soon http://www.politechbot.com/ is moving to MT too.
Mark Parnell
have stated:
> A software tool cannot help you here. Dreamweaver is like a hammer. It can
> put a nail into a wall, but its no good at telling you you've just hammered
> it into the wrong place.
LOL - I like that analogy. :-)
lawrence
about how a web page should be designed. Suppose something is the
banner for the page. It is the most important thing on the page. Then
it should be in an <h1> header. Something is a subheader. Then it
should be in an <h2> or <h3>. There is a list of options - no need for
it not to be in <ul>.
I get that now. You are all designers, and this is the way you think.
I started off thinking you were idiots, but now I realize that the job
of a designer is to make decisions just like this - the intelligent
discernment of what information has weight. In web design, one has to
make those discernments, then back them with the right kind of tag.
You are all in the business of final-output decisions.
Since the designers working on our control panels are working on final
designs, they should be making those kind of final-output decisions. I
only now realize they extent are still coding in the old style.
My perspective is different from the perspective of a designer because
mostly I program. We work to have a system that a graphic designer can
customize. I take out hard codings where I can, but much remain. It is
a shame, but it is a convenience, and for a small shop, convenience is
a powerful incentive. Not just for us, but for every small design shop
in the world that might use our product. Our goal has always been to
come up with a set of tools that would work well inside of
Dreamweaver, and anyone using Dreamweaver is going to put a premium on
convenience, or otherwise they would not be using Dreamweaver.
I use div tags for everything because it offers neutrality. It would
be wrong if I made final-output decisions and hardcoded that into the
software. What if I thought the web page headline was more important
than an entry headline but then some graphic designer wanted to do a
design that went flatly against my assumption. I fear someday seeing a
style sheet like this:
h1.webpageHeadline{
visibility:hidden;
line-height:0px;
}
h4.entrySubheading{
font-size: 120px;
font-weight: bold;
}
Here is a site that is being designed so that the entries are
important and the page title has been made invisible. In fact, my own
personal website is designed like this. But would Google understand
this? Would Google understand that on this website, the empty <h1> has
no importance, and the <h4> tag should be treated as if it was h1?
Rather than guess wrong, when I have to hardcode some code, I prefer
to put everything into div tags. They offer a neutrality of judgement.
lawrence
> > basic resistance to using something like <li> in the navigation bar is
> > that it seems to make an assumption of a homogenous list of links,
>
> You claim it is not a list? Then what is it?
It is whatever the user puts there. Maybe an image. Maybe some text
about themselves. Maybe a political quote they like. Maybe some PHP
code that does stuff I can't imagine. The thing is, I can't know ahead
of time what they are going to put there.
lawrence
> that are properly structured in this is needle and haystack time. How many
> of these 0.7% of sites are done by webdesigners exclusively relying on
> Frontpage or Dreamweaver? How many websites in the 99.3% of websites are
> created using Frontpage and Dreamweaver? You will find the vast majority of
> professional web designers in the 99.3% "haven't got the basics right"
> category.
This discussion is sadly degenerating into a flame war. To my mind
your statistics betray your argument. The web is a vibrant place. If
it can be so vibrant despite the fact 99.3% of the pages are done
wrong, then the "standards" defining validation are meaningless.
However, I don't wish to engage in what might turn out to be a long
and fruitless discussion, so I will leave the final word on the matter
to you.
Toby A Inkster
> This discussion is sadly degenerating into a flame war. To my mind
> your statistics betray your argument. The web is a vibrant place. If
> it can be so vibrant despite the fact 99.3% of the pages are done
> wrong, then the "standards" defining validation are meaningless.
No -- things happen to work because pages are close enough to the
standards for browsers to guess what the authors meant.
If it weren't for the standards, then this wouldn't be the case.
--
Toby A Inkster BSc (Hons) ARCS
Contact Me - http://www.goddamn.co.uk/tobyink/?id=132
playing://noartist/unknown_disc/15_track_15.ogg
lawrence
> Isofarro wrote:
>
> > Its no big
> > surprise that the most influential people on the web use MoveableType as
> > their blog software (apart from the guys who roll their own blogs).
>
> Soon http://www.politechbot.com/ is moving to MT too.
I'd like to again thank everyone for all the intelligent feedback.
I've forwarded a few of the better posts to everyone on my team.
Isofarro
> Isofarro <spam...@spamdetector.co.uk> wrote in message
>> > Drifting off the subject a bit, as we have been for several posts, my
>> > basic resistance to using something like <li> in the navigation bar is
>> > that it seems to make an assumption of a homogenous list of links,
>>
>> You claim it is not a list? Then what is it?
>
> It is whatever the user puts there. Maybe an image.
Collections of images linking to pages - still a list of links
> Maybe some text
> about themselves.
Collections of text linking to pages - still a list of links.
> Maybe a political quote they like.
You are using a political quote as a form of navigation? Nuts.
lawrence
> lawrence wrote:
>
> > Isofarro <spam...@spamdetector.co.uk> wrote in message
> >> > Drifting off the subject a bit, as we have been for several posts, my
> >> > basic resistance to using something like <li> in the navigation bar is
> >> > that it seems to make an assumption of a homogenous list of links,
> >>
> >> You claim it is not a list? Then what is it?
> >
> > It is whatever the user puts there. Maybe an image.
>
> Collections of images linking to pages - still a list of links
>
> > Maybe some text
> > about themselves.
>
> Collections of text linking to pages - still a list of links.
>
> > Maybe a political quote they like.
>
> You are using a political quote as a form of navigation? Nuts.
I hope you realize how agressive your tone is. I really have to wonder
if you'd talk like this if we were face to face having a chat at some
restaurant. I'm always a bit saddened by the agressive tone so many
people take when they talk online.
The side bar is there for people to put stuff into it. What they put
into it I cannot know a head of time. How they wish to order their
websites I can not know ahead of time. What they want to do with their
web sites I can not know ahead of time. I have to program for maximum
flexibility.
A link, followed by a political quote, followed by a link, followed by
an image of themselves, followed by a PHP script that gives the date,
followed by another link, seems to me like a heterogenous collection.
I'm not sure that such a collection should be each preceded by <li> .
lawrence
> lawrence wrote:
>
> > This discussion is sadly degenerating into a flame war. To my mind
> > your statistics betray your argument. The web is a vibrant place. If
> > it can be so vibrant despite the fact 99.3% of the pages are done
> > wrong, then the "standards" defining validation are meaningless.
>
> No -- things happen to work because pages are close enough to the
> standards for browsers to guess what the authors meant.
>
> If it weren't for the standards, then this wouldn't be the case.
I took nearly all of your posts and sent them to the two graphic
designers I work with. There reaction was basically "What is all
this?" They couldn't make sense of it. They had no idea what a
structured document was, nor why it might be important. They are
coming from a graphic design background, with a strong emphasis on the
word "graphic". They know how to make pages that look pretty. It was
interesting to watch their reaction. They felt like they were reading
a foreign language. They were wary. Using tags to offer logical
structure to a document sounded to them like a bit of programming, and
they both reminded me that they weren't programmers.
We've planned a meeting for Monday during which we're going to go over
this thread and I'll explain to them what a structured document is.
I'm not doing it to convert them, but I want them at least have a
sense about this point of view, which you do a fairly good job of
detailing.
Still, when they ask me why a structured document is important, I
admit I stutter. To my mind, the only reason that "structure" is
important is so that the document can be processed by software. But
HTML doesn't give me enough information to do anything useful. I can
not think of any interesting code I might write to process something
like <ul>. If I don't know what the content is, how might I process
it? So I'm not inclined to push it. Processing a structured document,
to my mind, is something one does with XML, because there the info is
so much richer. We've built Flash sites where the Flash reads info out
of an XML file to know what to show on screen. Also, I've a friend
who's in charge of the effort at Lexis/Nexus to convert their
datastore to an XML format. We've had some fun talks. Of course,
Lexis/Nexus doesn't dream of using HTML tags as their database,
because HTML tags aren't descriptive enough.
Still, I've come to see the value of using header tags and Blockquote
and Cite. Google looks at header tags, and I can write cool code to do
something with blockquote and cite.
Thanks for all the feedback.
lawrence
> lawrence wrote:
>
> > XML has semantics, HTML doesn't.
>
> Quite the opposite actually. How could anyone interpret the semantics of
> the following well-formed XML snippet?
>
> <flibble>
> <foo f="John" x="Steve" />
> <foo d="Chas" f="Dave" />
> <bar zibble="zonk">
> Potato<baz />Celery<baz />&russian;
> </bar>
> </flibble>
>
> There is simply no way to make any sense of it, unless you know what the
> elements flibble, foo, bar and baz mean, and what their attributes mean.
If it was my XML, then I'd know what it would mean, and I could write
the software to process it. If its not my XML, but it is a well known
XML standard, like some of Google's web services, or weblogs.com
services, then I can write code to process it. But I can't think of
any interesting code to write for HTML. The only software I can think
of writing for it is rendering software, that is, a webbrowser, which
I won't do because there are already plenty of web browsers out there.
But if I were writing such software, it seems like making the thing
look right would be my principal concern.
Isofarro
> Isofarro <spam...@spamdetector.co.uk> wrote in message
> news:<p7srmb...@sidious.isolani.co.uk>...
>> lawrence wrote:
>>
>> > Isofarro <spam...@spamdetector.co.uk> wrote in message
>> >> > Drifting off the subject a bit, as we have been for several posts,
>> >> > my basic resistance to using something like <li> in the navigation
>> >> > bar is that it seems to make an assumption of a homogenous list of
>> >> > links,
>> >>
>> >> You claim it is not a list? Then what is it?
>> >
>> > It is whatever the user puts there. Maybe an image.
>>
>> Collections of images linking to pages - still a list of links
>>
>> > Maybe some text
>> > about themselves.
>>
>> Collections of text linking to pages - still a list of links.
>>
>> > Maybe a political quote they like.
>>
>> You are using a political quote as a form of navigation? Nuts.
>
> The side bar is there for people to put stuff into it.
Right, now its a side bar, and not a navigational element. Serious
difference.
> What they put
> into it I cannot know a head of time.
That's what requirements gathering and analysis are for. You don't create a
website until you have requirements.
> I have to program for maximum flexibility.
Websites are not programmed. They are authored.
> A link, followed by a political quote, followed by a link, followed by
> an image of themselves, followed by a PHP script that gives the date,
> followed by another link, seems to me like a heterogenous collection.
Seems like a random clutter of ill-thoughout components. What user benefits
is this haphazard scattering supposed to achieve? Some respectable
references would be nice at this point.
Isofarro
> Toby A Inkster <UseTheAddr...@deadspam.com> wrote in message
> news:<pan.2003.10.18....@goddamn.co.uk>...
>> lawrence wrote:
>>
>> > This discussion is sadly degenerating into a flame war. To my mind
>> > your statistics betray your argument. The web is a vibrant place. If
>> > it can be so vibrant despite the fact 99.3% of the pages are done
>> > wrong, then the "standards" defining validation are meaningless.
>>
>> No -- things happen to work because pages are close enough to the
>> standards for browsers to guess what the authors meant.
>>
>> If it weren't for the standards, then this wouldn't be the case.
>
> I took nearly all of your posts and sent them to the two graphic
> designers I work with. There reaction was basically "What is all
> this?"
So they don't really know HTML. Sounds quite familiar.
> They couldn't make sense of it. They had no idea what a
> structured document was, nor why it might be important.
Quite a common failing with web designers. Nothing to be completely
embarrassed about.
> They are
> coming from a graphic design background, with a strong emphasis on the
> word "graphic". They know how to make pages that look pretty.
Glamour websites died along with boo.com.
> It was
> interesting to watch their reaction. They felt like they were reading
> a foreign language. They were wary.
That tends to happen when they are faced with the distinct possibility that
what they believe is wrong. Now they need to get over that bump in the
road, and work with the strengths of the medium (like every true artist).
> Using tags to offer logical
> structure to a document sounded to them like a bit of programming,
No, its called markup. Newspapers, books and magazines have been doing it
for centuries.
> We've planned a meeting for Monday during which we're going to go over
> this thread and I'll explain to them what a structured document is.
You'll probably have to explain the following concepts first: Hypertext,
then Markup, only then attempt structure. Pointing out that the web isn't
solely a visual experience will also help at this point. (and yes, blind
people do use the web).
> Still, when they ask me why a structured document is important, I
> admit I stutter.
Why is structure important in any content medium? Why are books divided into
logical structures like chapters and parts. Why are page numbers found in a
consistent location within a particular book? Why are magazines broken into
articles and tabulated into a tables of content? Why do newspapers have a
strongly hierarchical series of headlines. Why are the top stories on the
outer pages of a newspaper. These are all logical structures - so such a
concept should not be alien to anyone with basic literacy skills.
> To my mind, the only reason that "structure" is
> important is so that the document can be processed by software.
One tiny piece. Structure brings consistency and familiarity. When you look
at the front page of a newspaper - how do you tell which story is regarded
by that edition as the most important story? You tell by reading the main
headline (a logical structure). Now you can use that logical structure and
pick up almost any decent newspaper and figure out which is their most
important story (including a foreign language newspaper).
> I can
> not think of any interesting code I might write to process something
> like <ul>.
Consider reading through the thread properly before regurgitating. It is
common courtesy.
Toby A Inkster
> Toby A Inkster <UseTheAddr...@deadspam.com> wrote in message news:<pan.2003.10.14....@goddamn.co.uk>...
>> lawrence wrote:
>>
>>> XML has semantics, HTML doesn't.
>>
>> Quite the opposite actually. How could anyone interpret the semantics of
>> the following well-formed XML snippet?
[snip]
>> There is simply no way to make any sense of it, unless you know what the
>> elements flibble, foo, bar and baz mean, and what their attributes mean.
>
> If it was my XML, then I'd know what it would mean, and I could write
> the software to process it.
[snip]
> But I can't think of any interesting code to write for HTML. The only
> software I can think of writing for it is rendering software, that is, a
> webbrowser
Which only goes to show that you have a very limited imagination.
How about a spreadsheet that uses HTML tables?
Or a document display system that can generate automatic summaries of
documents, based on their <hX> headings?
A web crawler that searches documents for <dt>/<dd> pairs to populate a
dictionary?
A C++ compiler with a built in browser, so you can surf the web and then
press a button to compile the C++ code between <code> and </code>?
Presentation software (like Powerpoint)? (This has already been done by
Opera).
A word processor that can compile a bibliography at the end of the
document automatically, using links and cite attributes?
A crawler that uses the rel attribute of HTML links to build a site map
for an entire web site?
There are tonnes of things that can be done with HTML, but only if you use
the correct elements for everything.
This is because HTML has semantics. XML has no inherant semantics.
lawrence
>
> How about a spreadsheet that uses HTML tables?
>
> Or a document display system that can generate automatic summaries of
> documents, based on their <hX> headings?
>
> A web crawler that searches documents for <dt>/<dd> pairs to populate a
> dictionary?
>
> A C++ compiler with a built in browser, so you can surf the web and then
> press a button to compile the C++ code between <code> and </code>?
>
> Presentation software (like Powerpoint)? (This has already been done by
> Opera).
>
> A word processor that can compile a bibliography at the end of the
> document automatically, using links and cite attributes?
This is quite an interesting list of ideas for interacting with HTML.
I thank you for the list.
lawrence
question elsewhere. Can anyone point to a design site where people
discuss non-technical issues of design and visual presentation? I
already read Edward Tufte's site and Christina Wodtke's site, but
something specific to the web would be helpful.
Whitecrest <white...@zipzap.com> wrote in message news:<MPG.19e0bd45b...@news.charter.net>...
> In article <da7e68e8.03092...@posting.google.com>,
> lkru...@geocities.com says...
> > We (publicpen.com) are designing a new weblogging service, similar to
> > Typepad. Our graphic designers (Misty Vredenburg and Peter Agelasto)
> > are currently doing multiple design mock-ups of possible control
> > panels. These would be the panels you arrive after you, the owner of
> > the website, log into your secret, password-protected control-page. We
> > are trying to figure out what visual arrangement makes the options the
> > most clear.
> > None of the links work yet, but if you can guess their meaning from
> > their arrangement on the page, and the words that were choosen for
> > them, then our designers must be doing something right.
> > Comments welcome. One design can be seen here:
> > http://www.publicdomainsoftware.org/index.php?pageId=389
> > and here is another:
> > http://www.monkeyclaus.org/index.php?pageId=662
>
> Well I guess you could make them a little more plain if you really
> tried. If this is the best your graphic designers can come up with, you
> might want to look for different talent.
William Tasso
> This topic is apparently off-topic on alt.html, so I'll take my
> question elsewhere. Can anyone point to a design site where people
> discuss non-technical issues of design and visual presentation? I
> already read Edward Tufte's site and Christina Wodtke's site, but
> something specific to the web would be helpful.
The o/p is not available from my news service but from the subject header
and your notes,
news:comp.infosystems.www.authoring.site-design
seems like a sensible choice.
--
William Tasso - http://WilliamTasso.com
lawrence
to have a lot of empty tags, like this:
<h1></h1>
<h2></h2>
<h2></h2>
Or to have non-empty div tags like this:
<div class="headline1">Welcome to this site!</div>
<div class="headline2">Top stories:</div>
It occurs to me that the PHP code might just print the content, like:
$dbArray = $sql->getWeblogContent();
extract($dbArray);
<h1> $headline </h1>
<h2> $subheadline1 </h2>
<div> $maincontent </div>
In this way, the graphic designers who design the sites could have
total freedom about what HTML tags they wrap that code in. They'd
write all the tags themselves. However, sometimes there would be no
content and there would be a lot of empty tags. How bad is that? Is
Google smart enough to simply ignore empty tags?
The other possibe strategy is to have the code print nothing if there
is nothing to print, like this:
$dbArray = $sql->getWeblogContent();
extract($dbArray);
if ($headline) echo "<h1> $headline </h1>";
if ($subheadline1) echo "<h2> $subheadline1 </h2>";
if ($maincontent) echo "<div> $maincontent </div>";
The advantage of this is there are no empty tags, the disadvantage is
that the HTML is hard-coded, and the graphic designer can't change any
of it. And, as this thread makes clear, some designers will take issue
with the choice of tags. More so, true freedom of design should mean
changing those tags. On my own personal site, the weblog title is 12px
and the each weblog entry title is big. I'd use a <p> tag for the
weblog title and an <h2> for the weblog entry title. But on most
sites, the weblog title is the biggest thing on the page, and the
entry titles are more like 12px.
The question, really, is how damaging is it to the sense of
"structure", when speaking of a structured document, to have a lot of
empty tags? Toby's convinced me that the empty tag approach is best
because of the need to maximize the freedom of the designers. More so,
if the code writes no tags, then the separation between code and
design is cleaner. But, again, I wonder if it is considered poor form
to have a lot of empty tags on the page?
William Tasso
> ...
> I wonder if it is considered poor form
> to have a lot of empty tags on the page?
it is certainly a waste of bandwidth and it doesn't say much about the
structure of the document
lawrence
Thank you, your suggestion was useful.