META - Liquid design and text size

"brucie" <brucie@loser.brucies.com> wrote in message
news:bdd0kd$ofmte$1@ID-117621.news.dfncis.de...
> In post <xqmKa.913$GA6.136@newssvr17.news.prodigy.com>
> Chris Beall said...
>
> > I later found that I could 'liquefy' the image size
>
> images should only be displayed at their actual size (with a few
> exceptions)

OK, I'll bite. Why? And what are the exceptions?

>
> > by using something like WIDTH="50%" on IMG tags, but the
> > most-often-used browser responded by refusing to display the image
at
> > all.
>
> good, browser resized images look crap.

I can see how that would be the case if the browser was asked to display
the image at a resolution greater than that of the image I provided, but
it seems to me that if I provide an image at the highest-likely
resolution, the browser should be able to REDUCE the resolution, with
good fidelity for that target resolution. I understand this would make
the downloaded image large, even when sent to a low-resolution client,
thus making this a technique to be used cautiously.

Did I miss anything?

> > What I want is for the height and width of text characters to scale
with
> > changes to resolution and window size.
>
> stop wanting that. let the visitor decide what size they want their
> font.

I agree that the visitor should be able to select font size, however I
suspect that upwards of 95% of visitors have no idea that they can do
this, much less HOW to do it. I'm trying to provide a rational starting
point, not preempt the visitor's choice. (In Netscape, the visitor can
override what I specify; that seems a good approach for a browser to
take).

Regardless of whether we agree on whether or not this is a Good Thing to
do, the question remains: can it be done? I expect the current answer
is 'no'.

After further thought, I've concluded that the mechanism for doing this
exists, but is not implemented (in any browser I know of). It is the
'point' or 'pica' specification, which is SUPPOSED to define a font in
fractions of an inch. A 72-point font should therefore appear 1" high,
regardless of display resolution or glass size. At present, however, I
know of no standardized way for the browser to even be aware of the
glass size, though it could certainly be made part of the input provided
by a user in their (slightly distorting the meaning of the
word)Preferences.

>
>
> --
> brucie a. blackford. 26/June/2003 06:24:47 am kilo.
> http://loser.brucies.com/

Thanks for your comments,
Chris Beall

Chris Beall wrote:

> I agree that the visitor should be able to select font size
[...]
> I'm trying to provide a rational starting point

That would be 100% ;-)

--

Mark Parnell
http://www.clarkecomputers.com.au

"Mark Parnell" <webmaster@clarkecomputers.com.au> wrote in message
news:3efb8154$0$22145@echo-01.iinet.net.au...
> Chris Beall wrote:
>
> > I agree that the visitor should be able to select font size
> [...]
> > I'm trying to provide a rational starting point
>
> That would be 100% ;-)

Hmmm. OK, assume I did that, i.e. I didn't specify any font-size at
all. And assume the user has specified the default font size as 16
point, which is therefore 16/72" high. Three questions:
1. How high should a capital 'X' be, on the screen, in inches?
2. Should the answer change if the user changes screen resolutions?
3. Should the answer change if the user changes the window size?

I think the answers should be: 1. 16/72" 2. No 3. No (reversing
what I said in a previous post).

By asking the question this way, we get away from the religious issue
and back to the technical one. Regardless of who has specified the
target font size, the user or web designer, what should happen to it as
the display environment changes? Once we agree on that, we will
probably agree that the web designer who DOES chose to specify font
sizes is entitled to get the same result across environment changes as
the default would.

Regards,
Chris Beall

> --
>
> Mark Parnell
> http://www.clarkecomputers.com.au
>
>

Chris Beall wrote:
>
> What I want is a font that stays the same size at all resolutions

There is no font that stays exactly the same size at all resolutions.
There will always be a visual difference with any given absolute font
size at different screen resolutions. Even when it's a graphic. The
only thing that you will accomplish is making the text difficult reading
for some portion of your visitors, even impossible for some. There is
no benefit, so don't do it.

> and does NOT scale based on window size.

I'm not aware of any font-size unit that scales with browser window
size. So don't specify any size and you'll get your wish. If you
really mean screen size, then the answer is the same as above.

> I think that's what the 'point' specification should provide.

pt units are not appropriate for screen display, but are intended for
media such as print where the dot pitch can be reasonably predicted.
You cannot predict how pt will be rendered on screen, so it will
invariably be the "wrong" size for some portion of visitors, perhaps a
significant portion. There is no benefit, so don't do it.

> My concern was
> for the naive user who decides to use a high-resolution screen and can't
> understand why everything got so tiny.

And you believe that the first time that user notices this is when they
get to _your_ web page? Highly unlikely. Changing display settings to
a different resolution is going to have effects that are immediately
noticable in every app. By the time that user gets to your web page,
they have already decided what to do about it. Your misguided attempt
at compensating for this does not help them, but more likely has the
opposite effect. So don't do it.

--
To email a reply, remove (dash)ns(dash). Mail sent to the ns
address is automatically deleted and will not be read.

"kchayka" <kcha-ns-yka@sihope.com> wrote in message
news:3efdbefe$1@news.sihope.com...
(snip)
> Q: Should the user take the responsiblitiy of adjusting their default
> font size when they change resolutions?
> A: Yes. With an adjusted default font size, the capital 'X' will
still
> be as high as the user deems necessary for optimal readability. It
> doesn't matter if the actual size is 10pt or 36pt.

There seems to be an inconsistency here. On the one hand, you are
saying that when the user goes from web site to web site, they should
expect fonts to remain the same size, but when the user changes screen
resolution, they should expect to have to adjust the font size to get
back to the one they prefer. That seems to say that web site designers
should defer to the user's preference, but browser and OS coders need
not. This seems to be a widely accepted view, but I find it puzzling.

Chris Beall

Chris Beall wrote:

> "kchayka" <kcha-ns-yka@sihope.com> wrote in message
> news:3efdbefe$1@news.sihope.com...
> (snip)
>> Q: Should the user take the responsiblitiy of adjusting their default
>> font size when they change resolutions?
>> A: Yes. With an adjusted default font size, the capital 'X' will
> still
>> be as high as the user deems necessary for optimal readability. It
>> doesn't matter if the actual size is 10pt or 36pt.
>
> There seems to be an inconsistency here. On the one hand, you are
> saying that when the user goes from web site to web site, they should
> expect fonts to remain the same size,

When the user preferences take precedence, there is no reason to expect
drastic differences in font sizes from site to site. It's when an
author attempts to override user preferences that things start going badly.

> but when the user changes screen
> resolution, they should expect to have to adjust the font size to get
> back to the one they prefer.

Well, yeah, of course. What works well for a default font size at
800x600 96dpi is most likely not the same as that same monitor (whatever
size) at 1280x1024 96dpi. But this is a change the user only has to
make once, when the settings are initially changed. After that, it's
the same as above.

> That seems to say that web site designers
> should defer to the user's preference, but browser and OS coders need
> not. This seems to be a widely accepted view, but I find it puzzling.

I don't understand your confusion. Vendors release software with
"factory settings" that probably do OK for most users with default
hardware settings. If the factory settings aren't to your liking, you
should change them. Likewise, if you change those default hardware
settings (like screen resolution), those factory software settings may
no longer be suitable and should be changed.

--
To email a reply, remove (dash)ns(dash). Mail sent to the ns
address is automatically deleted and will not be read.

"kchayka" <kcha-ns-yka@sihope.com> wrote in message
news:3efdf5ca$1@news.sihope.com...
(snip)
> When the user preferences take precedence, there is no reason to
expect
> drastic differences in font sizes from site to site. It's when an
> author attempts to override user preferences that things start going
badly.

OK, I think I'm coming around here. Now, can you help me out by
pointing me to a few real, commercial sites that do it 'right' from your
perspective?

Thanks,
Chris Beall

Chris Beall wrote:

> "kchayka" <kcha-ns-yka@sihope.com> wrote in message
> news:3efdf5ca$1@news.sihope.com...
> (snip)
>> When the user preferences take precedence, there is no reason to
> expect
>> drastic differences in font sizes from site to site. It's when an
>> author attempts to override user preferences that things start going
> badly.
>
> OK, I think I'm coming around here. Now, can you help me out by
> pointing me to a few real, commercial sites that do it 'right' from your
> perspective?

Sure wish I could think of some, but alas no.

BTW, I don't really spend a huge amount of time at commercial sites. I
mostly frequent information sites. The big news sites, like cnn, are
just as bad as typical commercial sites. wired.com, however, redesigned
their whole site not long ago and it is much better than average. It's
not perfect, but somebody there is surely more clueful than most.

--
To email a reply, remove (dash)ns(dash). Mail sent to the ns
address is automatically deleted and will not be read.