Different screen resolutions

[Guest]

CSS *is* easy. It's browser bugs that make things difficult.

Point taken.

M

 

[C A Upsdell]

Which article was that? I'm intrigued. By the way, browser detection by
virtually any means is flawed. Too many vendors make determined attempts
to spoof themselves for it to be reliable.

Spoofing is not as much a problem as you might think. E.g., even when Opera
is spoofing another browser it is quite easy to recognize that it is Opera.
Difficulties in correctly detecting the browser are mainly due to programmer
incompetence: e.g., with Opera the userAgent string always contains the
substring 'Opera', even when configured (say) to spoof IE, but many an
incompetent programmer simply checks for 'MSIE' and then blithely assumes
Internet Explorer. Other browsers present similar problems, e.g. Safari can
be mistaken for Gecko if the programmer checks for 'Gecko' instead of first
checking for 'Safari'.

In practice spoofing becomes a problem only when the userAgent string not
only spoofs another browser, but also hides its own identify: and this is
rare.

CSS *is* easy. It's browser bugs that make things difficult. In my
experience, IE causes the most problems, but older versions of probably
every browser make a contribution.

Indeed.

 

[Dave Patton]

To Dave Patton (& his fans)

I doubt I have any "fans" in this group, nor do I make
postings in order to try and get "fans".
Brucie has "fans", but they just want to hear the voices ;-)

You may think that simply putting down the phrase "fluid design"

Which is not what I did. The substance of my reply was to
point out the flaw of assuming an 800 pixel desktop width
equates to 800 pixels of browser viewport width.

Something more specific like an example website

If you look at the timestamps, you'll see my posting
was made after David Dorward's, where he provided you
with a URL. It seemed pointless to me to repeat it.

 

[Beauregard T. Shagnasty]

In practice spoofing becomes a problem only when the userAgent
string not only spoofs another browser, but also hides its own
identify: and this is rare.

What browser am I using? <g> Unaltered UA string:

Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1)

 

[C A Upsdell]

What browser am I using? <g> Unaltered UA string:

Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1)

That is an IE6 userAgent for Windows XP.

It cannot be Opera. It could be iCab (iCab lets the user enter the entirety
of their own userAgent string, but iCab is uncommon); it could be Mozilla
(or maybe Firebird) if the user has used a plug-in which can do this (rare,
but possible if the user chooses to do so in order to use a site which is
too damn stupid to handle Gecko-based browsers ... i.e. a small % of sites).
Other possibilities are unlikely -- not impossible -- but sufficiently
unlikely (in % of users) that I would not worry.

 

[Michael Winter]

[snip]

Too many vendors make determined attempts to spoof themselves for
[browser detection] to be reliable.

Spoofing is not as much a problem as you might think.

I know, because the problem is browser detection itself.

[snip]

In practice spoofing becomes a problem only when the userAgent string
not only spoofs another browser, but also hides its own identify: and
this is rare.

It may be rare for your audience but it does happen, and as long as pages
are written, restricted to a particular browser, vendors will feel
inclined to spoof.

The incompetence you mentioned isn't so much how the detection is
accomplished, but that the author believes he needs to know which browser
is being used in the first place. Most of the time, it's completely
irrelevant.

[snip]

Mike

 

[Beauregard T. Shagnasty]

That is an IE6 userAgent for Windows XP.

That is correct. But I don't have XP.

It cannot be Opera.

Correct, it is not Opera.

It could be iCab (iCab lets the user enter the entirety of their
own userAgent string, but iCab is uncommon);

It's not iCab.

it could be Mozilla (or maybe Firebird) if the user has used a
plug-in which can do this (rare, but possible if the user chooses
to do so in order to use a site which is too damn stupid to handle
Gecko-based browsers ... i.e. a small % of sites).

Now you're getting somewhere. <g> It was Firefox, using the PrefBar
extension, which puts UA spoofing right there on the toolbar.
http://home.rochester.rr.com/bshagnasty/images/mozbar.png
...which works equally well with Moz or Firefox. See "Real UA" to the
right, which is where I selected "IE6 WinXP".

You can also add a completely different UA string by changing the
general.useragent.override in the prefs.js file.

Other possibilities are unlikely -- not impossible -- but
sufficiently unlikely (in % of users) that I would not worry.

Agreed that not too many people know how to do this. Oh, checking the
header of my post would have given a mighty clue. <g>

 

[Toby Inkster]

Understand you. My question was more to do with aesthetics. Right now, I've
left my banner left-aligned. At higher rez, that opens up wasted real estate
to the right that I want to "fill" with something that still pleases the
eyes. My banner is a photo, not suitable for tiling.

Google: CSS sliding doors

My site: scroll down to the "site style" form and choose "Modern 2" or
"Modern 3".

 

[Toby Inkster]

with Opera the userAgent string always contains the substring 'Opera',
even when configured (say) to spoof IE [...] Safari can be mistaken for
Gecko if the programmer checks for 'Gecko' instead of first checking for
'Safari'.

And don't forget: MSIE spoofs Mozilla!

 

[Toby Inkster]

Agreed that not too many people know how to do this. Oh, checking the
header of my post would have given a mighty clue. <g>

FWIW, there is a Windows registry setting that allows you to edit the MSIE
user-agent header.

 

[C A Upsdell]

with Opera the userAgent string always contains the substring 'Opera',
even when configured (say) to spoof IE [...] Safari can be mistaken for
Gecko if the programmer checks for 'Gecko' instead of first checking for
'Safari'.

And don't forget: MSIE spoofs Mozilla!

Well, the original Moz, true.

 

[C A Upsdell]

On Sat, 13 Nov 2004 16:19:05 -0500, C A Upsdell
The incompetence you mentioned isn't so much how the detection is
accomplished, but that the author believes he needs to know which browser
is being used in the first place. Most of the time, it's completely
irrelevant.

For most things, you're right, it is irrelevant: or at best, a last resort.
But sometimes it's an unavoidable last resort: sometimes a browser
misbehaves in a manner that must be avoided, and sometimes alternate
solutions like object detection and CSS trickery just don't do the job.

 

[C A Upsdell]

Now you're getting somewhere. <g> It was Firefox, using the PrefBar
extension, which puts UA spoofing right there on the toolbar.
http://home.rochester.rr.com/bshagnasty/images/mozbar.png
..which works equally well with Moz or Firefox. See "Real UA" to the
right, which is where I selected "IE6 WinXP".

But IMO this extension should only be used when nothing else will make a
particular site usable. And perhaps the userAgent string this extension
sets is poorly chosen. Instead of:

Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1)

I suggest something like:

Mozilla/5.0 (Windows; U; MSIE 6.0; Windows NT 5.1; en-US; rv:1.7.3)
Gecko/20040910 Firefox/1.0

which enables a competent programmer to correctly detect it as a Gecko
browser, but tricks a naive programmer into thinking it is IE 6.

You can also add a completely different UA string by changing the
general.useragent.override in the prefs.js file.

But again, it would be better to spoof another browser without hiding the
real identify of the browser. And in any case, such a technique should be
adopted only if a site intransigently refuses to work with your browser's
standard userAgent.

 

[kchayka]

FWIW, there is a Windows registry setting that allows you to edit the MSIE
user-agent header.

FWIW, I wouldn't recommend messing with it. I did once, and forgot about
it. The next time I went to Windows Update it totally barfed. It took a
while before I figured out the altered UA was the cause.

Stoopid MS.

 

[Beauregard T. Shagnasty]

But IMO this extension should only be used when nothing else will
make a particular site usable.

That's the idea, for all those clueless sites that require you to have
"Internet Explorer 4 or better"

And perhaps the userAgent string this extension sets is poorly
chosen. Instead of:

Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1)

(which is the UA when spoofing IE6)

I suggest something like:

Mozilla/5.0 (Windows; U; MSIE 6.0; Windows NT 5.1; en-US; rv:1.7.3)
Gecko/20040910 Firefox/1.0

Firefox 1.0 real UA is (or at least mine is):
Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7.5)
Gecko/20041107 Firefox/1.0
[with Win2K]

which enables a competent programmer to correctly detect it as a
Gecko browser, but tricks a naive programmer into thinking it is IE
6.

Trouble is most sites don't have competent programmers.

But again, it would be better to spoof another browser without
hiding the real identify of the browser.

But what does it matter? Why should I or anyone else care if the web
site knows if I'm using IE or something else - those sites that don't
sniff?

And in any case, such a technique should be adopted only if a site
intransigently refuses to work with your browser's standard
userAgent.

Again, that is the idea!

 

[Henry]

Please refrain from ever giving HTML advice again.

-Karl

Great!

You have an better idea how to design a liquid table for a banner, huh?

Soooo........???????

 

[Kris]

[tables for layout example]

You have an better idea how to design a liquid table for a banner, huh?

Can I choose something different than a table?
There are so many cuty elements to pick from.. let's see..

<div class="ad">
<a ...>
<img alt="Ad: Vote Karl Core" ...>
</a>
</div>

..ad {
text-align: center;
}

 

[Michael Winter]

On Sat, 13 Nov 2004 22:14:43 -0500, C A Upsdell

[snip]

I suggest something like:

Mozilla/5.0 (Windows; U; MSIE 6.0; Windows NT 5.1; en-US; rv:1.7.3)
Gecko/20040910 Firefox/1.0

which enables a competent programmer to correctly detect it as a Gecko
browser, but tricks a naive programmer into thinking it is IE 6.

The problem with that line of thinking is that you forget that incompetent
programmers use code produced by the competent. Now some clueless author
can correctly detect your browser, and bar you from the site.

[snip]

Mike

 

[Michael Winter]

On Sat, 13 Nov 2004 22:04:44 -0500, C A Upsdell

[snip]

But sometimes [browser detection is] an unavoidable last resort:
sometimes a browser misbehaves in a manner that must be avoided, [...]

True and I acknowledge that. However, it's important to consider whether
the behaviour can be detected by run-time before resorting to browser
detection. Most won't.

Mike

 

[Karl Core]

Great!

You have an better idea how to design a liquid table for a banner, huh?

Soooo........???????

A <div> would be a great start.

BTW - stop morphing, you fucking troll.

-Karl