jQuery Mobile--some people never learn

[David Mark]

http://jquerymobile.com/2010/08/announcing-the-jquery-mobile-project/

The dubious tagline appears to be "seriously cross-browser and cross-
device". Perhaps it escapes them that they have yet to produce a
cross-browser script for desktop browsers.

"Absolutely critical to us is that jQuery and the mobile UI framework
that we’re developing work across all major international mobile
platforms (not just a few of the most popular platforms in North
America)."

A framework that will work across "all browsers"?

"We’ve published a complete strategy overview detailing the work that
we’re doing and a chart showing all the browsers that we’re going to
support."

That's what I thought. Their "browser grading" chart demonstrates
they still have no idea what cross-browser scripting entails. And
their "strategy overview" has the nerve to mention progressive
enhancement (something else they've never supported in any meaningful
way).

All it should take is a look backward at all of the past browsers that
are no longer "supported" by jQuery to realize that they've been
faking it the whole time.

"“The jQuery community has focused on making the Web as productive and
fun as possible. When we heard the mission behind jQuery Mobile, we
wanted to help. With webOS we have shown that the Web platform is
fantastic for developers, so we are excited to help make jQuery Mobile
as good as it can be.” -Dion Almaer"

Isn't that one of the Ajaxian twins? Does it really escape these
people that jQuery is a horrible script (not to mention completely
inappropriate for mobile devices)?

This from Mozilla:-

"“As a longtime supporter of the jQuery project and its wider
community we are excited to extend our support to the jQuery Mobile
project. jQuery Mobile has the potential to make cross-platform Open
Web development significantly simpler.” -Pascal Finette"

What the hell are they thinking?

This from some "design studio" that apparently specializes in jQuery
add-ons:-

"“Filament is thrilled to sponsor and lead the design for the jQuery
Mobile Framework; it’s a great opportunity to extend our work on
ThemeRoller and jQuery UI. We’re especially pleased that progressive
enhancement will be built right into jQuery Mobile – we think this
positions it really well for broad accessibility and future
compatibility.”"

PE will be "built right into" jQuery? Not unless they are planning on
re-designing it from scratch (and Resig has already indicated that
they are going with the same old tired code). And what are the two
things that most modern mobile devices need the least? Queries and
scripted animations come to mind.

Clicking the link to the Filament site brought up the typical fixed-
width skinny sliver design that wastes most of the real estate on the
average monitor. That's fairly common for Web designers who seem to
be oblivious to wide-screen monitors. Sure, a fluid layout can be
"too wide" when the browser is maximized, but resizing the browser is
easy compared to changing the screen resolution, which would render
the typical font choices illegible anyway.

But that's not the worst of it. At the bottom of the page was a link
to a "desktop version", which (of course) rendered a horizontal scroll
bar as it was fixed too wide. Nothing to do there but reduce the
browser's font size (assuming IE is not in use) until the document
fits, at which point the text becomes completely illegible. Well,
that or disable CSS (something I find myself doing quite often, which
defeats the whole purpose of "designers").

And why they decided to initially present their "mobile version" to
Opera 10 is beyond me. Failed browser sniffing perhaps?

Then there is this uncredited note:-

"We’re thrilled to be working on this project. The mobile web is
desperately in need of a framework that is capable of working across
all browsers, allowing developers to build truly mobile web sites.
We’re doing all we can to ensure that jQuery Mobile fills that need."

This bunch is going to create a "framework" that is capable of working
across "all browsers"? Not if history (or their announced strategy)
is any indicator.

And their lead designers have certainly not figured out how to build
"truly mobile Websites". Hint: fluid layout + media queries +
handheld style sheets. One site fits all. No magic frameworks
required. In fact, very little in the way of scripting is needed,
even for the most animated of sites.

"Comments are closed"

Of course they are.

 

[S.T.]

Clicking the link to the Filament site brought up the typical fixed-
width skinny sliver design that wastes most of the real estate on the
average monitor. That's fairly common for Web designers who seem to
be oblivious to wide-screen monitors. Sure, a fluid layout can be
"too wide" when the browser is maximized, but resizing the browser is
easy compared to changing the screen resolution, which would render
the typical font choices illegible anyway.

But that's not the worst of it. At the bottom of the page was a link
to a "desktop version", which (of course) rendered a horizontal scroll
bar as it was fixed too wide. Nothing to do there but reduce the
browser's font size (assuming IE is not in use) until the document
fits, at which point the text becomes completely illegible. Well,
that or disable CSS (something I find myself doing quite often, which
defeats the whole purpose of "designers").

I get a little smile out of David "my site looks like an eighth-grade
class project" Mark critiquing web designs and layout.

 

[RobG]

http://jquerymobile.com/2010/08/announcing-the-jquery-mobile-project/

The dubious tagline appears to be "seriously cross-browser and cross-
device". Perhaps it escapes them that they have yet to produce a
cross-browser script for desktop browsers.

Of course no critical analysis is ever done of such claims. e.g.

"Yahoo has one of the best overall strategies regarding
desktop browser web development with their graded browser
support chart."

A strategy is a policy to achieve an outcome. What is the outcome that
a graded browser support chart is supposed to achieve? Its purpose is
to indicate which browsers may or may not be supported and to what
extent. That isn't a strategy, it is a communication tool.

The strategy is to categorise browsers into ones that are or aren't
"supported", you don't need a colour chart for that.

And how is "best" defined? What other strategies are available? How do
they compare? What are the criteria for a successful browser support
strategy?

The simplest and most obvious strategy to me is to *not* have a
specific list of supported browsers and to write defensively for
standards compliant browsers with a strategy for dealing with non-
standards compliant behaviour or missing functionality. Then you don't
care about any list of supported browsers, though you might keep a
list of quirks and non-standard behaviours that are or aren't dealt
with.

Isn't that the basis of the argument for feature detection? Didn't
jQuery finally realise that feature detection (and similar strategies)
are superior to browser detection? A supported browser matrix is
effectively equivalent to browser detection - if a user has an issue
with their browser, a quick look at the chart and they're likely to be
told "sorry, that browser isn't on my support matrix".

"Compared to mobile web development, the potential browser
choices in desktop web development seems downright simple.
In mobile development there are more engines, on more
platforms, and with more active versions of the browsers."

You can see the solution coming... they're going to pick an handful of
browsers they like and simply ignore the rest.


"When we look at the major browsers that are available, we
need to figure out what platforms they’re running on and
what versions of those browsers work well-enough to support."

So there it is. Having stated that there are a zillion mobile browsers
and platforms, the solution is to list a few browsers on a few
platforms and the rest be damned. Someone in the development team (a
"we") has decided specifically which browsers they will and won't
support. A decision apparently based on some unstated "quality" and
"relevance in the larger mobile market" criteria. In other words, some
unnamed person or person's opinion of what will and wont be supported.

Interesting that no "native" browser on any version of Windows Mobile
is to receive A grade support. Depending on whose statistics you
believe, Windows Mobile may have up to 20% market share in northern
USA, but that isn't enough for A grade support.

Isn't deciding on which browsers will and won't be supported the
antithesis of the ethos of the web? Isn't that precisely the attitude
that the web, and browsers in general, are designed to combat? Do we
have to keep going back over the same old arguments time and time
again?

Anyway, it will be wildly successful, web developers will clutter
their sites with crappy animations and overweight scripts and images
that clog the meagre bandwidth and processing power available to
mobile users. And they'll gleefully tell their clients that they
absolutely must have both a desktop and mobile version of their site,
and that it's impossible to have one site for everyone since the
browser support matrix for mobile is completely different to the one
for desktop.

Where do other devices fit into this picture like tablets, TVs and
consoles? Oh, that's right, they don't. So next there'll be jQueryTV
and jQueryConsole and so on - heck, more bespoke sites.

And browser developers won't be game to release a browser that "isn't
compatible with jQuery (mobile/TV/Console...)". I have seen the
future. Again.

 

[S.T.]

Many/most movie critics can't act yet are qualified to criticise movies.
Many/most art critics can't draw yet are qualified to criticise art.
Many/most sports commentators can't play sport at an elite level yet are
qualified to criticise elite sport.
Many/most political reporters have never been politicians yet are
qualified to criticise politics.

That you notice that a programmer can't design does not negate the
programmer's criticism of a design.

Critics should, in theory, understand the subject they're critiquing.
David doesn't, in this case.

I like running my browser full-width on a fairly large monitor. I like
it as there are sites that know how to utilize wide browser windows:

http://imgur.com/4teVo.png
http://imgur.com/NpB5F.png

.... yet also realize when *not* to use a wide-browser "real estate" and
adjust their layouts accordingly:

http://imgur.com/yywhJ.png

Unfortunately there's a decent population that has absolutely no idea
what they're doing. They probably read some article about fixed-vs-fluid
but didn't grasp the concept of context... and end up with an absolute
mess on wide browser windows:

http://imgur.com/9BmaE.png

Now it's fine that his design isn't flawless. Not that big a deal. It
would be absurd to expect every page on the internet to perfectly
accommodate unknown and seemingly infinite browser window dimensions.

The site David critiques, however, actually *does* get it right. The
primary layout is suited for 1024px+ monitors, perfectly reasonable,
with additional 'basic' (CSS doesn't touch layout) and mobile options.
The typography is immaculate and the readability is excellent. I'm sure
I could find faults if I devoted some time to it, but generally speaking
it's an excellent design and layout.

So when DM decides to write something like...

"Filament site brought up the typical fixed-width skinny sliver design
that wastes most of the real estate on the average monitor. That's
fairly common for Web designers who seem to be oblivious to wide-screen
monitors."

.... about a site that understands wide-browser layout, while his own
site perfectly illustrates what can go horribly wrong when you don't
prepare for wide browsers, I chuckle a little bit.

 

[Simon Brooke]

I like running my browser full-width on a fairly large monitor. I like
it as there are sites that know how to utilize wide browser windows:

http://imgur.com/4teVo.png
http://imgur.com/NpB5F.png

... yet also realize when *not* to use a wide-browser "real estate" and
adjust their layouts accordingly:

http://imgur.com/yywhJ.png

De gustibus nil disputandum est.

I think Google's design here is at best questionable. I hate designs
which cram all the content into the left hand gutter and leave a desert
of empty space to the right.

Very long text lines are not easy on the eye, so it makes sense to have a
column narrower than the page; but there's no rule that says one may have
only one column of main text content. And if one has only one column of
main content, then, for heaven's sake, float it into the middle of the
viewport!

 

[David Mark]

I get a little smile out of David "my site looks like an eighth-grade
class project" Mark critiquing web designs and layout.

Ah, it's the anonymous Stupid Twit. Here's the thing: my layouts are
way ahead of anything I've criticized here. You simply don't have a
clue what you are talking about (as usual).

And you are making the typical mistake of confusing graphics design
with layout. I don't do graphics, nor do I care about them (or clods
like you).

For more information on my (present and future) layout schemes:-

http://groups.google.com/group/my-library-general-discussion/browse_thread/thread/cd3a7e6781036b1d#

For those who still think fixed-width, pixel-sized "designs" make
sense, good luck writing new designs for each new type of device that
comes out. There will be a lot of them.

 

[David Mark]

Critics should, in theory, understand the subject they're critiquing.
David doesn't, in this case.

I don't understand layout? LOL.

I like running my browser full-width on a fairly large monitor.

So what? There's no issue with making it narrower as a fluid design
will adapt. On the other hand, a fixed-width design that is too large
(or too small) can't be helped. Get it?

I like
it as there are sites that know how to utilize wide browser windows:

   http://imgur.com/4teVo.png

What about that?

   http://imgur.com/NpB5F.png
Same.


... yet also realize when *not* to use a wide-browser "real estate" and
adjust their layouts accordingly:

   http://imgur.com/yywhJ.png

So they put a max-width rule on their search results. I certainly
agree that such a rule makes sense for some content. But again, so
what?

Unfortunately there's a decent population that has absolutely no idea
what they're doing.

Or saying. Like you.

They probably read some article about fixed-vs-fluid
but didn't grasp the concept of context... and end up with an absolute
mess on wide browser windows:

   http://imgur.com/9BmaE.png

That test page doesn't look messy at all. If it is too wide for your
tastes, then narrow *your* browser and it will happily oblige. You
can't do that with fixed-width designs, can you?

Now it's fine that his design isn't flawless.

Particularly as it is a test page using a three-year-old layout that I
slapped together in an hour. Try one of my sites on Opera Mini or
read the link I posted earlier that explains how easily such a design
can be retrofitted to be "flawless" on iPhones, iPads, etc. IOW, do
as I say, not as I did.

Not that big a deal. It
would be absurd to expect every page on the internet to perfectly
accommodate unknown and seemingly infinite browser window dimensions.

Define "perfect" in the context of such subjective criticism. Oh God,
what am I saying? I meant, shut the hell up and bugger off.

The site David critiques, however, actually *does* get it right.

It doesn't get anything right. Perhaps you should read my post
again. It came up in "Mobile view" in Opera 10, taking up maybe 10%
of my screen real estate (can't fix that by resizing the browser
window). On clicking "desktop view", it grew a huge horizontal scroll
bar (can't fix that either). I ended up disabling CSS entirely, thus
making every effort of the "designer" for naught.

The
primary layout is suited for 1024px+ monitors, perfectly reasonable,
with additional 'basic' (CSS doesn't touch layout) and mobile options.

See above. And what the hell is a "1024px monitor"?

The typography is immaculate and the readability is excellent.

Typography?! Ironic that my site is actually aligned to a baseline
grid. Was that one? I didn't notice before I turned off CSS so I
could *read* it.

You are projecting your own setup on to everyone else (e.g. "1024px
monitor", insistence on maximizing the browser, etc.) That's a
typical rookie mistake.

I'm sure
I could find faults if I devoted some time to it, but generally speaking
it's an excellent design and layout.

As you see it (hint).

So when DM decides to write something like...

  "Filament site brought up the typical fixed-width skinny sliver design
that wastes most of the real estate on the average monitor.  That's
fairly common for Web designers who seem to be oblivious to wide-screen
monitors."

That's what it did.

... about a site that understands wide-browser layout,

The *site* understands wide-browser layout?! I doubt that. Certainly
the developers do not as their "design" wastes a huge portion of it
for no reason. They used a pixel-based fixed-width layout (circa
1998) and (by coincidence) it looked pleasing to you on your monitor
and browser configuration. That's not understanding, but good luck on
your part.

while his own
site perfectly illustrates what can go horribly wrong when you don't
prepare for wide browsers, I chuckle a little bit.

You are simply a laughingstock. Or you would be if you even had an
identity. Scared of your own shadow?

 

[David Mark]

Many/most movie critics can't act yet are qualified to criticise movies.
Many/most art critics can't draw yet are qualified to criticise art.
Many/most sports commentators can't play sport at an elite level yet are
qualified to criticise elite sport.
Many/most political reporters have never been politicians yet are
qualified to criticise politics.

That you notice that a programmer can't design does not negate the
programmer's criticism of a design.

One quibble. Other than graphics, I certainly can design. Web design
is about far more than graphics.

 

[David Mark]

[...]

Typography?!  Ironic that my site is actually aligned to a baseline
grid.  Was that one?  I didn't notice before I turned off CSS so I
could *read* it.

Of course it isn't. Also, disable scripting and it loses *all* of its
"immaculate" design. That's about as clueless as it gets (and no
wonder it takes so long to ooze into place on load).

Graphics are commodities, competent Web designs are not.

 

[David Mark]

Of course no critical analysis is ever done of such claims. e.g.

 "Yahoo has one of the best overall strategies regarding
  desktop browser web development with their graded browser
  support chart."

Yes, we've got one group of fools copying another. The YUI developers
are so completely unsuited to their chosen task that they might as
well give it up. Like the Dojo developers, their mindset is all wrong
(e.g. "show me where it fails", "browser sniffing is pragmatic",
etc.) Virtually every blog post and commit of theirs is a howler.

A strategy is a policy to achieve an outcome. What is the outcome that
a graded browser support chart is supposed to achieve? Its purpose is
to indicate which browsers may or may not be supported and to what
extent. That isn't a strategy, it is a communication tool.

And an instantly dated communication. It amazes me that these guys
don't understand that today's observations may vanish tomorrow (and
may look radically different in some of today's environments as
well). If the purpose of these cults is to abstract all of the
browser variances out of the picture, which assumes there is a serious
need to do so and that there are far too many variations for any one
developer to handle on their own, how do they justify such folly as
assuming logic is sound based on whether it fails on their installed
browsers *today*.

They just keep rearranging their code until it seems to work in the
handful of browsers they aspire to "support" and agents that fail
(e.g. last year's browser versions) are deemed unworthy of their
arrangements (and therefore "downgraded" to unsupported status).

They've been doing the same Sisyphus act for years now and still no
inkling that maybe they are approaching this stuff the wrong way.
Along the way, they've turned JS into a bad joke and enabled
proprietary applications (e.g. the Apple store) to gain a significant
foothold. I've warned them all repeatedly, yet their brains seem to
short-circuit in the face of technical criticism, resulting in
accusations of jealousy.

The strategy is to categorise browsers into ones that are or aren't
"supported", you don't need a colour chart for that.

But color charts are really impressive (to the hopelessly
impressionable). That's why I put one on my front page. Never mind
that it has no units.

And how is "best" defined? What other strategies are available? How do
they compare? What are the criteria for a successful browser support
strategy?

Aw, you are thinking too much. But seriously, Web developers
generally don't want to have to use their brains. They want to be
told what to think in very simple and absolute terms.

The simplest and most obvious strategy to me is to *not* have a
specific list of supported browsers and to write defensively for
standards compliant browsers with a strategy for dealing with non-
standards compliant behaviour or missing functionality.
Then you don't
care about any list of supported browsers, though you might keep a
list of quirks and non-standard behaviours that are or aren't dealt
with.

Indeed. But defensive coding is not in these guys. They'll use any
old hack if it appears to work in the three browsers on their
desktops. Never mind how nonsensical it is. If you can't show them
where it fails *today*, they run with it. In their minds, they will
fiddle with it when it breaks tomorrow and all of their users will
instantly download, re-test and re-deploy their new versions. I guess
it keeps them off the streets anyway.

Isn't that the basis of the argument for feature detection? Didn't
jQuery finally realise that feature detection (and similar strategies)
are superior to browser detection?

Kinda sort of. About a year after Resig and I had out little
"discussion" about browser sniffing merits (or the lack thereof).
They moved up a notch on the learning curve to object inferences,
which are just prettied up browser sniffs.

A supported browser matrix is
effectively equivalent to browser detection - if a user has an issue
with their browser, a quick look at the chart and they're likely to be
told "sorry, that browser isn't on my support matrix".

Exactly. Whomever wrote that copy was talking out of both sides of
their mouths. Alternating paragraphs referred to "all browsers" and
"a list of supported browsers". How any reasonable developer could
read it and not burst out laughing is beyond me.

 "Compared to mobile web development, the potential browser
  choices in desktop web development seems downright simple.
  In mobile development there are more engines, on more
  platforms, and with more active versions of the browsers."

You can see the solution coming... they're going to pick an handful of
browsers they like and simply ignore the rest.

Of course. That's what they've always done. It's the dirty little
"secret" that none of them like to talk about. They come up with
grand, untenable designs and then spend years hacking and tweaking to
try to make them work in three or four browsers. If only they would
set reasonable goals, they could avoid the perpetual failures. But
reasonable goals don't make for good copy and (inexplicably) they all
seem to think they are in some important race (as opposed to
collectively and systematically killing off browser scripting with
endless floundering).

 "When we look at the major browsers that are available, we
  need to figure out what platforms they’re running on and
  what versions of those browsers work well-enough to support."

So there it is. Having stated that there are a zillion mobile browsers
and platforms, the solution is to list a few browsers on a few
platforms and the rest be damned.

All the better to make their untenable goals appear tenable.

Someone in the development team (a
"we") has decided specifically which browsers they will and won't
support.

Yes, how "pragmatic" of them. Go with a ludicrous design and block
out all of the browsers that won't cooperate. Apparently starting
with a reasonable design never seems to occur to them.

A decision apparently based on some unstated "quality" and
"relevance in the larger mobile market" criteria.

Yes, somebody wrote me recently to comment about a chapter in one of
Resig's book about "choosing your core browsers". Apparently a 1%
market share is "infinitesimally small". I know something else that
is infinitesimally small...

I suppose it never occurs to these guys that you can't choose browsers
on the Web. People are going to use what they want (or must) use.
The end-users don't read the color charts. They don't know what YUI
or jQuery are and they don't (and shouldn't have to) care.

In other words, some
unnamed person or person's opinion of what will and wont be supported.

Yes and they are usually in the deluded neophyte category. Hard to
see how that fails.

Interesting that no "native" browser on any version of Windows Mobile
is to receive A grade support.

I guess they need to study harder.

Depending on whose statistics you
believe, Windows Mobile may have up to 20% market share in northern
USA, but that isn't enough for A grade support.

I heard somebody babbling about an 80/20 rule recently. Maybe if it
was 21%...

Isn't deciding on which browsers will and won't be supported the
antithesis of the ethos of the web?
Yes.

Isn't that precisely the attitude
that the web, and browsers in general, are designed to combat?
Yes.

Do we
have to keep going back over the same old arguments time and time
again?

Apparently. For some it's 1998 forever.

Anyway, it will be wildly successful, web developers will clutter
their sites with crappy animations and overweight scripts and images
that clog the meagre bandwidth and processing power available to
mobile users.

Success is relative and their competition is stiff (the aforementioned
mobile apps).

And they'll gleefully tell their clients that they
absolutely must have both a desktop and mobile version of their site,
and that it's impossible to have one site for everyone since the
browser support matrix for mobile is completely different to the one
for desktop.

Yes, like broken records, they've been spinning that same crap for
years.

Where do other devices fit into this picture like tablets, TVs and
consoles? Oh, that's right, they don't. So next there'll be jQueryTV
and jQueryConsole and so on - heck, more bespoke sites.

I predict that jQuery (and the rest) will become less and less
relevant in the years to come. Proprietary apps will supplant Web
applications. The writing has been on the wall for some time.

And browser developers won't be game to release a browser that "isn't
compatible with jQuery (mobile/TV/Console...)". I have seen the
future. Again.

Rest assured, this sorry chapter in online history (i.e. the Ajax-Web-
Application craze) will soon be coming to an inglorious and
predictable end. HTML5 may help (if they ever finish it and if
developers will stop trying to make all input elements look exactly
the same on every platform), but it will likely be too little too
late.

Might not be a bad time to learn Objective-C. I've got dibs on the
name cQuery!

 

[S.T.]

And if one has only one column of
main content, then, for heaven's sake, float it into the middle of the
viewport!

I tend to agree with you on the centering issue.