Prev: ECMA-262-5 in detail. Chapter 2. Strict Mode.
Next: Simple hack to get $500 to your home.
From: David Mark on 2 Jun 2010 06:49
On Jun 2, 6:25 am, David Mark <dmark.cins...(a)gmail.com> wrote:
> On Jun 2, 12:36 am, Joe Nine <j...(a)yahoo.com> wrote:
>
> > David Mark wrote:
> > > The presence of QSA and non-QSA enabled browsers, as well as broken
>
> > What's this QSA I see mentioned occasionally here? Google was no helphttp://www.google.com/search?q=qsa
>
> Basically, a little while back, browsers developed the ability to do
> queries, without the help of dubious scripts.  Unfortunately, they are
> not compatible with the popular scripts.  Somehow it was decided that
> the best way for such scripts to proceed was to call QSA first and on
> failures use the old incompatible code.  The spin was that now the old
> scripts were now faster than ever (e.g. "Sizzle" was born).  The
> reality is evidenced here:-
>
> http://www.cinsoft.net/slickspeed.html
>
> ...which clearly demonstrates the differences between the non-QSA and
> QSA versions of several popular scripts.  This caused quite the public
> outcry when I published it.  Though much of it was directed at me
> rather than the scripts.  :)  After all, why would I test the old
> versions?  Because the new ones use QSA for 90% of the tested queries
> and AFAIK offer no way to turn it off.  And as we have rehashed with
> this quiz, testing in older browsers is quite revealing as well.  Why
> does this matter?  Because it demonstrates the effects of time on
> sniffy multi-browser scripts.  In a year, today's browsers will be on
> the scrap heap too and that's just silly.

Forgot to mention. Now roughly 98% of Sizzle's code exists to
"support" non-QSA browsers. Quotes indicate that could only be IE < 8
(and its Compatibility View), which they seem bound and determined to
stop "supporting" (and as we've seen that's a pretty strong word for
its performance in those browsers). So, if all of the supported
browsers have QSA then all you need are the roughly six lines used to
call QSA. Wouldn't it make a lot more sense to detect QSA, use it
when possible and degrade for everything else? Why use something that
is known to have tons of issues in some of the worst browsers ever
produced (e.g. IE7). Doesn't make a lot of sense, does it? Why are
all of the "major" projects dropping in 70K of outmoded (not to
mention irretrievably broken) code at this point to "support" browsers
they are all just itching to declare dead? Perhaps they haven't
thought this thing through. But then, they didn't need to as I've
explained and demonstrated these concepts ad nauseam. :)

The answer should be quite clear. If the scripts are reduced to six
lines, the associated cottage industries are out of business.
Granted, jQuery as a whole has more than just Sizzle. It's got some
element measurement methods that don't work and some very primitive
special effects (which also have issues in older versions of IE,
quirks mode, etc.) And don't forget "Live", which is a terrible way
to do basic event delegation (without learning anything of course),
relying on (you guessed it) queries. Hard to see those features as
beneficial for beginners.
From: Thomas 'PointedEars' Lahn on 2 Jun 2010 07:11
John G Harris wrote:

> Garrett Smith wrote:
>> Joe Nine wrote:
>>> David Mark wrote:
>>>> The presence of QSA and non-QSA enabled browsers, as well as broken
>>> What's this QSA I see mentioned occasionally here? Google was no help
>>> http://www.google.com/search?q=qsa
>> QSA - document.querySelectorAll. Handy lingo when you need to say
>> something in < 140 characters.
>
> According to Wikipedia it stands for Qt Script for Applications, which
> is a recently superseded ECMAScript engine.

"qSA", then?

> It's better if the less common abbreviations are expanded at least once
> in the threads using them.

ACK


PointedEars
--
Prototype.js was written by people who don't know javascript for people
who don't know javascript. People who don't know javascript are not
the best source of advice on designing systems that use javascript.
-- Richard Cornford, cljs, <f806at$ail$1$8300dec7(a)news.demon.co.uk>
From: news on 2 Jun 2010 07:14
On Jun 2, 2:53 am, RobG <rg...(a)iinet.net.au> wrote:
> On Jun 2, 11:14 am, Garrett Smith <dhtmlkitc...(a)gmail.com> wrote:
>
> > On 6/1/2010 4:28 PM, RobG wrote:
> [...]
> > > If the intention is to point out inconsistencies in the Sizzle
> > > selector engine, you're at risk of failing because of simple errors
> > > that affect the credibility of the exercise.
>
> > Indubitably stupefied by my own quiz and point taken.
>
> I wonder whether it might be better to just state the expected
> (technically correct) result, then the actual (jQuery) result and why
> jQuery fails on the first attempt regardless of which answers are
> given. Requiring visitors to "pass" the test to get to the interesting
> part may distract from the valuable point being made.
>

I agree - out of sheer nerdiness, I kept trying till I got them all
"right", but to be honest it was an effort.

Slightly hampered by your referring to the answers as a, b, c, d etc,
but on my browser at least, them being labelled 1, 2, 3, 4 etc.

You'd make more of a point just scoring people's answers once - show
the explanation of the correct answer regardless whether they got it
right or not. The point you are making is clear ONLY when you can see
the correct answers. The people you are trying to make that point to
won't get the answers right (in fact I don't think anyone can without
the research you did), so will wander away just as confused.

From: Matt Kruse on 2 Jun 2010 08:42
On Jun 1, 8:53 pm, RobG <rg...(a)iinet.net.au> wrote:
> I wonder whether it might be better to just state the expected
> (technically correct) result, then the actual (jQuery) result and why
> jQuery fails on the first attempt regardless of which answers are
> given. Requiring visitors to "pass" the test to get to the interesting
> part may distract from the valuable point being made.

Agreed. Obviously the point of this is to highlight the failings of
jQuery, but doing it in such a "sarcastic" way is annoying. I'd like
to see a page that just lists what jQuery gets wrong, rather than
making me study a test and pick answers. I don't have time for that,
and I don't think most people who might actually benefit from the
knowledge will take the time to actually do it.

Matt Kruse
From: Gregor Kofler on 2 Jun 2010 08:55
Am 2010-06-02 00:05, nick meinte:

> Got some issues with the scoring...
>
> "Good job! After just 1 try, you got all 10 questions right!"
>
> ..but I only got two right (7 and 9). Now I'm really curious what the
> right answers were? I'm guessing the non-strings are allowed, and the
> image inputs are getting treated as img tags or vice versa?

I was congratulated with the same message and had one out of ten right.
I'm somewhat puzzled what this quiz is all about...

Gregor


--
http://www.gregorkofler.com
First  |  Prev  |  Next  |  Last
Pages: 1 2 3 4 5
Prev: ECMA-262-5 in detail. Chapter 2. Strict Mode.
Next: Simple hack to get $500 to your home.