Garrett said:
David said:
Garrett said:
David Mark wrote:
Garrett Smith wrote:
David Mark wrote:
Garrett Smith wrote:
David Mark wrote:
David Mark wrote:
Garrett Smith wrote:
David Mark wrote:
David Mark wrote:
Garrett Smith wrote:
Eric Bednarz wrote:
[snip a bunch of noise]
You truly are hopeless. You really should be following me
("cinsoft" on
Twitter). Good luck with the FAQ.
I don't have much interest in following someone who cannot explain the
reason for his own code (but keep up the marketing and you're bound to
get some fool who will).
Once again, you are blind. Good luck with that!
No, I am not blind.
You've been acting like you can't see the forest for the trees with
numerous suggestions of mine. It's been going on for years.
Huh.
[...]
Wrong. We've had this discussion more than once. I've all but spelled
it out here. Now I have spelled it out (in code) on my site.
As with Twitter: That is your business.
My point was I made the information available after the initial
confusion. You should keep up with my site anyway as it is full of
useful information.
Although I find value in the *concept* of twitter, the UI is not worth
wrangling with.
Yes, the Web UI is a jQuery. Enough said.
I have decided that it is not worth the frustration of
trying to use Twitter UI. I am just not that interested in broadcasting
to many people.
That's your choice. I put up with it as it is great for generating
traffic, which begets more traffic, which begets contract offers.
I figure that the smart folks will find there way here
and the same folks who are promulgating the same bad practices that have
crap-flooded the industry with will flounder with their pop websites
(like Twitter).
I figured that once too. People seem to gravitate to the crap-ola. If
it has pretty graphics, it must be good!
Many of these individuals become millionaires by being involved in
trendy projects that sell. They sell despite the "engineering" with code
that seems incongruous with that term.
Yes. Take Google, which throws exceptions on virtually every page in
the _latest_ Opera (a very standards-compliant browser of course). It's
all browser sniffing madness.
What happens here on c.l.js is public discussion. If there is a good
reason for that test, such as a case where it helps find style value in
a particular situation, I haven't seen anybody posting it.
That's because (like many of my points on feature testing), nobody seems
to have thought of it (except Richard of course, who I assume is amused
that I with what I have built on top of many of his original ideas).
No, c.l.js is the right place to discuss these things. You can advertise
your twitter homepage or website or library, but those are not places
for open discussion. Here, there is free discussion.
I was just trying to point you to some example code that I posted.
If I am missing a point (a good reason) for determining
`doesPositionStyleAffectOffset` or whatever it should be called, then it
hasn't been made clear what that is.
The way to read styles is for browsers that implement ViewCSS, use
defaultView.getComputedStyle(obj, pseudoClass), and for IE, to use
currentStyle.
Not in many cases (e.g. left/top/right/bottom/height/width). First
there are lots of bugs, second IE won't return _anything_ unless there
is a corresponding declaration (and even then, it may not be in a
convenient format). That's why you have to balance the equation with
offset* properties. You don't have to take my word for it. Try out my
examples. I've tested in IE5.5-8 (all modes), Opera 7-10, FF1-3.6,
Chrome 3/4, Safari 4.0, an iPhone and several other odd agents. If you
get the equations right (which I knew I did for at least
left/top/height/width), you really can't go wrong (math is math).
Contrast that to what the "majors" do, which is often confused or even
completely backwards (e.g. jQuery's ridiculous "smoothing over" of box
model differences, which sometimes results in unusable dimensions).
Then there's that crazy MacGuyver hack, which I hate to say is actually
in My Library as well. Like I said, the simplified examples that I am
posting now (e.g. viewport, size, position, keyboard) will eventually
supplant the more complicated routines in My Library.
Both of those APIs are problematic. Some massaging is usually required,
both of the style values in the CSS and of the values returned, to get
consistent results.
I've got it down to s science (and have since around 2007).
This is an issue I've had to deal with a lot over the years and I do not
have a simple solution. You also suggested Eric to use "My Library" and
that is not going in the FAQ.
Who said it should; however, it seems idiotic to link to PrototypeJS and
not My Library.
OK, I think I see where you're coming from. I looked at each argument
separately, not as the same thing.
As you should have, but each time it was like you went off the course
for no good reason.
I agree that there is value in having a good simple answer to the
question. The problem is that the question can't be answered so simply.
What question?
Reading style values is a simple thing to want to do but a general
solution to solve for that is not going to be simple.
That's what you think.
