By "falls down a bit" do you mean unsuited to the task?
Possibly. It depends of course on the task. But it takes almost 4
seconds to run 40 queries on a 114KB HTML page. Three of the
libraries only take around 1 second for the same task. So whether
it's good enough for your task is up to you, but you should notice
that the other libraries are significantly faster here.
"self-aggrandizing"? Why do you care about any perceived personality
flaw in someone you most likely will never meet/know? Can't you just
objectively comment on the quality of the work?
I don't give a damn about his personality. But both the style and the
content of his posts get to be noxious. In the few months I've been
reading he makes fairly outrageous claims for his own skill and
regularly castigate the other libraries, claiming that anything the
other libraries get right, they steal from him.
When I go to test the claims he has made, and post my results, I'm met
with insults.
I didn't get the feeling that he didn't expect anyone to notice. Quite
the contrary, that everything was openly laid out before us.
But when challenged on the speed advantages he announced, he didn't
post a link to the speed test, nor did he post his results. He merely
makes these very strong, but non-specific claims.
At any rate, haven't the other libraries been around for a "long" time?
So shouldn't even a two year old version be an accurate representation
of the quality of the coding involved?
But it's not the quality of the coding that's under challenge here.
There should definitely be investigations here into that quality, but
what I was presenting was hard numbers that challenged his assertions
about the speed of his library compared to the competitors. Anyone
using such numbers as a factor in deciding which library to use would
expect the comparison to be among reasonably recent, if not the
latest, versions.
Perhaps you *are* testing the wrong thing. Sorry if I missed it but I
didn't notice your defence of your claim.
I'm not sure at this point "testing the wrong thing" has any
significant meaning. I posted the results of my tests of his library
in recent versions of the major browsers on a developer's Window's
machine. In what way could that be the wrong thing to test?
Additional tests on older machines, or other browsers are equally
legitimate. But people looking to use one library or another should
know how they perform in the environments in which they expect the
libraries to run. Perhaps in some locked-down corporate environment,
only IE 6-8 matters. If their application is to run only on iPhones,
presumably testing on those devices is what the user will care about.
For instance, I certainly don't give a damn personally about how any
of the libraries perform in FF1. Maybe David has a good use case for
that. But to try to claim that my tests are useless because they were
performed on a relatively modern machine is ridiculous and self-
serving.
To repeat, the people that I know that use the "common" js libraries are
unhappy with all of them.
I think it's great that David is bringing another library into the
fray. Certainly the survivors of the last rounds of the competition
have a great number of flaws, and the competition should help improve
all of them. But if he starts by making over-exaggerated claims about
his library, he is doing everyone (himself included!) a great
disservice.
-- Scott