Re: Migrating scripting additions to Mac OS X
Re: Migrating scripting additions to Mac OS X
- Subject: Re: Migrating scripting additions to Mac OS X
- From: Jolly Roger <email@hidden>
- Date: Tue, 26 Jun 2001 14:20:00 -0500
On 6/26/2001 10:14 AM, "Paul Berkowitz" <email@hidden> wrote:
>
On 6/26/01 5:29 AM, "JollyRoger" <email@hidden> wrote:
>
>
> If you have something other than speculation on speed differences, then
>
> please do share. I'd like to see proof that there is a more than negligible
>
> speed hit. Otherwise, please stop throwing speculation into this.
>
>
As you know, Bill Fancher made his two versions of his ticks osax: as a
>
scripting addition and as an FBA. I believe that he wrote you twice inviting
>
you to test them,
Why are we discussing this on the list after I asked that we end the thread?
Hmmm...
Are you Bill Fancher's publicity manager or something? Should I send all
correspondence to you directly from now on? ;)
Bill emailed me his sources once as I recall, at the end of a long,
drawn-out private discussion on the topic. To tell you the truth, I'm
growing tired of debating the issue with Bill, and I haven't the time it
takes to debate with him anyway, as I have already communicated to him. Did
Bill fail to tell you that?
>
and also provided his source code, in case you thought
>
there was something he could have done better. Have you tested them them
>
yourself?
He provided it in a form I could not test booted up in OS 9.1. When I got
around to booting up in Mac OS X last night, the example script would not
open (some strange Mac OS X error message appeared instead). I'll get
around to writing my own test script eventually.
In the meantime, I would ask that both he and you be patient. :)
>
Do you think Bill could have done the FBA version better? He did
>
ask you for suggestions. Bill had no preconceptions at all beforehand, and
>
developed his two versions as an honest means to compare performance. This
>
is something in which most of us have an interest.
My statements that speed differences are negligible are based on performance
on Mac OS 9.1.
>
I tested them. I've run them 20 times, 1000 repeats each time. The osax is
>
35 times faster on my machine in OS 10.0.4.
You mean performance took a nose-dive in Mac OS X?! CRAP! Someone better
get on the phone with Steve Jobs FAST! I bet the Mac OS X dev team isn't
aware of this performance issue at all!!! ;)
If you are wanting to test performance of *any* Mac OS X process, I think
you'd do better to wait until the OS is finished baking. Mac OS X is far
from finished in terms of performance. Something that may be slow now can
change drastically to something speedy later on, and vice versa. Until
things have settled down, I'd venture to say that any performance tests you
make on Mac OS X are pointless. Things can, are, and will be changing for a
while.
Have you tested them on Mac OS 9? I'd be interested in those results. Oh,
that's right - they won't run on OS 9. Guess not. I'd write my own that
would run on Mac OS 9, but I'm strapped for time as it is right now. It'll
have to wait until I can get around to it, I guess. :/
>
I do not consider a 3400% speed
>
differential, or even Bill's more conservative claim of 2900%, "negligible"
>
or "speculation". I consider it something of quite notable importance in
>
this discussion.
Me too, but as I said, why even try to measure performance on an OS that is
constantly changing and is not optimised yet?
>
The namespace argument is also important, although it's a
>
problem I myself have run into only once, even though I probably have about
>
500 osax commands installed in my OS 9.1 system. I would hope that keeping
>
communication lines open among osax developers in open forums and private
>
discussions could resolve any namespace conflicts in the future.
That hasn't worked so far. I see conflicts happen all the time on this
list. Not sure what makes you think things will change for the better in
the future...
>
I know which version of Bill's tick osax I'll be using. I'll also be happy
>
to compare both versions of any osax/FBA anyone develops. These are early
>
days yet.
Yes they are. We will see where we end up in the end.
Again, can we please end this thread, Paul?