Re: Bugs bugs, Finder bugs, yummy bugs
Re: Bugs bugs, Finder bugs, yummy bugs
- Subject: Re: Bugs bugs, Finder bugs, yummy bugs
- From: Matt Deatherage <email@hidden>
- Date: Tue, 23 Aug 2005 14:14:46 -0500
I'm honestly a bit puzzled at to why this produces so much emotion.
Richard is right - when you log in at <connect.apple.com>, the
members-only page doesn't have any links to bug reporting. There's one
on the front page of <http://developer.apple.com> that leads to
<http://developer.apple.com/bugreporter/>, with instructions and an
FAQ, but I agree it should have a link inside the members-only site.
So I'm going to file a bug on it. :-)
(Bugs about the ADC site don't go into RADAR, for whatever reason, so
I'm submitting it through <http://developer.apple.com/contact/>. That
page also has the top 5 most frequently asked questions, and #5 is "How
can I submit an enhancement request/bug report?")
--
Look, file bugs, don't file bugs, dance, eat oysters, do whatever makes
you happy. Spend your time arguing anything you like, if it fulfills
you, or feed the hungry, or do something nice for your kids, or write a
book convincing the world of something it wouldn't believe otherwise.
Life is short. Make the most of it.
All I'm saying is that if you take the 5 minutes to file a bug at
<http://bugreporter.apple.com/>, you increase the chances of that bug
getting fixed by about 1000X.
It used to be a real PITA (let's hope that doesn't start another
acronym war) to file a bug online, but it really did take me about 5
minutes last night to do this one. You choose product/component,
severity/classification, and reproducibility from pop-up menus. You
type in the version number of the component you're reporting, provide a
one-line bug title, and then type in the steps to reproduce it.
A handy pop-up window describes the standard six-part format: A
summary of the issue, what you did to cause the bug, what you expected
to happen, what actually happened, circumstances where it does not
occur (if any), and any additional notes. You then attach a profile
from System Profiler, upload any files associated with it, and click
"Submit."
Here's the bug I filed last night, with the private information
redacted so you can see how easy it was:
> Summary:
> Using the AppleScript "call xmlrpc" statement on a private
> server works most of the time, but always fails with a "transport
> error" for specific parameter values, even though the server is
> returning the entire reponse as expected. AppleScript returns
> "transport error", -916 (an old PPCToolbox error!), but a network spy
> tool shows the server sending the entire XML response as expected.
> AppleScript or WebServicesCore just can't see it.
>
> Steps to Reproduce:
> Enter and execute the following AppleScript code (line breaks like ¬
> are added just so wrapping doesn't impede reporting, and are not
> necessary for the problem):
>
> local myRPC
> set myStruct to -- a redacted struct for a private XML call
> with timeout of 60 seconds
> tell application "http://URL.redacted";
> set myRPC to (call xmlrpc {method name:"PrivateService.populate", ¬
> parameters:myStruct})
> end tell
> end timeout
>
> I'll attach a text-only script to the bug, too.
>
> Expected Results:
> I expect the script to return an XML structure that AppleScript
> parses into a result record. For *most* values of [myStruct], this
> is exactly what happens.
>
> Actual Results:
> For *some* values of [a variable in myStruct], like 24757 (and
> others), I get a PPCToolbox error and no result.
>
> Regression:
> I first noticed this under Mac OS X 10.3.7 but didn't know if it was
> a bug or not. Switching to the "debug" version of WebServicesCore at
> that time seemed to fix it, but that's not the case with Tiger.
>
> Notes: I can try to compile a list of more [variables] where it fails.
> The URL in this bug, by the way, is a *private* URL
> and should not be disseminated beyond what's necessary to fix this
> bug. Thanks!
There's not a lot different here than in a good mailing list post
talking about a bug, for that matter - you could file a bug and then
pretty much copy and paste the text here to initiate discussion.
You may wish that Apple paid people to read mailing lists and file bugs
they find therein, but Apple does not pay people to do this. You may
wish that plans for future engineering work came from mailing lists as
well as from RADAR, but they don't - the engineers fix the bugs that
are in RADAR, and any others that get fixed are pretty much accidental.
Unfixed RADAR bugs can stop a release. Posts on a mailing list cannot.
You may wish for a different world, but this is the one we have.
I've never seen an Apple engineer, during or since my time at Apple,
say "Don't file a bug on that." They ask for it all the time, because
RADAR sets priorities. <http://bugreporter.apple.com/> is your direct
line into that process.
Use it or don't use it, that's up to you. But if you believe the old
saying that "Those who don't vote don't have the right to complain
about the outcome," then those who don't file bugs in RADAR don't get
to complain when they don't get fixed.
If all you want is discussion and workarounds, hey, that's your right
as long as the listmom says it's fine. I think it's perfectly fine.
Discussion helps. If you want a bug *fixed*, you know what to do.
All the proof you need:
> On 8/22/05 at 6:29 PM, Dave Lyons <email@hidden> wrote:
>
> > We didn't mean to ship the "find" verb. Surprisingly, nobody
> > inside or outside of Apple filed a Radar bug about it until 11 days
> > ago...and that was me. (I see from the list archives that this
> > came up back in April, but until last Wednesday I wasn't a
> > subscriber.)
Martin's post that he mentioned earlier would *not* have solved this,
because he posted on 2005.04.30, and Tiger went GM around 2005.04.06.
There were, however, Finder revisions in 10.4.2, and it might have been
fixed there had it been in RADAR before 2005.08.11.
It's an imperfect system that could use any number of easily-conceived
improvements. I do not defend RADAR as perfection in defect tracking.
I merely note that it is how Apple works at present, and it's how you
can get your bugs fixed with a minimum of fuss. If you choose not to
use it, I support that, but don't expect any sympathy or credibility
from those who know you had the chance and passed on it.
Unless directly queried or personally attacked, I intend to say no more
on the subject.
--Matt
--
Matt Deatherage <email@hidden>
GCSF, Incorporated <http://www.macjournals.com>
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Applescript-users mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden