more bug explanation. (RE: Chipp & wouter)
frank at backtalk.com
Thu Jun 24 04:36:15 CDT 2004
I've been following this thread with interest, and feel compelled to
chime in again. I haven't looked at William's stacks as I'm on
deadline, but a couple of things need to be said.
First, menus work in RunRev -- popup menus work, pull down menus work,
option menus work, combo menus work, the menubar works. I've got 8
stacks, 8 menubars, 42 menus, and 6 popup menus in my app, and have had
no crashes that I haven't been able to account for in something I was
Second, is it possible to crash RunRev? Yes. But it's also possible
to write code that doesn't crash RunRev. To write clean code one must
test and debug it **as it's being written**, and learn how to **back
out code** in order to track down any crashes. The fact that it takes
"a minute delay to gather all that data and fill the menu" suggests to
me that something is seriously wrong with the code base, and it
probably needs to be re-architected and then re-written.
So, if I were William, I would start over again with a clean stack, and
see if I couldn't get the UI running -- and only after the UI was fully
debugged and working would I add in the animation external. And if I
were Kevin Miller, I wouldn't have any of my engineers spending time on
this until William can produce a self-contained example of the crashes
-- Frank Leahy
On Jun 24, 2004, at 1:12 AM, Brian Yennie wrote:
>> second type of popup, just a standard list of options to do anything,
>> be it save files and such, or switch group visibility. These corrupt
>> without known cause. They corrupt even when they only have 4 or 5
>> This occurs even when the only handler is a mouseUp.
> Are you able to narrow the stack down to one popup with just a mouseUp
> and a few items then?
> With all of the buttons in your test stack which you say don't
> actually work like in your stack, it's rather confusing to figure out
> exactly what to look at or what you really want to do. I know you've
> given detailed descriptions, but I think what we need is to work on
> boiling it down to the simplest case of the problem.
> - Brian
More information about the use-livecode