dburgun at dsl.pipex.com
Fri Mar 24 12:23:18 EST 2006
On 24 Mar 2006, at 16:12, Richard Gaskin wrote:
> I'm not sure it's helpful to the readers here to say that "me" is
> not 100% reliable without providing details;
The details are all in the posts from the other thread, I really
didn't want to cut and paste all that again.
> while narrowly accurate, it appears that "me" is 99.99% reliable,
> with only one highly specific circumstance required to expose a bug
> that to my knowledge has affects very few people in actual use,
f I had read something like "me" is not 100% reliable 6 months ago,
it would have saved me two days frustration and a lot of time and
effort by a few people on this list. As it is now I have to work the
weekend to get back on Track! Oh well, such is life!
If there are two ways of doing something, one that is 100% reliable
and one is 99.99% reliable, I think most people would choose the 100%
solution, especially when it comes to software.
> in a circumstance where using option a) above still relies on "me"
> and takes only a few seconds to correct for on one line with no
> changes required to any other code.
This is not *always* the case, for instance this requires more effort:
set line 1 of me to <something>
The thing is that I have:
put <something> into me
all over the place. The reason for this is that around 2 years ago
when I first started using RunRev, I saw an initialization handler
(in one of the Sample Stacks that came with RunRev) that did:
put empty into me
So I used this form from then on and the 99.99% solution got
replicated over and over again!
It is really obscure that you can have a tried and tested handler
that stops working as soon as it's called from a handler on an
unopened card and to get around the problem have to use an alternate
syntax which works 100% of the time (AFAIK!)
Also, seasoned "RunRev'ers) or maybe MetaCard'ers seem to think that
it's odd that a handler gets called on an unopened card. I guess I
just haven't come from the same background. I didn't think twice
about whether an object on an unopened card would be treated
differently or not. As far as I know it's not explicitly mentioned in
any of the documents.
This could be part of the problem facing newbies. The seasoned
developers have a mindset that has evolved from what *was* possible,
whereas the newbie doesn't know what *was* possible so they don't
think deeply about it and concentrate more on what *is* possible
given the sample stacks available and the documentation.
Now that I have changed all the occurrences of "me" to use the longer
syntax it works just fine, and I must say that using my Inter-Stack
Messaging (ISM) System makes re-using groups and building apps so
much more simple! In fact I think I am at least 95% of the way to
achieving the dream of truly re-usable objects.
I just did an experiment where I took 5 Groups developed using the
ISM and dropped them into a new stack. I then made a copy of all the
groups (so that there are two of each) and have them all interacting
perfectly. The whole operation took about 15 minutes, that's screen
design, code, everything!
FANTASTIC!!! This is what I love about RunRev!
All the Best
More information about the Use-livecode