Stacks whose names start with "rev"

[Mark Wieder]

On 09/28/2018 01:40 PM, Richard Gaskin via use-livecode wrote:
> I hate to say this is one of those rare times when I disagree with you, 
> but this is one of those rare times I disagree with you.

LOL

> Naming things has implications in many contexts across most languages. 
> Even well-written SQL will need to be revised if it refers to a field or 
> table whose name has changed.

Sure. But you wouldn't want to have a field name dependent on the 
contents of a record in another field. And that's my point. Or at least 
the one I was trying to make.

> And most importantly, this particular instance has less to do with 
> generic presumed "best practice" than what appears to be just a bug in 
> the IDE:  when it encounters stacks that it thinks are its own, even 
> using the prescribed method for editing the stacks doesn't work.
> 
> If setting gRevDevelopment did what it's supposed to do, she'd be able 
> to get back to work with no more inconvenience that seeing IDE object 
> names in its UI listings, and we wouldn't be having this discussion.
> 

Not so much of a disagreement, I think, as looking at different facets 
(Ha! I almost wrote "faucets"... we're about to get serious plumbing 
done here).

I think this does point out at least one serious bug in the IDE with 
respect to system/non-system stacks. There's a function somewhere in the 
IDE (I forget the name or location) that determines whether a stack is 
"special"... some of the stack names are hardwired, others are 
determined by naming convention.

Of course, part of the IDE problem comes from dealing with stacks by 
short name only, and this has been the subject of a long-standing bug 
report. The engine itself has no problem with multiple stacks, just the 
IDE. And this could possibly be eliminated with stack UUIDs or if the 
IDE looked at filenames in addition to stack names.

-- 
  Mark Wieder
  ahsoftware at gmail.com