Protected stacks in 2.61 (was For goodness sake!)

Richard Gaskin ambassador at fourthworld.com
Mon Sep 20 03:24:14 EDT 2004


Ken Ray wrote:

>>>>It may be appropriate to bring up again the default setting of the
>>>>destroy property for new stacks. Should it be made on or remain off?
>>>
>>>
>>>Made on by default - more often than not, you want it set to true, IMHO.
>>
>>In my own work it usually makes little difference.
>>
>>What are the benefits each way?
> 
> 
> Well, for me there are four reasons for having destroystack (and
> destroyWindow too btw) turned on by default:
> 
> 1) I like knowing that when I close something it's really closed. Not a lot
> to ask, IMHO.
> 
> 2) I sometimes like to copy a substack from one stack to another (i.e. one
> product to another), and it is easiest for me to open the substack, clone
> it, close the substack and it's mainstack, open the "other" stack, and then
> set the mainstack of the clone to the new mainstack. In the process, I
> change the name back to what it originally was, and if the destroystack
> wasn't set to true in the stack I was cloning, I'll get a "purge" error
> message (which is annoying).
> 
> 3) It's a good way to manage memory (although with today's computers (as
> well as those reaching several years back)) it's not as big of an issue, but
> better to err on the side of *not* keeping everything loaded in RAM. It
> seems to me to be extremely rare to have a circumstance where you actually
> appreciate that your stack opens faster since it's already "in memory" -
> MetaCard is pretty darn fast regardless. Perhaps this was important "back in
> the day", but it doesn't seem so now.  (Side note: I've *never* had a *need*
> to have the destroyStack false, and I've been using MC for 8 years+.)
> 
> 4) It just *seems* like the right thing to do - it's kind of like having the
> default itemDelimiter as "|" and you're always having to set it back to ",".
> 
> Anyway, that's my 2 cents...

Good points.  I'll add a preference for that in the MC IDE.

Of maybe better:  how about the ability to define a stack to be used as 
a template for new stacks?  That way we can handle destroyStack, 
alwaysbuffer, rect, and anything else we want exactly as we want it.

Worth pursing a change to the default behavior of the engine?
I'd love to see the alwaysBuffer set to true as well....

-- 
  Richard Gaskin
  Fourth World Media Corporation
  Developer of WebMerge: Publish any database on any Web site
  ___________________________________________________________
  Ambassador at FourthWorld.com       http://www.FourthWorld.com


More information about the metacard mailing list