Hi Joe,

Some of the thinking is that, even if no other code can get at these
things, if our own code modifies them in ways that might impact its
behaviour we could run into static mutability problems--or threading
problems for that matter.  If we're modifying a static (in a non-static
context) there had better be a very good reason for it; final would help us
catch such instances.

I'd be very interested in knowing in what situations you consider final not
to be appropriate?  Obviously where statics are used for counting the
number of instances of classes, or as temporary holding places (yuck!) they
shouldn't be final; where else have you found final problematic?

Cheers,
Neil
Neil Graham
XML Parser Development
IBM Toronto Lab
Phone:  905-413-3519, T/L 969-3519
E-mail:  [EMAIL PROTECTED]




|--------+-------------------------->
|        |          Joseph          |
|        |          Kesselman/Watson|
|        |          /IBM@IBMUS      |
|        |                          |
|        |          06/26/2002 05:16|
|        |          PM              |
|        |          Please respond  |
|        |          to xerces-j-dev |
|        |                          |
|--------+-------------------------->
  
>-----------------------------------------------------------------------------------------------------------------------|
  |                                                                                    
                                   |
  |      To:     [EMAIL PROTECTED]                                           
                                   |
  |      cc:                                                                           
                                   |
  |      Subject:     Re: Xerces and static mutability                                 
                                   |
  |                                                                                    
                                   |
  |                                                                                    
                                   |
  
>-----------------------------------------------------------------------------------------------------------------------|



Just a sanity-check, Neil: There are a bunch of things which you're making
final as well as private. Some of them really are manifest constants and we
can reasonably assert that nobody should be mucking with them in
subclasses; the kidOK table is one such instance. Is that true for all of
'em, or are some becoming final for other reasons?

(Final is a useful tool, but I've seen final overused/misused in some other
projects so I tend to be just a trifle paranoid about it.)

______________________________________
Joe Kesselman / IBM Research



---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to