dimiter wrote:

   > At least for me these generated methods don't add any value.


Ever written this code for 5 15+ field classes, that map objects to DB 
data? Pretty boring, and error prone.


   > If I want equals I prefer to write it myself and to determine precisely
   > which variables define the state of my class, 


So do I; but I start from a complete and functionning skeleton, that I 
tune if necessary.
Less error prone. Helps me be consistent, across classes.
The "template" side is an extra advantage for team consistency.


   > if I write toString - I want
   > to see concise description and not full memory dump (that's what the
   > debugger is for).


So do I; but ...


   > But if you feel more comfortable with standard implementations, 


I use Joshua Bloch - "Effective Java" - /standard/.
 
http://www.amazon.co.uk/exec/obidos/ASIN/0201310058/qid=1023090468/sr=2-1/ref=sr_2_3_1/026-2098539-4510843
A good enough standard to start with.


    > you could

   > use the generic versions and rewrite them yourself if the profiling shows
   > them as bottlenecks (I've seen this very rarely).


Having this task automated helps a lot.
No more "do I really need to spend 5-20 minutes making this object 
canonical, when all I need is a basic toString().


Having this task integrated would help even more : modifiying the class 
structure would trigger change/notification. It's not top priority, but 
like many IDEA cool "little" features, they quickly become invaluable.

    > so from me it gets -1


Does "-1" mean that
    - you would write a plug-in to remove it from the standard menus,
      to save space,

or simply that you think that
    - you would never use it
?

Alain

------------------------------------

 >>Martin Bayly wrote:
 >>>I find it tedious to write equals, hashcode and toString methods for
 >>>classes.  It would be great if IDEA could do this for me by examing
 >>>the fields of a class and generating/maintaining a default
 >>>implementation of these methods.


> Alain Ravet wrote:
>>+5
>>This is an old request - and battle - : make objects canonical.
>>In the meantime, I solved this with a doclet. Not perfect, but good
> enough.
>>Integration would be better, though.


_______________________________________________
Eap-features mailing list
[EMAIL PROTECTED]
http://lists.jetbrains.com/mailman/listinfo/eap-features

Reply via email to