I was thinking that GC in an IDE is a little different that usual. If
the user pauses to think, this might be a good time to do a quick GC,
even if memory is not that full.
You are best to do short GCs so that when the user suddenly kicks back
in, you get quick response. The total time spent going GC is not very
important.
I wondered how well Java.exe is designed for an IDE to work that way.

Signature
Canadian Mind Products, Roedy Green.
http://mindprod.com Java custom programming, consulting and coaching.
>I was thinking that GC in an IDE is a little different that usual. If
> the user pauses to think, this might be a good time to do a quick GC,
> even if memory is not that full.
I think this profile of user-activity is typically not only for IDEs, but
for most (if not all) GUI-centric applications. Most of the time the user
pauses to think or to read something currently displayed in the GUI.
Therefore I think a GC-strategy optimized for GUI-clients will be optimal
for IDEs as well.
> You are best to do short GCs so that when the user suddenly kicks back
> in, you get quick response. The total time spent going GC is not very
> important.
>
> I wondered how well Java.exe is designed for an IDE to work that way.

Signature
"TFritsch$t-online:de".replace(':','.').replace('$','@')
Thomas Hawtin - 05 Mar 2006 23:56 GMT
>> I was thinking that GC in an IDE is a little different that usual. If
>> the user pauses to think, this might be a good time to do a quick GC,
>> even if memory is not that full.
> I think this profile of user-activity is typically not only for IDEs, but
> for most (if not all) GUI-centric applications. Most of the time the user
> pauses to think or to read something currently displayed in the GUI.
> Therefore I think a GC-strategy optimized for GUI-clients will be optimal
> for IDEs as well.
A strategy like having GC run mostly concurrently in a low priority thread?
Tom Hawtin

Signature
Unemployed English Java programmer
http://jroller.com/page/tackline/