FYI.

I think this is a challenging problem, and I started to write about the
pros/cons/permutations, but then figured if Stefan/Anter's want help/input,
they can ask/discuss.

Since right now the behaviour seems non-deterministic (i.e. does what is
default for the environment) anything that can be done to tighten it should
be good. Luckily we'll "feel" the benefits first with Gump.

BTW: I wonder if build.sysclasspath ought influence that classpath aspects
of this.

regards,

Adam
----- Original Message ----- 
From: <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Monday, December 08, 2003 8:52 AM
Subject: DO NOT REPLY [Bug 25327] New: - Forked VMs != forking VM, breaking
script portability


> DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG
> RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
> <http://nagoya.apache.org/bugzilla/show_bug.cgi?id=25327>.
> ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND
> INSERTED IN THE BUG DATABASE.
>
> http://nagoya.apache.org/bugzilla/show_bug.cgi?id=25327
>
> Forked VMs != forking VM, breaking script portability
>
>            Summary: Forked VMs != forking VM, breaking script portability
>            Product: Ant
>            Version: 1.6Beta
>           Platform: All
>                URL: http://gump.covalent.com/log/ws-axis-test.html
>         OS/Version: All
>             Status: NEW
>           Severity: Enhancement
>           Priority: Other
>          Component: Core
>         AssignedTo: [EMAIL PROTECTED]
>         ReportedBy: [EMAIL PROTECTED]
>                 CC: [EMAIL PROTECTED]
>
>
> If a user (e.g. Gump, or whatever) selects a Java VM (perhaps a certain
> instance), or sets system properties (e.g. java.awt.headless, or
> bootclasspath ), forked VMs don't match this. As such, scripts can fail
based
> upon the default environment (which gives the new forked Ant VM).
>
> I can see this in three parts:
>
> 1) JVM instance (this might just have to be left to the OS path)
> 2) Classpaths (ought bootclasspath be expressed here, or ought we use
endorsded)
> 3) System properties (default behaviour ought be to clone all).
>
> I believe that this is a JVM problem (it ought allow a 'fork'), and I
wonder if
> others have solved it & ant could borrow that.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>


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

Reply via email to