You might want to send Sun a bug report about this one. Sun seems to like
getting JIT bug reports, but they just ignore you the rest of the time.
> Thanks for the tip about the JIT. I made this change and now the tests run.
I will have to try the Sun JDK sometime in the future. I gave up on Sun,
and switched to the IBM JDK because it is a lot faster and seems to have
fewer bugs than the Sun JDK.
> I get the following errors from the test run using Sun's JDK 1.1.8
> production release:
>
...
> tcljava/JavaInfoCmd.test
These test failures are really odd. They fail because there is an extra
',' char in the expected output. I do not remember making a change in
the java::info command, but I will have to double check to make sure
did not mess things up.
>
> ==== javaInfoCmd-6.3 fields
>
> ==== Contents of test case:
>
> list [catch {java::info fields junk1 junk2} msg] $msg
>
> ==== Result was:
>
> 1 {bad option "junk1": must be -type or -static}
>
> ---- Result should have been:
>
> 1 {bad option "junk1": must be -type, or -static}
>
> ---- javaInfoCmd-6.3 FAILED
>
>
...
> jacl/Interp.test
Don't worry about this "x" vs "::x" var name issue. This is something
that needs to be fixed in both Jacl and Tcl Blend.
> ==== Interp-1.2 dispose, variables are freed
>
> ==== Contents of test case:
>
> set interp [java::new tcl.lang.Interp]
>
> $interp eval {
>
> set x 10
>
> set y 10
>
> set z 10
>
> }
>
> set tr [java::new tests.VarTraceTest {}]
>
> set list ""
>
> lappend list [lsort [[java::field $tr varsCalled] toString]]
>
> $interp traceVar x [java::null] $tr [java::field tcl.lang.TCL
> TRACE_UNSETS]
>
> $interp traceVar y [java::null] $tr [java::field tcl.lang.TCL
> TRACE_UNSETS]
>
> $interp traceVar z [java::null] $tr [java::field tcl.lang.TCL
> TRACE_UNSETS]
>
> $interp dispose
>
> lappend list [lsort [[java::field $tr varsCalled] toString]]
>
> set list
>
> ==== Result was:
>
> {} {::x(null) ::y(null) ::z(null)}
>
> ---- Result should have been:
>
> {} {x(null) y(null) z(null)}
>
> ---- Interp-1.2 FAILED
...
> tcl/clock.test
This is really strange. Are you sure you were not running this is Tcl
Blend? This looks like the clock test case results from a newer
version of the Tcl core.
> ==== clock-3.7 clock format tests
>
> ==== Contents of test case:
>
> list [catch {clock format 123 -bad arg} msg] $msg
>
> ==== Result was:
>
> 1 {bad switch "-bad": must be -format or -gmt}
>
> ---- Result should have been:
>
> 1 {bad switch "-bad": must be -format, or -gmt}
>
> ---- clock-3.7 FAILED
>
The rest of the errors you got are known problems. They should
be fixed but they no break "normal operation" so they are kind of on
the back burner.
later
mo
----------------------------------------------------------------
The TclJava mailing list is sponsored by WebNet Technologies.
To subscribe: send mail to [EMAIL PROTECTED]
with the word SUBSCRIBE as the subject.
To unsubscribe: send mail to [EMAIL PROTECTED]
with the word UNSUBSCRIBE as the subject.
To send to the list, send email to '[EMAIL PROTECTED]'.
A list archive is at: http://www.findmail.com/listsaver/tcldallas/
