Thread: [tcljava-user] Jacl: tcl.lang.TclRuntimeError
Brought to you by:
mdejong
From: Jeff S. <js...@on...> - 2001-06-18 21:53:11
|
(Apologies in advance if this is a known problem... I've been away from the list for a while now.) Every now and then, our application logs an error like: Jun 14 20:48:40 taubmanweb1 ERR: tcl.lang.TclRuntimeError: (find) table entry "java.lang.String.17035487" mapped to an invalid entry, Searched (Class = "java.lang.String" identityHashCode = "17035487" hashCode = "1828610235") Found (refID = "java0xd099" Class = "java.lang.String" identityHashCode = "17035487" hashCode = "1828610294") Equality Tests ( Class == "true" Object == "false" Object.equals() == "false" Interp == "true") Apparently, the JDK can duplicate an identityHashCode. (I'm using Sun's JDK 1.2.2 on Solaris.) Jeff |
From: Mo D. <md...@cy...> - 2001-06-18 23:12:46
|
On Mon, 18 Jun 2001, Jeff Sturm wrote: > (Apologies in advance if this is a known problem... I've been away from > the list for a while now.) > > Every now and then, our application logs an error like: > > Jun 14 20:48:40 taubmanweb1 ERR: tcl.lang.TclRuntimeError: (find) table > entry "java.lang.String.17035487" mapped to an invalid entry, Searched (Class = > "java.lang.String" identityHashCode = "17035487" hashCode = "1828610235") > Found (refID = "java0xd099" Class = "java.lang.String" identityHashCode = > "17035487" hashCode = "1828610294") Equality Tests ( Class == "true" > Object == "false" Object.equals() == "false" Interp == "true") > > Apparently, the JDK can duplicate an identityHashCode. (I'm using Sun's > JDK 1.2.2 on Solaris.) > > Jeff That is a really old and nasty problem that has been fixed in the 1.3 version in the CVS. You should use the CVS version if you want to avoid this problem. You could also try back porting the rewrite of ReflectObject, but it would be no fun at all. Mo |
From: Dan S. <sc...@is...> - 2001-06-19 01:52:32
|
Mo, What are the plans for releasing 1.3 into the mainstream? Dan ----- Original Message ----- From: "Mo DeJong" <md...@cy...> To: <tcl...@li...> Sent: Monday, June 18, 2001 6:12 PM Subject: Re: [tcljava-user] Jacl: tcl.lang.TclRuntimeError > On Mon, 18 Jun 2001, Jeff Sturm wrote: > > > (Apologies in advance if this is a known problem... I've been away from > > the list for a while now.) > > > > Every now and then, our application logs an error like: > > > > Jun 14 20:48:40 taubmanweb1 ERR: tcl.lang.TclRuntimeError: (find) table > > entry "java.lang.String.17035487" mapped to an invalid entry, Searched (Class = > > "java.lang.String" identityHashCode = "17035487" hashCode = "1828610235") > > Found (refID = "java0xd099" Class = "java.lang.String" identityHashCode = > > "17035487" hashCode = "1828610294") Equality Tests ( Class == "true" > > Object == "false" Object.equals() == "false" Interp == "true") > > > > Apparently, the JDK can duplicate an identityHashCode. (I'm using Sun's > > JDK 1.2.2 on Solaris.) > > > > Jeff > > That is a really old and nasty problem that has been fixed in > the 1.3 version in the CVS. You should use the CVS version > if you want to avoid this problem. You could also try back > porting the rewrite of ReflectObject, but it would be no > fun at all. > > Mo > > _______________________________________________ > tcljava-user mailing list > tcl...@li... > http://lists.sourceforge.net/lists/listinfo/tcljava-user > |
From: Mo D. <md...@cy...> - 2001-06-19 03:43:51
|
On Mon, 18 Jun 2001, Dan Schenck wrote: > Mo, > > What are the plans for releasing 1.3 into the mainstream? > > Dan I had not really made any recently. There was a lot of new development going on a couple of months back, but recently things have slowed down. I would like to get a few more features into Jacl before calling 1.3 "stable". For example, the IO layer work needs to be merged. Tcl Blend is mostly stable, there is some work that needs to be done to get the thread shutdown code finished. There are also quite a few test cases related to the new thread code in Tcl Blend that need to be written. In addition, an HTML based into to programming with the Tcl/Java API needs to be written. The existing demos are ok, but folks need a more simple walk-through of the API. If all/most of those things are in place, I think 1.3 would be ready to roll. Perhaps you would be interested in the release manager role? It would involve drawing up a release schedule and then getting folks to do the needed tasks (me included). There are some nice tools on Source Forge to help with this process, if you are interested. Mo |
From: Jeff S. <js...@on...> - 2001-06-19 03:58:36
|
On Mon, 18 Jun 2001, Mo DeJong wrote: > > Apparently, the JDK can duplicate an identityHashCode. (I'm using Sun's > > JDK 1.2.2 on Solaris.) > > That is a really old and nasty problem that has been fixed in > the 1.3 version in the CVS. You should use the CVS version > if you want to avoid this problem. You could also try back > porting the rewrite of ReflectObject, but it would be no > fun at all. Thanks Mo. We're actually using sources imported from CVS a year or so ago, extensively modified in our tree [*]. I'll look into backporting the changes. Jeff [*] We initially tried to use Jacl as-is. It became difficult to extend the package because most of the classes are package-private and some are not well abstracted. For example, we needed to create a StreamChannel to wrap an arbitrary Java io stream as a Tcl channel. |