Never mind. This isn't worth the heart-ache and I'm probably wrong about workdir being defined in terms of tempdir. I made this probably wrong assumption by looking at the startup scripts .
George Sexton MH Software, Inc. http://www.mhsoftware.com/ Voice: 303 438 9585 > -----Original Message----- > From: George Sexton [mailto:[EMAIL PROTECTED] > Sent: Tuesday, February 21, 2006 10:37 AM > To: 'Tomcat Developers List' > Subject: RE: never say never... > > > > > > -----Original Message----- > > From: Mladen Turk [mailto:[EMAIL PROTECTED] > > Sent: Tuesday, February 21, 2006 4:48 AM > > To: Tomcat Developers List > > Subject: Re: never say never... > > > > > http://issues.apache.org/bugzilla/show_bug.cgi?id=38352 > > > > > > > From your bz report: > > > > > ... > > > I think to be compliant with the spec, this must be allowed. > > > ... > > > Again, I think the spec requires this. > > > ... > > > > > > So, why would anyone even spend a single microsecond to look at > > the patch if *you* could not guarantee that the patch is > > compliant with the specs? > > I guess this is my mistake, for trying to sound humble. I was > trying to be RESPECTFUL. I'll say it PLAINLY. > > Servlet Specification 2.4 says: > > SRV.3.7.1 TEMPORARY WORKING DIRECTORYS > > A temporary storage directory is required for each servlet > context. Servlet > containers must provide a private temporary directory for > each servlet context, and > make it available via the javax.servlet.context.tempdir > context attribute. > > > This just flat says that it must be there, and it is a > working directory, which implies its writable. > > > > > > > I applaud to your willing to contribute, but if you wish to > > do that, and eventually become a member of the elite, you must > > follow some rules, and the first one, like in any company is to > > respect your colleagues. > > Respect is a two way street. When someone like Bill Barker > creates a logic puzzle as blatantly wrong as he did, then > what level of respect is required? > > Specifically, Bill Barker's comments: > > >> Don't see the need. If you depend on this, your app is > non-portable since > >> there is no requirement that javax.servlet.context.tempdir > has any relation to > >>java.io.tmpdir. In fact, a servlet container is perfectly > free to set > >> java.io.tmpdir to /dev/null if it wants. > > >> > Directory specified by java.io.tmpdir (which is what > tomcat points > >> > javax.servlet.context.tempdir to) is now read, write, > delete. Again, I think > >> the > >> > spec requires this. > > I was fixing an implementation specific issue. The spec says > that the container MUST make temporary working directories available. > > The IMPLEMENTATION that tomcat uses is to set java.io.tempdir > and javax.servlet.context.tempdir > > So, my making that writable fixes an implementation specific > issue for tomcat. I'll say it again in case I wasn't clear. > > 1) The spec says javax.servlet.context.tempdir must be a > working directory > 2) TOMCAT sets that value to the value of java.io.tmpdir > > THEREFORE > > FOR THE TOMCAT IMPLEMENTATION > > java.io.tmpdir MUST BE WRITABLE. > > So, in short. The ELITE should spend a little more time > thinking about things instead of just reflexively trashing > people. That is respect. > > George Sexton > MH Software, Inc. > http://www.mhsoftware.com/ > Voice: 303 438 9585 > > --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]