home *** CD-ROM | disk | FTP | other *** search
- To: tlm@newton.physics.purdue.edu
- Subject: Re: testing dnet
- Precedence: special-delivery
-
- You need to unsetenv DNETDIR ... have DNETDIR not
- exist. Then, create the directories NET1 and NET2 and from the parent
- run the test program. Don't setup DNETHOST either.
-
- % unsetenv DNETDIR
- % cp ~/.DNET/dnet.servers NET1
- % cp ~/.DNET/dnet.servers NET2
- % mkdir NET1 NET2
- % test
-
- I found the bug... since there is no DNETDIR, the server tries to
- use the current directory. But since the current directory is
- specified by dnet.servers, it would not be in NET1 or NET2. You need to
- modify dnet.servers in NET1 and NET2 to set the base directory to NET1
- and NET2 respectively (the last entry on each line).
-
- DNETHOST should be set to "newton.physics.purdue.edu". It does what?
- The u gets eaten? Hmmm.. check the size of the array I stick the
- stuff in, you may have to recompile.
-
- Then, to run a client, it must be run FROM within NET1 or NET2 depending
- on where you want the source to be.
-
- -Matt
-
-
-