From mboxrd@z Thu Jan 1 00:00:00 1970 From: Vincent D Murphy To: 9fans@cse.psu.edu Message-ID: <20010114165606.A18906@student.cs.ucc.ie> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii User-Agent: Mutt/1.1.1i Subject: [9fans] heterogeneous cpu servers? Date: Sun, 14 Jan 2001 16:56:06 +0000 Topicbox-Message-UUID: 4c551464-eac9-11e9-9e20-41e7f4b1d025 can a NT, linux, or solaris box 'be' a plan9 cpu server? i fear not; if this is the case, does anybody have an idea what would need to be done to make it happen? rationale: i have to develop java at work -- my duties dictate it. however i have found that i am far more productive working in plan9/acme, and now i want to use it for java, perl and oracle development work. maybe i could even win over a few converts among my co-workers. :) to do this, i need to access tools on other hosts, running operating systems which are better supported than plan9 by the developers of those tools. hopefully it is possible to do this as transparently as possible, through creating a distributed development environment. for example, it would be nice to run 'java/javac foo' or similar on my plan9 terminal, and have javac run on a solaris box, with the obvious multiplexing of /dev/cons and friends taken care of. then i could process the stderr output of the compiler such that it is acme-friendly. needless to say, such a capability would be immensely useful for distributed systems in general, rather that just 'hacking foo with acme'. i have looked at cpu.c, but i remain in the dark as to how plan9 cpu servers work from having read the source. any pointers to further (prosey) info, apart from cpu(1)?