Hi Doug: Very interesting. A tool like Voyager is what I was hoping for. It allows 100% Java development and yet provides CORBA compatibility. The license is somewhat restrictive for public distribution. But we should study Voyager and probably use it in our project. I try to prepare something for the seminar. Doug, have you written the Demeter/Java page to be posted in the JavaCC pages? Demeter/Java is good enough for such a posting especially if we adopt a bazaar style. -- Karl >From dougo@ccs.neu.edu Tue Jan 6 01:56:50 1998 >From: Doug Orleans >To: jayantha@ccs.neu.edu, binoy@ccs.neu.edu, lieber@ccs.neu.edu, > johan@ccs.neu.edu >Subject: Re: Remote Methods & JavaCC > >Has anyone here checked out Voyager? Sounds intriguing. > >------- start of forwarded message ------- >Resent-Date: Mon, 5 Jan 1998 12:23:46 -0500 (EST) >Resent-Message-ID: <"pdWR5C.A.m-H.OaRs0"@charlie> >Resent-From: javacc-interest@charlie.metamata.com >Resent-Sender: javacc-interest-request@charlie.metamata.com >From: m.mower@unl.ac.uk (Matt Mower) >To: Mike Karlesky >Cc: javacc-interest@metamata.com >Subject: Re: Remote Methods & JavaCC >Date: Mon, 05 Jan 1998 17:07:51 +0000 (GMT) > >On Mon, 05 Jan 1998 10:10:20 -0500 (EST), Mike Karlesky > wrote: > >> >>Here's my situation: I'm using JavaCC and JJTree to accept programming >>statements for industrial robots. The produced parser/lexer is hosted on a >>server with client connections on the web. Users submit code, once it's >>syntactically correct, a VRML model of the robot is brought up and the >>instructions are simulated using the traversal provided by JJTree. I have >>the parsing pretty much done, I now have to work on the interpreter. >> >>I'm wondering what more knowledgeable java programmers think of using RMI. >>I think it would be an excellent alternative to developing a protocol >>between the client and server to traverse the instructions since the >>JavaCC and JJTree final products will be on the server. >> >>I would appreciate any thoughts on Pros/Cons, I'm new to RMI. Any pointers >>to RMI help would be appreciated as well. >> > >While RMI is a perfectly acceptable solution I would suggest taking a >look at Voyager (http://www.objectspace.com) which is IMO looking more >and more like a general solution for writing mobile/distributed java >code. The current production version 1 is quite usable, and version 2 >(in beta at the moment) has "transparent" links to both CORBA and RMI! > >In practice it's very similar to writing RMI code with the immediate >advantage of mobility (object instances can hop from one Voyager server >to another). To take your example you might use a mobile class to carry >the robot source to the server, have the server do the parsing & tree >generation, then have the mobile class hop back to the client to do >display work. This way network traffic is kept to a couple of short >bursts. Or you could just do it by remote method invocation - your >choice. > >Best regards. > >Matt. > >-- >Matt Mower, Information Systems Team, University of North London >T: +44-(0)171-753-3288 F: +44-(0)171-753-5120 E: m.mower@unl.ac.uk > >-- >The JavaCC home page is at http://www.suntest.com/JavaCC >The latest version of JavaCC is 0.7 >The JavaCC FAQ is at http://www.metamata.com/JavaCC/FAQ >Check out the JavaCC newsgroup at comp.compilers.tools.javacc >Send list administration mail to javacc-interest-request@metamata.com >Subscribe to javacc-interest-d-request@metamata.com for the digest > >------- end of forwarded message ------- >