Wang Zeep <G.ZEEP%MIT-EECS@MIT-MC.ARPA> (01/16/85)
1) MSDOS 2.0 and up have big chunks written in C. (Just FYI) 2) Why don't 40-80 hackers get together, sign incorporation papers, get a contract with AT&T as a company and split the cost of the source license. Call yourself a consulting consortium or something. While this solution may be a bit unwieldy, anyone who has experience in hacking the internals of UNIX can make lots of money as a consultant and could easily make back the $40K total in 2-3 days each! [yet another modest proposal] wz -------
chuqui@nsc.UUCP (Chuqui Q. Koala) (01/18/85)
>2) Why don't 40-80 hackers get together, sign incorporation papers, >get a contract with AT&T as a company and split the cost of the source >license. Call yourself a consulting consortium or something. While >this solution may be a bit unwieldy, anyone who has experience in hacking >the internals of UNIX can make lots of money as a consultant and could >easily make back the $40K total in 2-3 days each! AT&T licensing is on a per CPU license. Those 40-80 hackers all need to be able to share a single computer, or do something to allow multiple CPU's (#2 is, I believe $16K, #3-n $1k under contractors provisions). Otherwise, you're STILL illegal. You know, I remember way back in the good old days when we talked about technical issues in unix-wizards. I'm glad the net is working so well now we don't need to anymore.... -- From the ministry of silly talks: Chuq Von Rospach {allegra,cbosgd,decwrl,hplabs,ihnp4,seismo}!nsc!chuqui nsc!chuqui@decwrl.ARPA Vote for the fascist of your choice, but vote!
tim@callan.UUCP (Tim Smith) (01/25/85)
In article <2229@nsc.UUCP> chuqui@nsc.UUCP (Chuqui Q. Koala) writes: >AT&T licensing is on a per CPU license. Those 40-80 hackers all need to be >able to share a single computer, or do something to allow multiple CPU's >(#2 is, I believe $16K, #3-n $1k under contractors provisions). Otherwise, >you're STILL illegal. > How are multi-CPU computers gonna fit in? Will it really cost more to get a source license for an N cpu system, when they are running as one computer? Imagine the kludges needed for a company that can only afford one source license! I can see it now... open( some args... ) { . . . if ( ip->st_mode & S_SRCFILE ) p->p_srccnt++; . . } sched() { . . if ( p->p_srccnt ) if ( available[SRCCPU] == NO ) goto schedloop; /* * Every vital kernel routine * is required to have a goto! */ else cpu = SRCCPU; else cpu = select_cpu(); . . } :-) And how do networks fit in? The machines I am on at Callan are hooked up with Ethernet. When I am going to edit a source file, I usually copy it from the development system to my office system. I also compile it and test it on my office system. Then, when I am done with it, I send it back and rm it from my system. Is this a violation of the license? And how about smart terminals? Suppose I had an editor that down- loaded the file to be edited into the terminal, and the editing is done by the terminal, with the changes uploaded to the host? Does this violate the license? -- Duty Now for the Future Tim Smith ihnp4!wlbr!callan!tim or ihnp4!cithep!tim
jbn@wdl1.UUCP (01/28/85)
Now that the Sun/Pyramid/Vax/Gould distributed file system is becoming available, we are all going to need a lot fewer source licences. Not to mention a lot less disk space.