Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!ucsd!ucbvax!hplabs!hpfcso!bruno From: bruno@hpfcso.HP.COM (Bruno Melli) Newsgroups: comp.lang.smalltalk Subject: Re: Re: GNU Smalltalk/Emacs Interactor Problem (and STIX) Message-ID: <9050001@hpfcso.HP.COM> Date: 27 Aug 90 17:24:09 GMT References: <15368@reed.UUCP> Organization: Hewlett-Packard, Fort Collins, CO, USA Lines: 41 >> "Also, anyone have STIX working on an HP 9000?" I just made quick and VERY dirty changes to compile STIX on a 9000/345 running 7.0. I replaced calls to alloca() in mstcomp.c, mstsave.c, and mstsym.c to calls to malloc(), and then calls to free() just before the return of whatever function was calling alloca(). Note that there is still a call to alloca() in mst.tab.c. In this case, I just wrote a dummy alloca that just calls malloc(). (This is the VERY dirty part :-) I'll look a little more into it when I get a chance. I also coded my own version of bcopy() : void bcopy(s1, s2, l) /* assume no overlapping */ char *s1, *s2; int l; { for(;l--;) *s2++ = *s1++; } There is also a call to getdtablesize() in socket.c. You can get the desired value by calling sysconf(_SC_OPEN_MAX). You need to #include . Note that I didn't try this. Over the weekend I didn't know how to get the desired value so I just put some arbitrary number and it worked. When I first ran mst, it didn't work until I added my machine name on line 75 of t.st. "Create the initial display connection" d _ Display host: '' display: 0. "Create the initial display connection" d _ Display host: 'my machine' display: 0. After these changes, everything went fine. Have fun, bruno.