jantypas@hope.UUCP (John Antypas) (03/13/88)
I don't know if anyone else has tried this, but I have the sourcees for Sendmailand I reently tried to compiler them on Microport Sys V/AT 2.3. I used the V7 definition and removed all of the IPC, sockets, and logging code. Things went along pretty well -- files were compiling and then the compiler complains of syntax errors in /usr/include/sys/types.h! This is the first compilable line in the file! Why????? Microport's include files should not contain any errors.I use them in other things. What would confuse the compiler like this? John Antypas -- Soft21 --21st Century Software: New Product from FastBuck Enterprises! New Puzzle -- Rubik's Cling Wrap! The puzzle you get stuck on! (Along with everything else ... almost) UUCP: {garp, killer, pyramid, reed, sdcsvax!ucsd!ucrmath}!soft21!jantypas Internet: jantypas%soft21.UUCP@{eddie.MIT.EDU, UCSD.EDU, uunet.UU.NET}
bae@ati.tis.llnl.gov (Hwa Jin Bae) (03/14/88)
In article <152@ucrmath.UUCP> jantypas@hope.UUCP (John Antypas) writes: >I don't know if anyone else has tried this, but I have the sourcees for Sendmailand I reently tried to compiler them on Microport Sys V/AT 2.3. I used the V7 >definition and removed all of the IPC, sockets, and logging code. Things went >along pretty well -- files were compiling and then the compiler complains of >syntax errors in /usr/include/sys/types.h! This is the first compilable line >in the file! Why????? Microport's include files should not contain any errors.I use them in other things. What would confuse the compiler like this? I don't use V/AT but on 386 V.3, the a couple of sendmail source files include the "sendmail.h" file which in turn includes sys/types.h. The system V compilers seem to behave very sensitively when an include file such as types.h is included twice for some reason. I just removed the sys/types.h include line and moved the sendmail.h include line above other include lines - so that the ones that use sys/types.h will still be happy... and everything compiled just fine. In addition, if you look at daemon.c file, it uses "ifndef DAEMON" and "else" in very strange way (syntactically wrong, I think) which system V compiler refuses to compile (which is good). Some intuitive replacing/removing "else" will solve the problem. Incidentally, is anyone running a working SMTP based sendmail on microport V.3 based system? I would like to talk to you!!! Especially if you are using Excelan's TCP/IP software. I've got all of our custom software ported to this 386/ Excelan ethernet TCP/IP system but still having problems getting sendmail to work - debugging sendmail running in SMTP mode is kind of difficult due to its heavy use of "fork/exec" and IPC calls. Since the sendmail is a piece of code that tries to do what "inetd" or "bsd_inetd" should do AND SMTP and etc. it is unnecessarily complex... Hwa Jin Bae Control Data Corp. bae@{ati,aftac}.tis.llnl.gov (Internet) 4234 Hacienda Drive {ames,ihnp4,lll-crg}!lll-tis!bae (UUCP) Pleasanton, CA 94566 hbae@plseca (smail)