home *** CD-ROM | disk | FTP | other *** search
- Path: tut!osu-cis!cbosgd!clyde!rutgers!iuvax!iucs!bobmon
- From: bobmon@iucs.UUCP (RAMontante [condition that I not be identified])
- Newsgroups: comp.sources.bugs
- Subject: Re: Microemacs 3.9 spawn.c #if problems
- Message-ID: <4755@iucs.UUCP>
- Date: 24 Nov 87 17:51:16 GMT
- References: <2151@tut.cis.ohio-state.edu>
- Reply-To: bobmon@iucs.UUCP (RAMontante [condition that I not be identified])
- Organization: Indiana University, Bloomington
- Lines: 30
-
- mdf@tut.cis.ohio-state.edu (Mark D. Freeman) writes:
- >
- >In spawn.c, the "#if"s are inconsistant in the case of MSDOS and MSC.
- >This concerns the use of "system()" vs. functions defined in spawn.c
- >
- >I played with it a little, and got it to compile, but I can't push a
- >shell.
- >
- >Does anyone have the correct fixes?
-
- [On rereading, I'm not as sure whether Mark is referring to v3.9 or v3.9e.
- I assume he means v3.9 the original, not 3.9e.]
-
- I had the same problem, using Turbo C. I think that the problem is, the
- system() call executes command.com in batch mode only, on a null input. So
- it just returns immediately. A fix, though not a good one, is to change
- a line that looks like
- system("");
- to system("\\command.com");
- (sorry, I'm not sure exactly where this is, but it's in the spawncli function,
- near the middle/end (?:-)) This causes the batched command.com to run another,
- interactive copy, wasting memory but giving you the effect you want. This
- probelm was in v3.8i, also.
-
- BY THE WAY: in v3.9e, this is all VERY different. Apparently Dan Lawrence is
- replacing the system() call entirely with his own call, with some additional
- functionality. Unfortunately, it doesn't work either for me (TurboC, MSDOS3.1)
- so I've stuck the system() calls back in [in the shellprog() function, and
- disabled the execprg() function] until I can get it to work. Any fixes for
- this new stuff, anyone?
-