peregrin@hulaw1.harvard.edu (03/05/91)
Does anyone have a working version of Purdue's RCS under DOS? I've tried the version on SIMTEL and it is so shaky as to be unusable. I've spent a couple of days porting the source from comp.binaries which is in MS C 5.0 to Turbo C 2.0. It works, (except rlog) but I don't trust it yet. I'd rather not buy a commercial version control system, since I have simple needs (only me programming, no network). James +----------------------------------------------------------------------------+ + James Peregrino | JPEREGRINO@HBSSTG.HARVARD.EDU + + Programmer/Analyst | PEREGRIN@STIG.HARVARD.EDU + + Science & Technology Interest Group | JPEREGRINO@HBS.HBS.HARVARD.EDU + + Harvard Business School | PEREGRIN@HULAW1.BITNET + + Boston, MA 02163 +-----------------------------------+ + Voice: (617)495-6307 | My opinions are only my own and + + FAX: (617)495-0351 | never H.B.S's. + +----------------------------------------------------------------------------+
NU013809@NDSUVM1.BITNET (Greg Wettstein) (03/06/91)
> Does anyone have a working version of Purdue's RCS under DOS? > I've tried the version on SIMTEL and it is so shaky as to be unusable. > > I've spent a couple of days porting the source from comp.binaries whic > is in MS C 5.0 to Turbo C 2.0. It works, (except rlog) but I don't trust it > yet. > > I'd rather not buy a commercial version control system, since I have > simple needs (only me programming, no network). > > James I was recently faced with the same troubling thoughts about the various versions of RCS floating around the PC world. One hesitates before committing sources of importance to a source code control system of questionable reliability... To avoid any concerns on my part I picked up the 5.5 sources from prep.ai.mit.edu and ported them to MS-DOS under MSC 5.5. I use ms_shell exclusively as my command interpreter and with the 5.5 sources comes a very nice Bourne Shell script which tests RCS fairly thoroughly. I currently have the port in a condition in which it passes the rcstest script completely. My intention (if there is interest) is to pack up the binaries and ship them to our eminent moderator for distribution. The only thing that has delayed me from shipping them is that I am writing a C version of the merge script which is called by the rcsmerge program to call diff3 and run the three-way file comparison through ed to produce the final merged output. Since not everyone uses (or wants to use) a version of sh under MS-DOS I thought that this would be needed to make the package fully usable under most flavors/environments of MS-DOS. As soon as the C version of the merge script passes rcstest I will ship the binaries to Mr. Davidsen. The package is currently set up to use the MS-DOS version of GNU DIFF as available from ocf.berkeley.edu as part of the GNU-ish MS-DOS project. The rcsmerge program also requires the availability of an MS-DOS versio of ed. I found a version of ed whose original source came from the MINIX project and appears to be freely distributable. I will enclose a copy of ed in the binaries package so that users will have access to pretty much everything they need to get started with version control under MS-DOS. I assume that everyone has a copy of GNU DIFF 1.14 or has the ability to get at it. If this assumption is incorrect please advise me and I could package up diff and diff3 as well. As to rcsfreeze and rcsclean which are implemented in Bournce script I will include these but leave it up to the final user to reimplement them or to get a copy of ms_shell or something similar. If anybody has any questions or comments please feel free to drop me a note. Please use the e-mail address in my .sig as that will route mail to me much faster than a reply to this account. As always, Dr. G.W. Wettstein Oncology Research Division Computing Facility Fargo Clinic / MeritCare UUCP: uunet!plains!wind!greg INTERNET: greg%wind.uucp@plains.nodak.edu Phone: 701-234-2833 `The truest mark of a man's wisdom is his ability to listen to other men expound their wisdom.'
robl@idca.tds.PHILIPS.nl (R. Luursema) (03/09/91)
In article <91065.084335NU013809@NDSUVM1.BITNET> greg%wind.uucp@plains.nodak.edu (Greg Wettstein) writes: ><...> I picked up the 5.5 sources from >prep.ai.mit.edu and ported them to MS-DOS under MSC 5.5. <...> I currently >have the port in a condition in which it passes the rcstest script completely. >My intention (if there is interest) is to pack up the binaries and ship >them to our eminent moderator for distribution. I think thats a very good idea! I picked up RCS from c.b.i.p last year (Posted 17 Aug 90, V07i145), but I failed to get it working properly. And it couldn't be fixed without recompiling with the right set of flags. So I put this pakage aside. >I will enclose a copy of ed in the binaries package so >that users will have access to pretty much everything they need to get >started with version control under MS-DOS. I assume that everyone has >a copy of GNU DIFF 1.14 or has the ability to get at it. If this >assumption is incorrect please advise me and I could package up diff and >diff3 as well. Please Mr. Wettstein, make the package as complete as possible, with the right set of tools. Make it independent of tools someone may have laying around but are of version 1.1 while RCS requires version 2.0 for example. So include the tools required and not obvious a dos-user would have, and provided they are free/share-ware. DIFF(3) and SH are not obvious since there are dozens versions of it. ED is also not obvious, while EDLIN is. I have MKS toolkit which contain the necessary tools, but I am not sure if it all works flawless. II'd rather stick to something thoroughly tested. -- _ _ / U | Rob Luursema, Philips Information Systems Apeldoorn /__ < robl@idca.tds.philips.nl 88 |_\ "The trouble with everyone is that they generalize too much"