Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!usc!cs.utexas.edu!uunet!zephyr.ens.tek.com!uw-beaver!ubc-cs!news.UVic.CA!sol.UVic.CA!denhaana
From: denhaana@sol.uvic.ca (Albert Den Haan)
Newsgroups: comp.sys.amiga.programmer
Subject: RCS (and porting GNU stuff)
Message-ID: <1991Jun11.224756.18225@sol.UVic.CA>
Date: 11 Jun 91 22:47:56 GMT
Sender: denhaana@sol.UVic.CA (Albert Den Haan)
Organization: University of Victoria, Victoria B.C. CANADA
Lines: 57


	Here are a couple of questions about (almost) related topics,
one should be dear to almost all Amiga programmers and the other to users of
GNU source on the amiga.

	First some background. When I program I try to make changes to my code 
in an incremental manner.  That is, I try to get one module working before I 
work on the next in an attempt to keep the amount of information I HAVE to 
keep track of to a minimum.  (there are other rationals for this sort of 
develop ment process but I wont get into that here :-)  
	Of course this sort of process and the edit-test-debug cycle produces 
LOTS of different source code revisions and the problem I am worried about is 
keeping track of the revisions I make at each stage without maintaining 
archives of the full text of each revision.

	I know of two systems that automate this: SCCS and RCS.  Both systems 
keep track of the revision history of a given set of files by storing the 
CHANGES in the files from revison to revision.  This saves a great deal of disk 
space and allows more sophisticated analysis than 'It worked yesterday...'.  
These systems also help in multiprogrammer situations to avoid 'race' 
conditions where two (or more) programmers modify the same file in different 
ways at the same time.  I am currently using RCS on UN*X and enjoying the 
facilities it has for determining the *actual* changes I make to a file that 
cause things to break.  (This only happens to me right? :-)  Also a look at the 
blunders you consistently make in developing applications can be very 
educational!  Just make sure your boss doesent see them! 

	Enough for the intro.  I ftp'ed the RCS implementation from ab20
last week in hopes of getting the same facilities on my amiga.  No such luck.
I tried to run the binaries on my 2000HD (WB 1.3.2) and recieved various 
amusing crashes and gurus.  Attempts to un-RCS the source provided in RCS format on a UN*X box to attempt a rebuild with SAS C 5.10 provided more interesting
crashes.  Are those binaries so broken that the later patch is *necessary* or 
is it me?  Has anyone ported a later, working version of RCS (5.5 is the 
version I am using on UN*X) and would they mind posting it to 
comp.amiga.sources?  Once I started using this system at school a lot of 
revision-related headaches 'went away' an I would LOVE to have the same 
system at home.

	My second question (how many is that so far, really?) is how does one
go about porting a GNU program/system to the Amiga platform?  The SAS C run-time
library has quite a few UN*X 'compatible' functions but where does most of the
effort appear?  I wouldnt mind hearing from the people who have already ported
stuff like GNUgrep, GNUawk, bison, flex and yes, GCC to hear what efforts are
required for a good GNU port.  Maybe a GNU.port.FAQ report/posting could be
generated from the responses or a GNU.lib library built to save previous work
for later porting efforts to prevent wheel re-inventing.


				Albert.

P.S.  how does one go about getting patch 5.10a (or whatever is latest) to 
SAS C anyway?  e-mail me on this one.

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Albert den Haan.   (yes den Haan is my LAST name)
denhaana@sanjuan.UVic.CA
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=