[net.unix-wizards] Sccs vs. Rcs

david@ukma.UUCP (David Herron, NPR Lover) (05/10/85)

Over last christmass break I put all our 4.2 sources under RCS.  I'm not
pleased with the results entirely.

The problem basically is this.  Part of the reason we have a source liscence
is for students to be able to study the source code.  But, the most economical
way of using rcs (disk space wise) is to keep sources checked in all the time.
If you keep them checked in, then they aren't easily available.  You have to
check them out.  But if you're not on our staff then you don't know the
super-user password and can't check them out to this directory.  You have to
check them out to your home directory.  But that requires you to become a bit
of an expert in using RCS, which you've never seen before much less used.
And you only find out you have to use rcs if there is a *super*user* around
to answer questions.

What it boils down to is.  Neither rcs nor sccs do the thing I'd like to
be able to do, using a minimum of disk space.  That is, keep the current
version "checked out" all the time, and keep ONLY diffs in the delta file.

I can ALMOST do this with rcs.  But rcs wants to keep a copy of the file
in the delta file.  That doubles the disk space used.  Which might be
worth it if we had more eagles.  Buuutttt....

I've been thinking about this off and on for a couple of weeks and I've
decided that the editor needs to be brought in on the act.  i.e. we need
a front end for the editor which will save changes away automatically.
But it also shouldn't get in the way of using the editor either.

(Kind of like the description for Xanadu I read once.  Maybe I'll dig that
back out again.)

Any ideas?
-- 
--- David Herron
--- ARPA-> ukma!david@ANL-MCS.ARPA or ukma!david<@ANL-MCS> 
---	   or david%ukma.uucp@anl-mcs.arpa
---        Or even anlams!ukma!david@ucbvax.arpa
--- UUCP-> {ucbvax,unmvax,boulder,oddjob}!anlams!ukma!david
---        {ihnp4,decvax,ucbvax}!cbosgd!ukma!david

	"It's *Super*User* to the rescue!"

chris@umcp-cs.UUCP (Chris Torek) (05/13/85)

Suggestion: a ``co-like'' program that extracts an RCS revison, putting
the result on stdout.  I imagine this would be an easy hack to co (just
write to stdout instead of fopen()ing and writing there).

Then anyone who wants to examine a source file can say

	cocat sourfile.c | more		# cocat == co+cat...

or whatnot.
-- 
In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 4251)
UUCP:	seismo!umcp-cs!chris
CSNet:	chris@umcp-cs		ARPA:	chris@maryland

john@genrad.UUCP (John P. Nelson) (05/13/85)

In article <5688@umcp-cs.UUCP> chris@umcp-cs.UUCP (Chris Torek) writes:
>Suggestion: a ``co-like'' program that extracts an RCS revison, putting
>the result on stdout.  I imagine this would be an easy hack to co (just
>write to stdout instead of fopen()ing and writing there).
>

RTFM!  The "co" program already has a -p option that causes output to
go to standard out rather than to a file.

chris@umcp-cs.UUCP (Chris Torek) (05/14/85)

Guilty as charged:  I've never read the co manual.  (Me, read manuals? :-) )

However, co -p should do what the original poster was asking about, right?
-- 
In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 4251)
UUCP:	seismo!umcp-cs!chris
CSNet:	chris@umcp-cs		ARPA:	chris@maryland

david@ukma.UUCP (David Herron, NPR Lover) (05/17/85)

In article <5715@umcp-cs.UUCP>, chris@umcp-cs.UUCP (Chris Torek) writes:
> However, co -p should do what the original poster was asking about, right?

Not quite.  In my original posting I was complaining that it's nontrivial
to learn how to use rcs.  Therefore you want to keep the source file out
where it's easy to look at it.  (We're talking here about graduate students
who aren't system administrators ... not the type of people who would
already know what rcs is).  But if you also want to use rcs (or even 
sccs) to keep track of versions, you have an extra copy on disk plus diff's
between versions.  High cost making the sources available.  More than doubling
the disk space used, not to mention backup tapes, etc.


-- 
--- David Herron
--- ARPA-> ukma!david@ANL-MCS.ARPA or ukma!david<@ANL-MCS> 
---	   or david%ukma.uucp@anl-mcs.arpa
---        Or even anlams!ukma!david@ucbvax.arpa
--- UUCP-> {ucbvax,unmvax,boulder,oddjob}!anlams!ukma!david
---        {ihnp4,decvax,ucbvax}!cbosgd!ukma!david

	"It's *Super*User* to the rescue!"

david@ukma.UUCP (David Herron, NPR Lover) (05/17/85)

Oops!   I meant to add something else ...

I too didn't know about co -p.  It does ALMOST what I want.  See previous
article for reason why.
-- 
--- David Herron
--- ARPA-> ukma!david@ANL-MCS.ARPA or ukma!david<@ANL-MCS> 
---	   or david%ukma.uucp@anl-mcs.arpa
---        Or even anlams!ukma!david@ucbvax.arpa
--- UUCP-> {ucbvax,unmvax,boulder,oddjob}!anlams!ukma!david
---        {ihnp4,decvax,ucbvax}!cbosgd!ukma!david

	"It's *Super*User* to the rescue!"

olson@fortune.UUCP (Dave Olson) (05/20/85)

If you want to check out an RCS file, and have the file
appear on stdout, try using 'co -p'.  It works on the RCS
version I have, ich is pretty old, and I'm pretty sure it's
in the man page.
	Dave Olson