| Re: Version Control Strategies For Block-Based Forths? |
|
 |
|
 |
|
 |
|
 |
Group: comp.lang.forth · Group Profile
Author: spamspam Date: Sep 20, 2008 10:52
On Sat, 20 Sep 2008, Dennis Ruffer wrote:
> Date: Sat, 20 Sep 2008 14:17:52 GMT
> From: Dennis Ruffer speakeasy.net>
>
> I remember doing many paper comparisons, along the lines of the thin paper
> overlay method described by John Passaniti earlier. The "triad" was a
> classic printing pattern for blocks in polyForth where 3 blocks would fit on
> one printed page. In the end, the shadow would also fit with a smaller font.
> The index pages would just print the 1st 64 character line of each block and
> the convention to put a revision date at the end of that line helped
> significantly.
>
> I don't remember the tool that Elizabeth mentioned, but I remember many
> manual comparisons. Since most systems filled their available block ranges,
> having two sets of blocks online at the same time was rare. The best we
> could usually do was to compare the printouts.
>
Given the requirement for a block based system, and the OP's stated goal
of "not re-inventing the wheel", I'm just wondering whether the OP's
system has (or more to the point is lacking) an umbilicus to a hosted
system (Linux/BSD/Unix), which has any of a number of source management
tools (sccs, rcs, cvs, svn, git ... whatever) available to it. Why
attempt to force an ancient block based system accomplish a very prevalent
task which can easily (trivially) be accomplished on any modern operating
system ... or even Windows ... ???
Surely there is a simple way to xfer source from a modern source
repository/config tool either to a X-dev'nt system or directly onto the
target processor, no?
Cheers,
Rob.
| 
