[Simh] binary packages...

[J. David Bryan]

On 5 Sep 2007 at 11:48, Jason Stevens wrote:

> Let me know if there is something that you feel should be on the site..

As the current lead for the HP simulator, I tend to agree with Bob's 
comments here:

  http://mailman.trailing-edge.com/pipermail/simh/2007-August/001299.html

...regarding distributed source management.  I think that having a 
separate, parallel SIMH project would not be as valuable as making 
persuasive arguments to Bob or the simulator leads to accept changes 
directly into Bob's code base.

However, what would be very valuable, in my view, would be for the 
Sourceforge project to host diffs for users' locally modified SIMH 
versions.  I know that a bunch of people maintain local versions (myself 
included), consisting of patches either never submitted or submitted but 
rejected.  I think it'd be helpful to offer those to the community.

For example, I am maintaining about a dozen local enhancements that have 
been rejected by Bob.  Some are minor (e.g., a change to the NOBR command 
to clear the breakpoint at the current PC if no argument is supplied); some 
are major ("string" breakpoints based on console output from the simulated 
program, scripted responses to allow automated regression testing with 
diagnostics, etc.).  I know that Mark Pizzolato has made a bunch of 
scripting enhancements to the DO command ("On Error", "goto") that I would 
dearly love to use.  Undoubtedly, there are other useful enhancements that 
people have made.  Having these available to the community would be 
terrific.

One might argue that we could achieve this with a collaborative effort on 
shared source.  However, someone might want just my NOBR change and not my 
scripting change (or all of the other collaborative changes).  Diffing a 
source checkout against the official SIMH release and then sorting through 
the six hundred changes to identify just the ones pertaining to my NOBR 
change would be difficult.  I think it'd be easier for someone to pick the 
appropriate bits out of my diff, especially if a bit of subsidiary 
documentation were included.

                                      -- Dave