Re: Vector programming

At 11:40 AM 9/28/99 -0500, you wrote:
>> Have you thought of using a standard CVS system ? (The kind of things the
>> various GNU projects use to check-in and check-out parts of code that are
>> being worked on).
>
>I'm familiar with the basic idea of CVS, but I haven't used it before. How
>granular of an item can you check out? Is it at the file level, or can you
>check out a specific section of a file? (If it is at the file level, I
>could break down the assembler into 1K segments or so.)
>
>But then I have to wonder if CVS is a bit of overkill, since we're just
>modifying or attaching comments to lines of code and not changing the code
>itself. The "step on your toes" aspect isn't there as much as, say, a
>large programming project would be.
>
>Hmmmmm. Your thoughts?
>
>
>

On all of the version control systems I've used, version control granularity
is to the file level. I haven't used CVS, but I'm sure that it is the same.
I think version control would be a good idea. You are still developing
the file, and you can track what you are doing much better than just with
files. It also lets you back out of changes if you make a mistake!

I keep document files under version control here at work, not just code.

Comments on your comments:
>004E AD:AA 03 LDA:abs $03AA # If the zapper has already
>0051 C9:02 CMP:imm #02 # been used twice, then
>0053 B0:08 BCS:rel Branch->$005D # branch to $005D.

I like the "If the zapper has already been used twice", that tells me what
the code is doing. However, the "branch to $005d" is redundant. The
assembler tells me that. I would suggest that you don't restate what the
assembler nmonics state but *what* you are doing. Like, "If the zapper
has already been used twice, then ignore super zap request."

Just my $0.02,
-Chris

--
Christopher V. Moore -- Principal Engineeer
Heartlab, Inc. - 101 Airport Rd - Westerly, RI 02891 -- www.heartlab.com
Phone: (401) 596-0592 x113 - Fax: (401) 596-8562 - Email: cmoore@heartlab.com