commit-then-review

[Jim Jagielski <ji...@jaguNET.com>]

One thing I would like is that before people work on stuff
that they simply say, for example, "I'm updating foobar in flurbl.c"
For example, there are many places in the code where it could
be tightened up a bit. Dean is a maniac with that (I'm just waiting
for assembler :) ) but others also like tuning up code (myself
included). Let's just say that both Dean and Ken get it into their
minds to rework the code in util.c... Both would be working on
the same stuff which could be a waste of at least some time.
Maybe Dean commits his stuff first and then Ken needs to compare
his work with Dean's and "replace" things that might be "better."
By announcing things first (1) it's possible that they could
work it together offline, instead of seperately since they both
know what they have in mind (2) it reduces wasted cycles.

Nothing fancy... maybe just a short blurb in STATUS similar to
the CONCEPT stuff...

Yea? Nea? Shut the hell up?
-- 
====================================================================
      Jim Jagielski            |       jaguNET Access Services
     jim@jaguNET.com           |       http://www.jaguNET.com/
            "Look at me! I'm wearing a cardboard belt!"

Re: commit-then-review

[Rodent of Unusual Size <Ke...@Golux.Com>]

Jim Jagielski wrote:
> 
> One thing I would like is that before people work on stuff
> that they simply say, for example, "I'm updating foobar in flurbl.c"
	:
> Nothing fancy... maybe just a short blurb in STATUS similar to
> the CONCEPT stuff...
> 
> Yea? Nea? Shut the hell up?

Yea.  To the concept, not to you shutting up. <g>

#ken	P-)}

Re: commit-then-review

[Michael Douglass <mi...@texas.net>]

On Fri, Jan 09, 1998 at 05:58:17PM -0600, Michael Douglass said:

> to take the CVS scripts that (already) email out the diffs of a
> commit and add to that the assignment of a 'commital' number to
> which it can be referenced for approval.

s/approval/review/

-- 
Michael Douglass
Texas Networking, Inc.

<tnet admin> anyway, I'm off, perl code is making me [a] crosseyed toady

Re: commit-then-review

[Michael Douglass <mi...@texas.net>]

On Fri, Jan 09, 1998 at 06:29:37PM -0500, Jim Jagielski said:

> All of us are "promising" that committed code will get a serious
> look through. There's also the note that pre-work communication
> will go on, which was absent from the original concept (or maybe
> just not vocalized)...

Okay, a voice that prolly doesn't count, but one way to do this is
to take the CVS scripts that (already) email out the diffs of a
commit and add to that the assignment of a 'commital' number to
which it can be referenced for approval.

In this way, instead of voting prior to commital with +1, you review
on a post-commital basis with a +1 in some file where the commitals
are tracked.  In this way, too, you can fairly quickly find any
commits that were not reviewed by a certain number of people, etc.

Just random thoughts as I process the thoughts of change.

-- 
Michael Douglass
Texas Networking, Inc.

<tnet admin> anyway, I'm off, perl code is making me [a] crosseyed toady

Re: commit-then-review

[Dean Gaudet <dg...@arctic.org>]

I'm not advocating that we stop communicating with each other... I'll
still babble on about crap I'm working on.  I like babbling.

Dean

On Fri, 9 Jan 1998, Jim Jagielski wrote:

> One thing I would like is that before people work on stuff
> that they simply say, for example, "I'm updating foobar in flurbl.c"
> For example, there are many places in the code where it could
> be tightened up a bit. Dean is a maniac with that (I'm just waiting
> for assembler :) ) but others also like tuning up code (myself
> included). Let's just say that both Dean and Ken get it into their
> minds to rework the code in util.c... Both would be working on
> the same stuff which could be a waste of at least some time.
> Maybe Dean commits his stuff first and then Ken needs to compare
> his work with Dean's and "replace" things that might be "better."
> By announcing things first (1) it's possible that they could
> work it together offline, instead of seperately since they both
> know what they have in mind (2) it reduces wasted cycles.
> 
> Nothing fancy... maybe just a short blurb in STATUS similar to
> the CONCEPT stuff...
> 
> Yea? Nea? Shut the hell up?
> -- 
> ====================================================================
>       Jim Jagielski            |       jaguNET Access Services
>      jim@jaguNET.com           |       http://www.jaguNET.com/
>             "Look at me! I'm wearing a cardboard belt!"
>