My thoughts were, I guess, somewhere in the middle.

Sarah's suggestion of just keeping track of the configs in the Distribution Builder covers the major chunk of the code, and I hadn't thought of using the config settings as a sort of cm tool, but that makes sense. But (and it may be that I just am not thinking right about how the Dist Bldr works) several of my stacks don't need to be "built" (template stacks and stacks which retain preference settings, for instance, but contain very little 'code'). So, slightly along the lines of the arcane sccs, I need a clean way to ensure that I've checked the latest versions of the templates into the "build" that I'm creating.

I wasn't really looking for the full monte (pun intended) of something like Visual Age, but if that existed, would clearly warrant a try. I guess, then, Monte, my high-level objective statements might look (quickly) like this:

1st Objective: tracking file/stack/substack revision date/time/changedby/changenotes easily. (oh, and size)

2nd Objective: clean, easy-to-navigate object encyclopedia which shows created on/last modified/exists in or usedby/changenotes/changedby (and size)

3rd Objective: build tracker, showing builds of projects by target platform (really what I'd like right now). I'm not spun-in enough to do either of the first two, but may end up taking a crack at the 3rd before too much longer if nothing pops up.

Thanks for the feedback.
Cheers,
Ray

On Monday, Feb 10, 2003, at 20:18 America/New_York, Monte Goulding wrote:


Once worked with IBM Visual Age for Java. Other than being slow and
using a monstrous amount of RAM, it was a pretty cool IDE. Every
resource in the project was version controlled, automatically. No
messing with CVS or anything. So in the course of developing your
project, it would create a Library with the entire history of changes
documented and you could revert or visit snapshots at any step of the
way. This type of functionality, and also a nice "diff" viewer for
scripts would definitely get my $.

I'm not sure about automated history but it would certainly be possible to
store the changes in an objects properties (say at the end of every
session). Perhaps it would be best to let the user just take a snapshot of
all objects when they feel it would be useful? Personally I'd probably want
to link some notes to the snapshot.

One of the main things I'm looking at is distribution management. If I had
access to this kind of history data I could work out which files needed to
be distributed in an upgrade.

Monte

_______________________________________________
use-revolution mailing list
[EMAIL PROTECTED]
http://lists.runrev.com/mailman/listinfo/use-revolution

_______________________________________________
use-revolution mailing list
[EMAIL PROTECTED]
http://lists.runrev.com/mailman/listinfo/use-revolution

Reply via email to