Richard Gaskin wrote:
So lets dive in with lcVCS in v7 today, and with any luck the project
will attract enough contributors that they'll be able to handle at
least some of whatever work may be needed to port it to v8 later,
allowing you to maximize the time you spend on your externals which
the community depends on as well.
Good positive move to take the energy from this somewhat tense thread to
pour into a useful direction. Though I still think it behooves Kevin to
consider VCS for the whole community -- it would be "HUGE" for his
goals to make LC one of the world's top languages.
I did study the Git book and that level of code control, played with it
for a while using some scripts on the web server... I found myself
spending more and more time on the cmd line than I would have liked. No
doubt one who is using GIT a lot will become very efficient.. It
certainly is a powerful tool. But for one level of user it's a bit time
consuming and feels like it gets in the way...
Meanwhile... I guess what I'm saying is, a full blown GIT management of
scripts is scary to me when I would be content with "document"
control... where a stack is a document and in some contexts it can
simply be shared with someone else or "checked out" they work on it and
"check it back in" ... while it is "check out" I can't touch it. If
there were some way to regress and view changes that would be super, but
not necessarily required. A simple approach is, Person A gives it Person
B and B makes improvements. If nothing is broken... keep on going.. if
person B messes up... we delete his version and regress back one and
keep going...
I made my own "magic carpet" in-house for InDesign document RCS and our
team loves it. We have, in 4 years since we abandoned Adobe's version
control, not lost any work or the the ability to regress to a previous
version. 12 people working on the same document repositories on the LAN
server.
It would be simple for me to adapt my model to include HTTP calls to
the server. The model is super simple: document is archived and checked
in... if it is checked out by someone else, you can't touch it. When
someone else checks it back in, another copy is made both on the server
and locally. At anytime something breaks (iteraton21.livecode) there's
copies of the last revision (iteration20.livecode) in 3 places, on
user's A hard drive, the server and on user B hard drive. We can always
recover. Its simple but robust "pass the baton." RCS
I realize that the super coders would find that simply too limiting...
but I think it works for a lot of not-so-edge cases.
A strong Video screen tutorial on lcVCS might be useful. I want to see
if that's where I want to go, or resurrect Magic Carpet... Perhaps there
is, within lcVCS a way to keep it that simple.
Monte... do you have documentation I can read somewhere? I have a need
coming up here soon. I'm in the middle of working on a mobile app, and
will shortly reach my limits and then I'll want to pass it off to others
to improve, re-factor my code if necessary and fill out the features
that are beyond my competency. So I'm scratching my head right now
about just how to do that. Methods now are painful: FTP to server...
send someone an email. manually change file names etc...
Maybe we need to move this to a new thread? Anyone ever hear from Chip
in Texas? (author of Magic Carpet) Altuit.com not longer seems to be
up. Chipp seems to have moved on to other planets: http://blog.chipp.com/
Cheers from Hawaii. Monte, I hope your farm is not too cold down there!
Brahmanathaswami
_______________________________________________
use-livecode mailing list
use-livecode@lists.runrev.com
Please visit this url to subscribe, unsubscribe and manage your subscription
preferences:
http://lists.runrev.com/mailman/listinfo/use-livecode
