The thing that I find most strange about the popularity of distributed version control is that
people like that they can work off-line with them. Yet networking has become pretty much
ubiquitous, and the remaining non-networked places will certainly shrink even more over the
next few years. Is there some other factor involved?
Posted May 1, 2008 11:20 UTC (Thu) by boudewijn (subscriber, #14185)
[Link]
Working offline can be much more productive. Of course, if you're using a
compiled language and frequently have to wait ten minutes for a compile
to finish, it doesn't make much difference, but if you find yourself
quickly checking lwn.net between every ten keystrokes, going offline
might help recover productivity.
Offline working
Posted May 1, 2008 14:19 UTC (Thu) by engla (guest, #47454)
[Link]
The network always has latency far greater than the linux disk cache. A git diff can literally
be computed and displayed in the blink of an eye. This shows even more with being able to
merge, look at the result, re-merge to do it better (if needed) etc. All operations become
'cheaper', and many have noticed, just like Linus himself claimed, that there is a huge
difference between having to wait 10 seconds for an operation or just having it done by the
moment you press enter.
Offline working
Posted May 1, 2008 14:45 UTC (Thu) by scripter (subscriber, #2654)
[Link]
Taking a laptop on an airplane is a common occurrence. There usually no net connectivity
there. Programmers in third world countries often don't have net access from home, but they
might have connectivity from work or from an internet cafe. A remote office in the US may have
a very slow link to the main subversion server. Using a distributed tool (like git-svn)
reduces the pain.