A lot has been said about Git. There are tons of articles written about how cool and efficient this new SCM is. I was already comfortable with all the features of Subversion and was reluctant to move. But then after a strong urging from my colleague Sankara, I thought to give Git a try.
One thing I can say is that its very very rich in features. It makes Subversion look like a small subset of it. Furthermore it keeps the entire repository in my development machine. It has reduced my dependency on network to reach the Subversion by 90%.
So myself and Sankara are planning to convince other (rather lazy) developers in my company to move to Git. As a sampler we are using git-svn. But it has some issues brought about by the Subversion end of thing. But it works great, surely better than just using Subversion client. But we are planning to move the entire code base to git repository for better leverage.
It there is anyone who is starting a new project and is thinking about which SCM to choose from, think no more. Git is the panacea of all the troubles of tracking code. I have been a huge fan of Linus, and he continues to amaze people like me.
I have been keeping a copy of Donald E. Knuth, The Art of Computer Programming as a reference for few years now. The level of attention to detail is just mind boggling. But the entire book is written for a hypothetical architecture which is like a ASM. Its not only frustrating but makes me skip over pages when I find the reading little painful. I naturally resort to wikipedia for pseudo code or algorithm mentioned in the sections of TAOCP. I would like to see a version of TAOCP upgraded to a more contemporary language or base it on better programming/psuedo-code constructs. Even with these drawbacks, without doubt TAOCP is the best source of reference I know of today.
Subscribe to:
Posts (Atom)
