characterization of other version control systems

[cboettig]

Wow, great stuff. A few nitpicks I will file as issues. starting off:

Intro make it sounds a bit too much like git is the only distributed control system. (mecurial is also distributed). Perhaps unnecessary to emphasize the 'distributed' part so much, since I'm guessing most the audience comes from the final_draft_V3.doc school of management rather than the subversion school.

[yoavram]

I felt the same - I used mercurial until I wanted to use github and found it only supports git...
But they are really almost the same, the only differences I noticed are that mercurial is a little nicer if you want to interface it from python, and that mercurial has some nice plugins - hg-git allows you to have a local hg repository that talks to a remote git repository, and docdiff and diffpdf allow you to diff doc/docx and pdfs (see discussion at stackoverflow), although not as comfortably as plain text diff.

[karthik]

Great comments. I agree. I'm a bit stuck here because I really don't want to make this an long debate about the pros and cons of using different types of VCS (not the goal of the article). I didn't even want to mention centralized vcs but it's hard to jump into a specific kind without mentioning the alternatives.

The one big difference between git and Mercurial is that the latter does not support a branching model and is somewhat less granular than git (albeit more user friendly). That's one of the features I talk about as being useful in a scientific context. I've made a small note (for now) that there are other options but not really room to go into that.

Maybe the best way to address this without ignoring Mercurial might be to explain that other, more user friendly options are available and most of the use-cases described below could still be applied with other systems.

1 Git vs. Mercurial: Please Relax
2 The Differences Between Mercurial and Git

[cboettig]

Agreed. I think you are best off avoiding any statements that one version
control system is better than another to avoid silly flamewars. You can
simply state that there are a variety of well-known options and that you
will stick with git for simplicity of discussion.

In mentioning the advantages of distributed version control, say that
version control systems can be distributed, rather then dwelling on the
fact that SVN and CVS etc cannot be distributed.

On Thu, Jan 17, 2013 at 11:51 AM, Karthik Ram [email protected]:

Great comments. I agree. I'm a bit stuck here because I really don't want
to make this an long debate about the pros and cons of using different
types of VCS (not the goal of the article). I didn't even want to mention
centralized vcs but it's hard to jump into a specific kind without
mentioning the alternatives.

The one big difference between git and Mercurial is that the latter does
not support a branching model and is somewhat less granular than git
(albeit more user friendly). That's one of the features I talk about as
being useful in a scientific context. I've made a small note (for now) that
there are other options but not really room to go into that.

Maybe the best way to address this without ignoring Mercurial might be to
explain that other, more user friendly options are available and most of
the use-cases described below could still be applied with other systems.

1 Git vs. Mercurial: Please Relaxhttp://importantshock.wordpress.com/2008/08/07/git-vs-mercurial/
2 The Differences Between Mercurial and Githttp://www.rockstarprogrammer.org/post/2008/apr/06/differences-between-mercurial-and-git/

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/1#issuecomment-12387069.

Carl Boettiger
UC Santa Cruz
http://www.carlboettiger.info/

[sckott]

If git is just discussed as an example of many VCSs, then should the title of the ms just mention VCSs instead of git specifically

[karthik]

Wouldn't that be more misleading then? If the title were Review of tasty meats but the article said, I only discuss the awesomeness of bacon wouldn't that disappoint some people? Instead I'm choosing to call it Reasons why bacon is tasty but in the article I provide context and say while bacon is indeed tasty, people suffered through other, less tasty meats before. That way someone who has never eaten meat before (in this analogy, VCS newbies) wouldn't be led into thinking that there is no other kind of meat besides bacon.

Does my rationale make sense or do you think I should still reword the title?

[sckott]

makes sense

[yoavram]

So there's bacon and there's ham, both are tasty, not much of a difference, and you're writing about bacon because it is more sexy than ham.

[karthik]

Absolutely right. Admittedly the article is intentionally biased because I like bacon and I want others to know reasons why I (and my colleagues) think bacon is tasty. There are other cuts of meat that are also tasty but describing them is not my goal.