The semantics component of the Web Standards Project forever rewired my brain. Applying structure and meaning to things seeped into all my thinking. It’s not a paralyzing process internally, but ignoring it feels lazy. Thanks Zeldman. This vexes me.
On GitHub each repository needs a name, so naturally it has to meet these criteria:
We all think this way. Right guys?
For many Git projects I can’t discern anything from the name. Surely there’s a standard to follow? Nope. Letter casing and punctuation are oddly devil-may-care for a development construct. So I tweeted and Bryan Veloso of GitHub answered:
@Falkowski — Depends on the type of project and the language. I usually follow their conventions.— Bryan Veloso (@bryanveloso) March 7, 2012
@Falkowski — Yep! I'd follow the domain.— Bryan Veloso (@avalonstar) March 7, 2012
My Git projects are usually for websites though, so many languages are in play. We concluded modeling after the domain makes sense.
http://domain.com ➔ domain.com.git http://sub.domain.com ➔ sub.domain.com.git
For other projects, I’m inclined to keep the lowercase and dashes pattern:
the-fellowship-of-the-ring.git the-two-towers.git the-return-of-the-king.git
The GitHub for Mac ‘Clone In Mac’ feature creates a new folder named for the repo — potentially breaching a meticulously crafted project folder. I’m terribly vexed.
Git’s command line swoops in to save the day and my sanity:
# Clone the repo into a new folder named after the GitHub project git clone email@example.com:username/project-name.git # Clone the repo into a new folder named 'zebra' git clone firstname.lastname@example.org:username/project-name.git zebra # Clone the repo into the current directory (must be empty) git clone email@example.com:username/project-name.git .
Now go forth and name/clone Git projects like a standardista.