4 ms·

And if you learned SVN, you would have had a solid base for understanding GIT. Would you expect students to learn source code control in the abstract or not at

by crusso 10y ago

And if you learned SVN, you would have had a solid base for understanding GIT. Would you expect students to learn source code control in the abstract or not at all? You have to do programming assignments anyway. Why wouldn't you require students to learn and use the latest source code control tools while they're doing their development? Teach students to write tests, use source code control, utilize continuous integration, etc. Although the specific tools, languages, and approaches will evolve in the coming years - none of the above are going away soon.

sf_rob 10y ago
Exactly, we were checking in Java assignments in the form of a ZIP file in 2009 that we had validated with print statements. It wouldn't have been that much more work to structure the algorithm assignments in that class in a way more similar to industry workflows even if the workflows are an evolving target.
lacampbell 10y ago
Git is a very complicated piece of software. it's not intuitive, it has a famously poor CLI, and it takes time to learn. Time spent making undergraduates use git is time that could be spent teaching them long-term, fundamental skills.
- crusso 10y ago
  vi isn't very intuitive either, but there's no better way to learn something difficult than to learn it when you're young. I've been using vi professionally for 30 years thanks to my early GT classes. It's probably the single-most valuable skill that I learned there that I still use today.
  - lacampbell 10y ago
    I'm a vi(m) user, but I have to say - it's not a fundamental part of computing at all. It's just a very popular tool. A lot of people don't know how to use it and manage to make amazing things. there's no better way to learn something difficult than to learn it when you're young Hmm, define 'young'. I'm in my late 20s and I find it easier to learn new things more than ever - including things I failed to learn in my teens and early 20s. Maybe I'm just a late bloomer, and it took me a while to "learn how to learn". But maybe I'm still young in the eyes of someone who has been using vi for 30 years (:
hackermailman 10y ago
If you know Graph theory then you know git, all that remains is just reading the man page for specific commands. Intro to Graphs/Graph Theory is generally in the curriculum at all university compsci departments Testing ect is usually covered in all intro classes (assert libraries) or industry type testing like JUnit by a software engineering elective typically taught in Java
- lacampbell 10y ago
  If you know Graph theory then you know git, all that remains is just reading the man page for specific commands. Just because one of gits key abstractions is based on a kind of graph, I don't think it follows that knowing graph theory means you know git. I mean LISP is based on a graph structure as well but plenty of people find that confusing.

by crusso 10y ago

sf_rob 10y ago

Exactly, we were checking in Java assignments in the form of a ZIP file in 2009 that we had validated with print statements. It wouldn't have been that much more work to structure the algorithm assignments in that class in a way more similar to industry workflows even if the workflows are an evolving target.

lacampbell 10y ago

Git is a very complicated piece of software. it's not intuitive, it has a famously poor CLI, and it takes time to learn. Time spent making undergraduates use git is time that could be spent teaching them long-term, fundamental skills.

crusso 10y ago
vi isn't very intuitive either, but there's no better way to learn something difficult than to learn it when you're young. I've been using vi professionally for 30 years thanks to my early GT classes. It's probably the single-most valuable skill that I learned there that I still use today.
- lacampbell 10y ago
  I'm a vi(m) user, but I have to say - it's not a fundamental part of computing at all. It's just a very popular tool. A lot of people don't know how to use it and manage to make amazing things. there's no better way to learn something difficult than to learn it when you're young Hmm, define 'young'. I'm in my late 20s and I find it easier to learn new things more than ever - including things I failed to learn in my teens and early 20s. Maybe I'm just a late bloomer, and it took me a while to "learn how to learn". But maybe I'm still young in the eyes of someone who has been using vi for 30 years (:

hackermailman 10y ago

If you know Graph theory then you know git, all that remains is just reading the man page for specific commands. Intro to Graphs/Graph Theory is generally in the curriculum at all university compsci departments Testing ect is usually covered in all intro classes (assert libraries) or industry type testing like JUnit by a software engineering elective typically taught in Java

lacampbell 10y ago
If you know Graph theory then you know git, all that remains is just reading the man page for specific commands. Just because one of gits key abstractions is based on a kind of graph, I don't think it follows that knowing graph theory means you know git. I mean LISP is based on a graph structure as well but plenty of people find that confusing.