Coffee-script now a local dependency #98

merged 4 commits into from Jan 16, 2012

3 participants


This commit pulls in @pmuellr's #96 / #97 code and adds some small changes. renamed to to make it more obvious that you can 'coffee' to run the command line with your globally installed coffee-script if you wanted. Also removed some complexity from server/express/bin/server, it works the same now as it did before I believe, but let me know if I'm missing something important about having absolute paths in the require.

As a side note, I need a new name for server/express/bin/server, let me know if anyone has any good ideas, i'm leaning towards sfw-express, but i'm not totally sold.

pmuellr and others added some commits Jan 15, 2012
@pmuellr pmuellr make coffee-script a module of express server
These are some changes for issue #96 :

**express server should include coffee-script as a module**

* moved the bulk of what used to be in bin/server to

* changed bin/coffee to be a js file, not coffee, and have
  it invoke lib/ via require()

* added coffee-script to package.json
@nrn nrn Merge branches 'issue-96' and 'master' of… 4af8403
@nrn nrn Merge branch 'master' of…
@nrn nrn express: minor modifications to pmuellr's issue-96 fix 27def90

This looks good, thx for updating the doc as well.

Note that you included a "drive-by" fix for server/Wikiduino/, in commit 6bf16f9 . No problem for me, but not sure how "clean" you're trying to keep your commits/pulls.


Hmm, yeah, we've had a lot of turbulent stuff like that, it seems to be ok, but if you've got a better work flow that doesn't leave those artifact merge commits i'd be glad to hear about it.

@WardCunningham WardCunningham merged commit 4848fba into WardCunningham:master Jan 16, 2012

For our edification only: what would have been a good way to handle these commits and end up with a squeaky clean history. (My own understanding is limited to pull --rebase before pushing. I gather there has been more going on here.)


Nick is looking to rename "server/express/bin/server". Is there a problem with that name? The name "sfw-express" seems to encode information already present in the path to the file.


Something about the redundancy strikes me as off, maybe. I was thinking it could be more descriptive, but that's a good point that the path really does describe it just fine :)


Typically, to keep things squeaky clean, I create a new branch for everything that I want lumped together. Only make changes in that branch for things you want lumped together. For example, for every "bug" I work on, I create a new branch called issue-17 (where 17 is the bug number). Do all my work in that branch, when I'm done, go to the "master" branch (or whatever), and do a "git merge issue-17", then push to GitHub, then pull request.

For a days-long sort of fix, I'd work out of issue-17 and push it to my GitHub fork every night, then when I'm done merge it into master and do the pull request. I think if you're working out of a fork, merging to master probably isn't even needed (or desired) in most cases; just leave your work in your issue-17 branch and issue the pull request from there.

You can work on multiple bugs/new features simultaneously this way. Sort of loosely based on a development flow that I've used in the past:

One nice thing about this is, if you start making changes for issue-17 but you've forgotten to create the branch, you can create the branch then, and all the non-committed changes go with you. That is if you were starting from a clean branch already.

I'm less familiar with "fixing" a set of commits in a branch using --rebase, et al. I try to keep it from getting to that point :-) If I had to, I'd go look at one of the online Git books like

In the end, doing "drive by" fixes like this isn't really a big issue, I don't think. Sometimes it's too much work to be squeaky clean, when all you really need to do is be pretty clean. I wouldn't try to "fix" the commit with rebases unless you're familiar with the processes. OTOH, good opportunity to get familar with the processes.


For the server renaming, I've opened issue #99.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment