I think seed-element shouldn't advise ignoring .bowerrc

[mco-gh]

The http://www.polymer-project.org/docs/start/reusableelements.html doc prescribes this step:

Add .bowerrc to a .gitignore file so it doesn’t get checked in

This can be done using the echo command or a text editor

echo '.bowerrc' > .gitignore

I think this causes problems. Our component import paths are setup to use the prefix ../, rather than bower_components so if you ignore this .bowerrc setting, then next time you check out a component and run bower install, it will put everything in the bower_components dir, not the parent dir, and your import paths will not see the newly installed comps. So you end up with a working dir with some stuff in the parent and some stuff in a subdir. This is very confusing. I suggest leaving the .bowerrc file intact for seed-element and replacing this prescription with a block of text better explaining what's going on here and how any comp derived from seed-element is free to override or ignore .bowerrc per their desire, but let's make the simplest, default case just work.

If this issue is approved, I volunteer to write the text proposed above.

[robdodson]

Instead of gitignoring it, we could add it to the ignore block in the bower.json

[mco-gh]

The point is that I don't think it should be ignored. I think it should be
retained, for reasons outlined in the issue text.

On Tue, Jun 24, 2014 at 5:02 PM, Rob Dodson [email protected]
wrote:

Instead of gitignoring it, we could add it to the ignore block in the
bower.json

—
Reply to this email directly or view it on GitHub
#533 (comment).

Marc Cohen
[email protected] [email protected]
http://about.me/marc1

[robdodson]

There's a difference between a gitignore and a bower.json ignore. I agree with you, we should keep it in the git repo.

Adding it to the ignore block in bower.json just means that when a user runs bower install x-foo, they won't end up with a .bowerrc in their bower_components/x-foo dir. At that point, it would be useless because bower never checks a bower_components dir for .bowerrc files.

[mco-gh]

That seems reasonable. Let's see what Addy and Eric think (anyone else we
should consult?). If they agree, I'll submit a PR undoing the gitignore
instruction in the docs and add the ignore block to seed-element's
bower.json.

Thanks Rob!

On Tue, Jun 24, 2014 at 7:27 PM, Rob Dodson [email protected]
wrote:

There's a difference between a gitignore and a bower.json ignore. I agree
with you, we should keep it in the git repo.

Adding it to the ignore block in bower.json just means that when a user
runs bower install x-foo, they won't end up with a .bowerrc in their
bower_components/x-foo dir. At that point, it would be useless because
bower never checks a bower_components dir for .bowerrc files.

—
Reply to this email directly or view it on GitHub
#533 (comment).

Marc Cohen
[email protected] [email protected]
http://about.me/marc1

[robdodson]

Yeah I think @ebidel and @addyosmani can weigh in. We're all slammed with I/O stuff at the moment but I want to give seed-element and the yeoman generator some more love next week.

[addyosmani]

There's a difference between a gitignore and a bower.json ignore. I agree with you, we should keep it in the git repo.

I agree too.

Adding it to the ignore block in bower.json just means that when a user runs bower install x-foo, they won't end up with a .bowerrc in their bower_components/x-foo

I'm totally down for the ignore block being included in bower.json.

[robdodson]

Moving discussion to seed-element repo.
googlearchive/seed-element#18