search

LockerProject / shared-ui

UI code shared in common between the Locker Project and Singly's hosted offering.
https://singly.com/
4 stars 2 forks source link

dust compiled_templates.js shouldn't be committed #2

Closed smurthas closed 12 years ago

smurthas commented 12 years ago

there are almost always conflicts

mdz commented 12 years ago

Seems like the kind of thing which should be done once, during the build in CI

smurthas commented 12 years ago

dust is not compatible with node v0.6 and doesn't look like it is going to be moved forward (straight forward pull req to get it working has been sitting there for 2.5 months untouched).

I think we should move to the very similar, and more widely compatible mustache. the transition should be mostly adding a second pair of curly braces around things, with a bit of updating to the compile scripts and likely a few subtle style changes.

Does that should reasonable? (haven't done a ton of work on templating solutions so looking for a bit of guidance, but also need to move quickly as this is about to punch us in the face if we don't fix it.)

/cc @kristjan @temas (of course, all are welcome to chime in!)

othiym23 commented 12 years ago

+1 for the simple reason that mustache.js seems to have much greater community adoption.

Caveat: Having multiple template systems is always going to be at least mildly confusing. Is there a reason to not do all the templating on the server side?

mdz commented 12 years ago

Sounds good, let me know if I can help

smurthas commented 12 years ago

If templating is all server-side, then we can't do things like talk directly to the registry from the front end. That might not be that bad, but it will result in a bunch of semi-common back end code (in addition to our growing amount of semi-common front end code), which I think would be taking us in the wrong direction.

smurthas commented 12 years ago

also, mustache can be used on the back end, just as well!

kristjan commented 12 years ago

+1, Moustache sounds good

As far as rendering server-side, I don't have any opinions for or against at the moment. Though it might be a significant rewrite.

Interestingly, it appears LinkedIn is doing it client-side, and in fact with dust: http://engineering.linkedin.com/frontend/leaving-jsps-dust-moving-linkedin-dustjs-client-side-templates

kristjan commented 12 years ago

Little more linky research:

http://stackoverflow.com/questions/5494839/templating-language-for-both-client-side-and-server-side-rendering http://stackoverflow.com/questions/7738345/should-i-be-rendering-html-on-the-server-side-or-client-side

Interesting uncommon approach: http://blog.nodejitsu.com/micro-templates-are-dead

smurthas commented 12 years ago

Yes, that linked in blog post was what lead us to dust in the first place. :)

I like the idea of Weld, but given that mustache will be a much closer drop in, I'd say we move to that now.

I'm going to go forward on this.

othiym23 commented 12 years ago

I really like the look of Weld, and it really doesn't look like using it is that much work over switching from dust to mustache.