dukechronicle / chronline Goto Github PK

RDiscount is faster and compiles gorgeously on Windows... apparently bluecloth's header files are out of order. You could make this conditional if you're really set on using bluecloth

Ffaker = faster faker

Success on Windows (ish)

See https://devcenter.heroku.com/articles/error-pages

Currently redirects to http://s3.amazonaws.com/heroku_pages/error.html and http://s3.amazonaws.com/heroku_pages/maintenance.html

Administrators should be able to post articles without publishing them for general viewing. Articles may only be seen once published_at is set.

DRY and toasty goodness:

class ArticlesController < ActionController::Base
  def index(); @articles; end
end
class Admin::ArticlesController < ArticlesController
  def create(); end
end
class API::ArticlesController < ArticlesController
  def index(); super; render json: @articles; end
end

Only issue to resolve would be the BaseController.

http://www.quantcast.com/

If the photo should be credited to a non-staff member (Special to the Chronicle), it should use this field instead of photographer_id.

http://webdevchecklist.com/

I'm going to tweak the lib directory a bit to aid with autoloading the gem extensions and what not.

I'll have a sample up in the config branch in my fork soon enough.

https://developer.apple.com/videos/ios/?id=12

Someone with a mac and an interest in IOS should take the lead on this one.

Not sure whether to use mailto or send the email from the server.

Support IE 8, not 7. http://www.browserstack.com/.

Right now authors can be created on the fly from the submission form. The behavior should be restricted via some validation

And don't make everyone else install Node and Bower just for assets...

Write RSS source feeds for the whole site, individual sections, and each author.

Instead of /article/--, the route should be /articles////. Also consider the issue of changing slugs.

You'll probably want to talk to @fny if you have any questions about using the search API.

Chronline is a perfect use case.

And use Redis for caching.

yesh says either just a reference list somewhere of the articles posted that day, or to prepopulate the search box before text is typed with the recetn would be good.

this way they dont have to remember the names of all the articles when they are doign the layouts

http://wiki.apache.org/solr/MoreLikeThis

Image deletion should not leave article with hanging reference.

Right now there's no distinction between admin/mobile/www.

There's no flash message on failed login.

Also I get the following warning in the logs:

WARNING: Can't verify CSRF token authenticity

scrollHeight is not working for some reason.

They're all the same...

It's kind of messed up.

There should be many different fail pictures that are stored (or the links are stored), and one is selected at random each time a 404 page is hit.

If I search for "Duke", then duke should be bold in all of the search results. See how it's implemented now if you need help.

Anchor tags outside of block level elements (e.g. <a><h1>whatever</h1></a>) are invalid HTML.

Sections such as .content-container a hx and related CSS need fixing.

Note that this also restricts the use of the Rails #link_to() helper.

You'll want to use Single Table Inheritance to have both Article and BlogPost share a table. A working example is the Staff model. Also, with respect to tagging it appears there are quite a few gems that do it, so don't roll out your own tagging system.

Its missing.

Pages with indefinite titles currently render with "| The Chronicle" (<-- that's a pipe). It should just be chronicle

If the skolar font doesn't load, the views get funky. I noticed this during dev because I kept getting a "403 (FORBIDDEN)" from Typekit.