Add description blurbs that happen to be picked up by facebook’s link blurb generator, somewhat addresses bug #90.

We have per-tab blurbs expanding the sections and a page-wide blurb listing the courses.

Fix support for displaying the number of credit hours in a permutation for schools with course slots, reported by bot10berg.

https://www.facebook.com/zimjimmy/posts/231316953666045?comment_id=886608&offset=0&total_comments=11
The linked schedule was displaying 24 credit hours instead of 14
credit hours because there were two course slots with two sections
each (each of these course slots having 4 credit hours) causing 8
credit hours to be doubly-counted.

Add some hacks to support pure HTTPS for WebAdvisor XSS automatic registration hack.

Instead of using a JSONP callback and PHP sessions to store the
TOKENIDX, the XSS script now just updates the URL parameter for the
WebAdvisor login page so that the TOKENIDX will be transferred to
webadvisor.php by a GET variable right when it is needed instead of
asynchronously.

A separate hack to help support HTTPS is that automatic uploading of
assets to Amazon S3 (which has HTTPS access) allows the XSS script to
be served over HTTPS. This eliminates the browser warnings about
accessing mixed secure/insecure content and, thus, hopefully supports
browsers which automatically block insecure content on secure pages.

Fix the support for automatic WebAdvisor registration to not enter infinite loops.

The automatic registration process used to enter an infinite loop
every time it was used if the user had been previously issued a
TOKENIDX by WebAdvisor. This means that it would work the first time
in most cases but rarely a second time, if ever. Now we use
LASTTOKEN=NULL instead of TOKENIDX= to request a new token and it
works pretty well.

Don’t automatically redirect to the somewhat broken automatic WebAdvisor registration page.

The automatic WebAdvisor registration page is sometimes broken and the user may genuinely
just want the list of section synonyms.

Fix qTips not appearing for certain courses because of partial XML underquoting.

Some courses which had ampersands in their titles would cause XML parsing errors in
the browser when trying to display the qTip, resulting in the qTip not showing
for those courses. This is because the qTip’s content is inserted as XHTML into
the document, so it needs to be doubly encoded.

Add support for manually specifying the base_uri.

This is necessary for me to continue developing on my messed up mpm_peruser
installation of slate_permutate since my apache's idea of its hostname is
wrong.

Add ability to distinguish, in most cases, between labs and lectures at Hope.

Hope's scheduling paradigm requires the use of course slots, so
distinguishing between labs and normal sections is vital.

Don't disable autocomplete for schools which failed to crawl in the last rehash.

This is especially important for hope as http://plus.hope.edu/ seems to always 503
in the middle of the night when we crawl it. Before this commit, that means that the
otherwise valid crawl data associated with hope would be available in cache/ but
would be ignored because hope was marked as uncrawled.

Add basic crawler support for Hope College in light of their winning the Calvin v. Hope basketball game last weekend.

The crawler is unable to, from the data provided by hope, properly recognize
and handle course labs in any fashion. To do that, the human-formatted comment
fields will need to be parsed. These fields are placed in the same column as
the course title is and are in the subsequent row for a particular section --
these fields are currently ignored through the rule which throws out rows with
too few columns.

Crawl and store credit-hours per section. Display credit-hours, but provide no UI for updating them. Fixes bug #114.

Credit-hour crawling support for calvin and cedarville.

Fix PHP strict warnings due to using Schedule::url() as both static and object method.

Introduce Schedule::my_url() as an object method and convert Schedule::url() into a static method.