history.txt

See ./README.txt for an overview.


== Nov 5 2012 ==

 # Created original project at csmarlboro.org,
 # /home/mahone/site/flaks/pe/ 
 # i.e. http://csmarlboro.org/mahoney/flask/pe/
 csmarlboro$ cd ~/site/flask
 csmarlboro$ mkdir pe; chdir pe  # "Planet Express" project

 # created ./README.txt and ./history.txt

 # initialize local python environment for the project
 csmarlboro$ virtualenv -p python2.7 env

 # initialize folder as git repository
 csmarlboro$ git init
 csmarlboro$ git add *
 csmarlboro$ git commit -m "initial commit"

 # Also did (as prompted; first time with git on this machine)
 csmarlboro$ git config --global user.name "Jim Mahoney"
 csmarlboro$ git config --global user.email "mahoney@marlboro.edu"

 # Now on my laptop, thirty-two
 # This folder is csmarlboro.org:/home/mahoney/site/flask/pe/
 # in addition to its web address, http://csmarlboro.org/mahoney/flask/pe/
 # I've cloned it to my laptop, using ssh access with public/private keys
 thirty-two$ cd ~/academics/web_development; mkdir flask; cd flask;
 thirty-two$ git clone csmarlboro.org:site/flask/pe/
 # The workflow on my laptop is now :
 # 1. edit files
 # 2. $ git add <files>; git commit "message"; git push

Hit a complication. As I originally thought of this, 
I wanted the files visible at csmarlboro.org, but
also wanted the working copy on my laptop able to
ush changes there. Turns out git doesn't like this model.
Any git repo with visible files is considered to be
a "working" copy, with its own .git tree of changes.
Pushing to a working copy is strongly discouraged,
since effectly you'd be chaning the .git tree without
changing the working files, which would then need
to be manually reset. See e.g. 
http://www.bitflop.com/document/111

So, after a bit of playing around, I've set up a 
"bare" git repository (i.e. one designed to act
as just a push/pull location, without any working
files, essentially just the .git part), from which
both csmarlboro and thirty-two can push and pull
changes. This means that to propogate changes
from thirty-two to csmarlboro's working directory
I'll need to (1) push from thirty-two to the
bare repo, and then later (2) pull from csmarlboro
from the bare repo.

I've set the bare repo as flask/.pe_bare_repo, 
and effectively created it this way:

 # starting from the working pe/ on csmarlboro,
 # create a bare copy :
 csmarlboro flask$ git clone --bare -l pe .pe_bare_repo
 # get rid of that working copy, and clone the 
 # the bare repo so that push/pull go to the bare one
 csmarlboro flask$ rm -rf pe
 csmarlboro flask$ git clone .pe_bare_repo pe
 # Then on thirty-two, edit pe/.git/config 
 # to point at flask/.pe_bare_repo/ on csmarlboro

== Nov 12 ==

Futzed around with the directory structure. Ended up putting virtualenv
stuff in env/ folder with the intention of *not* putting that in the
git repo, since the stuff in it depends on the OS system specifics.

 * Added a number of static files (Bootstrap, jQuery, underscore.js, and
   some of my "default" html5 files).

 * Added pip's requirements.txt ; with virtualenv should be 
   simple to create virtualenv's env/ and put in dependencies.
   See that file for the recipe.

 * Added .gitignore ; put env/ in it.

Started work on putting SQLAlchemy into the project.  Am trying
Flask-SQLAlchemy ... not sure if it's going to be much better than
just doing that stuff without it.  Also not sure whether it's bringing
the database connection up and down at the right times in the request.

There's a long list of possibly interesting Flask extensions;
see docs/thoughts.txt :
 flask-assets  		css, js, compiling
 flask-debugtoolbar	django-debug-toolbar clone
 flask-lesscss
 flask-login		... eh
 flask-restless

 flask-admin		library for an admin interface
 flask-superadmin 
 flask-dashed

 flask-mustachejs	fancier templates, I think
 flask-geokit
 flask-security

 flask-coffee2js

== Nov 16 ==

Looking for images & fonts

Planet Express: Our crew is replaceable. Your package isn't.

http://www.fanpop.com/clubs/futurama/images/603850/title/futurama-wallpaper
the frytrix - matrix hack futurama poster

http://www.fanpop.com/clubs/futurama/images/3305283/title/planet-express-ship-wallpaper
planet express ship

http://en.wikinoticia.com/Technology/general-technology/78033-the-ten-most-iconic-spaceships-of-science-fiction
http://www.wikinoticia.com/images/alt1040/cdn.alt1040.com.files.2011.03.Planet-Express-Ship-futurama-3305283-1024-768-w640.jpg
 planet express ship (black background)

http://newsentimentality.wordpress.com/2012/04/09/space-truckin/
http://newsentimentality.files.wordpress.com/2012/04/planet-express.jpg
 planet express ship (blue background)

(cc) skeletal planet express ship
http://www.flickr.com/photos/gabrielsond/4665229709/

(cc) planet express logo yellow/red/black
http://www.flickr.com/photos/patdavid/7562368060/
cleaned up in gimp
 * added transparency to backround
 * put insides of letters to yellow
 * saved as 

google.com/webfonts
 droid serif (normal, bold, italic, bold-italic) & sans (normal, bold)
  html:
    <link href='http://fonts.googleapis.com/css?family=Droid+Sans:400,700|Droid+Serif' 
          rel='stylesheet' type='text/css'>
  css:
    font-family: 'Droid Sans', sans-serif;
    font-family: 'Droid Serif', serif;

The bold Droid Sans looks close enough to the logo to use.

gradient backgroud , from http://webdesignerwall.com/tutorials/cross-browser-css-gradient

 /* for non-css3 browsers */
 background: #999; 
 /* for IE */
 filter: progid:DXImageTransform.Microsoft.gradient(startColorstr='#cccccc', endColorstr='#000000'); 
 /* for webkit browsers */
 background: -webkit-gradient(linear, left top, left bottom, from(#ccc), to(#000)); 
 /* for firefox 3.6+ */
 background: -moz-linear-gradient(top,  #ccc,  #000); 

== Nov 19 ==

Got src/model.py into a reasonable state,
using SqlAlchemy relations and classes,
and roughly equivalent to the SqlSoup work
I did earlier this semester.

The console is also in a working state,
and does allow modifications to the database.

For a template, adopt from fedex.com/us ; see attached.

== Nov 23 ==

To add variables or functions to the template context, flask/jinja2 
provides the following syntax (see http://flask.pocoo.org/docs/templating/)

@app.context_processor
def template_context_additions:
    # return dict of variables & functions to put in template namespace
    return {func = func, name = value, ...}

You can also define template filters, which modify the result of 
variables in templates in a unix-ish pipe notation

 {% keyword ...  %}          statement in jinja2 language, e.g. if, for, block, ...
 {{ name }}                  insert variable into document at that spot
 {{ func('arg') | upcase }}  produce text, pipe through filter

These functions can be 

jinja can also define its own functions, which it calls "macros",
and can import those from other files.

... Got a simple bootstrap example running as a flask template,
with the static urls translated via a filter 

== Nov 26 ==

Themes for bootstrap : http://bootswatch.com/

== June 2013

 Looking back at this much later.

 Here's the basic idea of how it runs :
   # Get to the base directory
   $ cd /Users/mahoney/academics/web_development/flask/planet_express
   # Initiate the virtualenv python etc.
   $ . env/bin/activate
   # Run the development server.
   $ python planet_express.py
   # ... and point a browser at localhost:8090/ (port set in planet_express.py)
   # Or
   $ ./console       # Interact with the database via python-sql stuff.

 It wasn't finished; I'm not sure how many of the routes worked.
 (I thought more than I'm seeing ... maybe the 'localhost'
 is failing for the path stuff?)