Part 1 Continued - Saving to mongodb

I decided to build an online course aggregator and introduced the project in this post, and last week I started part 1 explaining the backend and creating the first provider class for coursera. This post I will dive into the storage engines, initially I created two storage engines; mongodb and postgres using sql alchemy.

Part 1 - Creating a python based backend

I decided to build an online course aggregator and introduced the project in this post. This post is part 1 of the series and I will break down how I build the backend aggregation framework. The whole point of this project is to build a full stack application as both a learning exercise and an in-depth blogging topic.

MOOCs are Massive Open Online Courses, and I have had the time to take a few of them. These courses have probably been the best college courses I have taken of any kind, in terms of how much I learned and have been able to apply to my day to day work. This makes sense as I have been able to pick courses that interest me or apply specifically to my job versus my normal course work where the college decides what I need. The other reason they have been great is because they are provided by some of the best universities or professors in that topic.

Everyone in the world was programmed by the place they were born, hemmed in by their beliefs, but you had to at least try to grow your own brain.

Scott Westerfeld, Pretties

My VPS service from myhosting.com is up for renewal, and honestly it isn't worth the money anymore. I only use it for my blog, and it has been a year since I have posted anything. I also used to host other websites and my private github repos, but now I have a paid github account and no longer host those websites. I can also spin up an Amazon server if I need to play with around with the cloud. Github will host static html sites for free and allows for custom domains. The problem is a blog is dynamic, but that is where Jekyll comes in.

I am starting to think I am at a crossroads in my career.  I love coding, and more specifically I love solving problems with technology.  I love the team I a m on and the role I play on that team.  When I think about where I want to be in 5, 10 or even 20 years down the road I can't see myself in engineering.  The ladder in engineering still goes higher than the rung I am currently on, but it climbs up to the rooftop of a building I don't want to be on.   The engineering ladder leads to senior engineers, lead engineers, and engineering managers.  It's still all engineering, and I want to keep learning m ore about the process, more about marketing, more about sales, and more about finance. More importantly, I hate the fact that I have very little strategic in put on our product or what I am building. The problem is in order to change ladders I will have to give up coding. Maybe I am ready to do that maybe I am not , but the opportunity to jump to another ladder is opening up now.

At my company several Product Management jobs are opening up at various points this year and it is making me consider making a switch.  Product Management gives me the opportunity to work with engineering, but also with sales, marketing , and finance.  The Product Management ladder is different then the engineering ladder in that it forks off in many directions instead of going straight up.  This means that I could go a number of different directions as I climb, and in 20 years I can really be anywhere in the organization (except for m aybe HR, fine by me).  The best part of a jump to product management is that I still get to solve problems with technology, and that could ease the pain of quitting coding cold turkey.  Have any of you faced similar crossroads? I would love to hear some thoughts in the comments.

Sulia.com is a good concept and a decent website.  I hate the oversized nav links at the top but other than that I have no other gripes about the site.  However, many of my favorite journalists use it to post thoughts and analysis when 140 characters isn't enough.  It seems like twitlonger but i t is actually more useful to posters as it categorizes the post by topic and allows people to view content by topic.  

The problem is that sulia pays the content creators for clicks, and that tempts the creators to post everything on sulia and simply post a link to sulia on twitter.   I hate twitlonger links in my feed as well. I would prefer a 140 character teaser with a link to a blog in the tweet, because I mostly use a mobile device fo r twitter consumption.  The worst behavior is when the journalist posts a short statement (less than 140 characters) and a link to their blog on su lia, and then tweets a teaser with a link to sulia.  THIS IS A HORRIBLE THING TO DO TO YOUR READERS.  Unfortunately, many of my favorite journalist (Denverpost.com I am talking to you) do the double link sulia post several times a day.  I want to reply to all sulia tweets begging for them to stop, but I respect these writers and don't want to be rude.  How can they use sulia more effectively?

A few months back I posted that I wanted to away from PHP for web development and started learning some of the major MVC web development frameworks. The first one I tried was Pyramid, and I wrote about creating a blog engine with Pyramid and MongoDB. This was an awesome experience and I loved using it, so next I wanted to try using Django.

Django's Strengths

You can find many good comparisons on the web. So these are my thoughts after build a blog engine in both systems. Getting started with Django was quicker and easier than with Pyramid because it had everything built in and had much better documentation and tutorials. I loved the kitchen sink manage.py utility vs several Pyramid commands like pserve and pcreate. Also the settings.py was easier and had a better structure than Pyramid's ini files.

My wife's brain works very differently then mine. She is very emotional and I am very logical, that is probably very common in any couple as it probably a Mars/Venus thing. She worries about things that are very unlikely to happen, often asking me questions like "what if the house burns down when we are gone". I typically respond with a snarky reply like "what if a plane falls from the sky and lands on top of our car". I usually don't think about things that are statistically unlikely. Even things that are more likely: cancer, heart disease, car crashes, etc don't bother me. I honestly just didn't worry or fear much of anything.

All that changed last January when my first child was born. The though of anything happening to my son, not matter the odds, scares the living shit out of me. It is irrational, illogical, paralyzing and uncontrollable. I also now worry more about myself because the thought of leaving my son without a father scares me as well. This last week the fear and emotion have hit an all time high because of the horrible story that is unfolding in my backyard. There is a triangle of key locations in the case that I spend most of my days within. In fact, last Friday just an hour or so after the kidnapping, I was at the dog park with my two golden retrievers, less than a mile from where she disappeared from.

Two nights ago when a body was discovered, near our house, I couldn't sleep. My wife and I started discussing how we would try and protect our son from this type of danger. The though of something like this happening to our son is paralyzing, no matter how statistically unlikely! This is the part of becoming a parent that nobody warns you about. Everyone tells you about the joy (which is true) and how it changes your life (very true), but nobody said one word about the all-consuming fear. My normal defense, logic, has no chance against this new brand of anxiety.

I hope and pray that the monster in my backyard is captured and brought to justice!

In my previous post about creating a blog with Pyramid, I mentioned that I really didn't like Chameleon or Mako (To be fair I never really tried Mako; I just read through the documentation trying to understand how to perform inheritance) templates. A commenter(thanks Matt) suggested I give Jinja2 a try. Some of my gripes (or my lack of understanding) about the those template engines were:

• I couldn't find an easy way to break my templates into various pieces like header, footer, sidebar etc. Then have a layout template that pulls them in, and lastly inserts my individual view on top of that portion of the template.
• The syntax should be easy and I believe template portions of your markup should stand out from the static markup.
• I don't have any concern over speed or jython compatibility and this seems to be a major focus for chameleon.

Chameleon uses a macro language called METAL and it adds a bunch of extra markup to your document and making it less clear over which parts were part of the template and which were static. This became apparent when I was retrofitting my templates to Jinja2 and ripping out the chameleon parts, I kept finding remains of the chameleon templates after I had made the switch.