Managing my S3 WebSite

It is a good security practice to use the least amount of privilege required. So I followed Keita’s instructions on granting user access to a S3 bucket to create a new user that can only access a single bucket in S3. Because my AWS usage is only S3 and Route53, I also gave this user access to AWS billing. That is a minor violation of least privilege, but the increased convenience well worth the small decrease in security.

There are plenty of options for copying files to S3. It is important that your choice sets the content-type correctly. I use Forklift from binarynights to copy my files to Amazon S3. If you add the site as a Favorite from the Favorites pane, then you have the option to save the secret key. Add S3 Favorite in Forklift

Tin Foil Hat

Just because you’re paranoid doesn’t mean they aren’t after you

Joseph Heller

Maybe I shouldn’t advertise, but I have definite tin foil hat tendencies. So I’ve been following the Heartbleed Bug pretty closely. This is scary stuff. An important piece of infrastructure that we all trusted turned out to be untrustworthy.

Security conflicts with convenience. That applies to users and developers alike. Users make their own choices and take their own chances. But developers make choices on behalf of their users. They need to have an appropriate level of paranoia on their behalf as well.

I'm Back

I needed a new home after my two lifetimes ran out. I quickly realized that I had some problems:

  1. My eyes were bigger than my stomach - I kept wanting more.
  2. I’m cheap - I didn’t want to overbuy.
  3. It’s a snake pit out there - the affiliate programs, sponsered links and aggressive SEO obscure the truth.

Now I’m back and better than ever, courtesy of Amazon S3 and Route53. It was the best combination of high quality, small scope, and low cost out there. The pay as you go model just works.

If you want to do this yourself, here are some helpful links:

  1. Amazon’s Hosting a Static Website on Amazon S3
  2. Chad Thompson’s Static Web Hosting with Amazon S3

The End of Innocence

You don’t buy an apartment to make money. You buy it to make a home, a place of your own.

I didn’t become one of the TextDrive VC200, add the Mixed Grill and a lifetime accelerator to save money. I joined to establish my home on the web. A place to which I could always return.

That’s part of what TextDrive took away from me. I thought I had a home for life.

Other members of the VC200 may have received good value on their investment. But I never had more than one website and I never launched a live application on my accelerator. But I always knew that I could.

I suppose that I should have known that commercial lifetime isn’t the lifetime of the seller or the buyer, but the lifetime of the product being sold. But I somehow thought it would be different this time. That a medium in which people were brands and reputation mattered would be different.

That’s another thing that TextDrive tried to take away from me. I’m hoping that they failed.

Hello, Hexo

Time for a new beginning. “Take the First Step” is now brought to you by Hexo.

Why Hexo?

It is built in node.js. I’ve been meaning to start exploring node.js and JavaScript. A journey of a thousand miles begins with a single step. Here is my first step.

It generates static sites. My hosting situation is in flux right now. Static sites are fast, lightweight and portable. Have site, will travel.

My search quickly brought me to Static Site Generators, the self described “Definitive list of Static Site Generators”. Sorting the list by language narrowed the list from 237 to 39. Looking at github commits and contributors narrowed the list to two: Assemble and Hexo. I really wanted to like Assemble, but it was just too ambitious for my simple little weblog. So I went with Hexo.

Two Lifetimes

Well, those two lifetimes went by quickly.

My first TextDrive lifetime lasted about nine years. That ended in disappointment as parent Joyent discarded TextDrive and its obligations for lifetime hosting and lifetime accelerators (VPS).

My second TextDrive lifetime lasted just over a year. That ended on Friday when I noticed my weblog was down. A visit to TextDrive yielded this:

As anyone looking for decent support or even useful information over the past few months can attest, the revival of TextDrive has not been a success.

What began in mid-2012 as an exciting challenge fuelled by good intentions and lean resources quickly turned into a cleanup project with almost no resources.

It is disappointing to report that after a year and a half of uphill battles and unimagined setbacks, after several costly efforts to regroup and find another way, options to keep TextDrive growing have run out, and we will cease operations on the 14th of March, 2014.

For those who wish to know, details of what went wrong will be made available once shutdown operations have completed.

Sorry to have let you down.


I had several months warning the first time around. This time I was completely unprepared. Fortunately, Jacques Marneweck at Kaizen Garden has stepped in to help recover data. Thank goodness for the kindness of strangers.