I’ve had a lot of fun over the past few weeks playing more and more with cloud services and seeing how they can extend the functionality you normally get with a shared hosting plan. I blogged about my adventures setting up Amazon Cloudfront as a CDN (Content Delivery Network) using W3 Total Cache (which is one hell of a plugin the more I play with it). Tonight I decided to really mix things up and get a Rackspace account to see how the services differed and what they had to offer.

This is certainly not an ad or evaluation of Rackspace as a service but I want to just put this out there: I’ve already gotten a personal email offering to connect me to human beings with names and phone numbers and someone on their team reached out on Twitter with some help when I mentioned the URL structure for their CDN being extremely long. Needless to say they’re off to a killer start just in terms of customer service. W3 Total Cache supports using Rackspace as a CDN in the same way that they support Amazon CloudFront so I won’t rehash how to go about setting that up, read the post here and you’ll have a decent idea of how that works.

As a quick recap the purpose of a CDN is that your blog likely has a decent amount of images in addition to CSS and Javascript files that get displayed when a user visits your site. These are static files and typically served from whatever server your hosting company is on (in my case Hippie Hosting‘s server is in Canada) and geographic location makes a difference in speed. A CDN takes a copy of every image in your media folder, all your theme files, and any other custom files you’d like to serve up and puts it up on their cloud storage which gets replicated to servers all over the globe. So someone in Australia isn’t pulling images from the server in Canada anymore, they’re getting them from a server right there in Australia and hopefully notice quite a speed bump. All of this is fairly automated using W3 Total Cache. Give it your Rackspace authentication info, upload all your files one time, and any new files you add (like the featured image of the post I’m writing) get mirrored to their server and served up using their CDN. If I ever turn it off it will switch right back to being served locally from my blog.

Not only is this fairly easy to setup but Rackspace is dirt cheap. $0.10/month will get me 1GB of storage and they don’t charge anything extra for the bandwidth to upload, download, serve via CDN, etc. I like their pricing scheme a lot more than Amazon in this regard because it’s very straight forward. Up to 1TB is $0.10/GB with a few different tiers to serve massive amounts of files (you can see full pricing here). All my blog images and files amounted to just 200MB so for one freakin’ dime per month I get the functionality of serving a copy of all images and files on my blog as close geographically to the reader as possible. That’s pretty awesome!

The other service I had wanted to try out was using Rackspace as a DNS server. I briefly played with this on Amazon which costs $1/month for 1 domain but I was pleased to see the service is actually free on Rackspace. Why would you want your DNS to run on another server? Speed is one benefit in some cases but the biggest advantage is reliability. If Hippie Hosting goes down I can easily point the DNS somewhere like a Tumblr blog to keep everyone up-to-date on what’s going on. Now the downside is that any changes I make to my cPanel account I need to manually reflect on Rackspace so if I create a subdomain in cPanel I need to create an A record in Rackspace. That’s a tradeoff I’m ok with though because the changes are almost instantaneous once your Rackspace nameservers have propagated and I don’t edit my DNS often enough for that to be a major concern.

So yeah, I’m flyin’ high in the clouds and I’m loving it. I’ll still keep my Amazon account around for EC2 access and some other things I like to do over there but it’s fun to try out these different services to see what they have to offer and Rackspace is top notch in that regard.

I have to confess I’ve been vaguely following the GitHub scene for awhile now but always considered it a place for hardcore programmers. I had heard that people were using it for other interesting things, most recently last year when App Dot Net put their Terms of Service out there and encouraged folks to fork it and offer their suggestions for modifications. When Alan posted recently about moving Feed2JS over there I decided to follow his lead and start actively using my account to explore the space. I’ll second his recommendation to go through the free CodeSchool lesson on Git to get a general feel for the terminology and how things work.

One model that I really like is how the CUNY Academic Commons and CHNM have their organizations on there to publish code they’re working on. On our best day DTLT could hardly be described as a group of programmers (more like self-described hacks) but I’ve no doubt that we’ll likely have something to contribute what with all the little snippets that Martha writes for WordPress to interact with various plugins and themes. To that end I setup a DTLT organization on there and added Jim and Martha (and yes, this is a callout to Andy Rush that he has to get an account. Get on the bus!). For our first stab at a repo I decided to see if we could use it as a collaborative space to build a proper Terms of Service for the Domain of One’s Own project. DoOO (how do you like that for an acronym?) is an interesting project because it centers around openness and portability of your data. And of course most TOS exist solely to remove certain rights from users. I want to believe with the help of others we can craft something that marries the two ideals. I’d encourage anyone interested to fork the project. This is just one example of a case where you don’t have to be a programmer to help out.

The more I play with the tool the more I start to get a feel for how the platform could be used similar to a wiki to allow versioning and collaboration for writing and research with a strict focus on sharing. I’m experimenting with having a living CV in a repo that I can use to track changes over time. I’ve only just begun to dip my toes in the water but I did my first pull request on Feed2JS today and the more I play the more confident I get with the possibilities.

A few weeks ago when Google announced the discontinuation of Google Reader effective July 1st what began as a moderate approach to reclaiming things here and there that were important to me was sent into a tailspin. Maybe that seems dramatic but the Reader shutdown really hits home to me possibly because it’s one of the first services to shut its doors on me that I personally used every single day, often several times a day. It’s hard to trust a company after something like that and more importantly it raised larger issues about what the RSS landscape looked like before a heavy like Google came in and wiped out the competition only to bow out now. I had briefly flirted with the idea of getting off Google servers but now it was personal. So here’s what I’ve done in the meantime:

Fever

Their have been tons of blog posts detailing all the various Google Reader alternatives out there. Most of them feel like switching from one proprietary web platform to another. The idea of hosting my own solution appealed to me and paying a nominal fee to a developer I’ve admired and respected for years was the icing on the cake. I’ve been using Fever since the day the shutdown was announced and it’s been absolutely great. I save items I enjoy which feed into a widget at the bottom of this blog. If there’s one thing I wish it had that would be a “mark as unread” feature since I used that previously to keep posts available to me. I’ve been trying to retrain myself to “save” items I want to come back to which works ok but not as well. I will say I’m a bit hesitant to how well Fever will be supported given Shaun’s blog post about the current state of it but regardless it’s working well now and since it’s hosted in my space I can use it without worry of a company attempting to monetize or close its doors on me.

Mail

Gmail was always expected to be one of the hardest ones for me to give up. I’m not sure why that is, I’m certainly not a power user of many of its features and most of time I’m checking my mail through a client with its own featureset anyway. I connected Apple Mail to Gmail over IMAP to download all of my email and then setup a mailbox on my hosting server, connected that to Apple Mail, and dragged it all into an Archive folder there. I setup a forward on Gmail to my new address, changed the contact for as many of the services I could think of (that’s an ongoing thing), and never looked back. I have to say this was one of those switches that I expected to be worse than it really was. I’m rarely using the web client but when I need to Roundcube on Hippie Hosting is pretty nice (and there’s a whole host of plugins for it I haven’t even explored much yet). The hardest part is getting people to know the new address but with all email from the Gmail address coming to me and my replies coming from the new address I’m hopeful over time that will self-correct (I thought about setting an auto-reply letting the user know of the change but figured that could cause real problems with mailing lists so I didn’t do it).

Google Docs

Switching away from Google Docs has been a lot harder. Getting an archive of everything I had was pretty easy thanks to Google Takeout so I grabbed everything and dropped it into an archive folder in my ownCloud folder. But the fact remains there are no good web alternatives to Google Docs. There are plenty of repositories but not many collaborative editing suites and none as robust. I’ve played with a hosted install of etherpad-lite which works ok but it’s not great and the server demands make it a non-starter for most people that don’t have their own dedicated box. For now I’ll have to keep my account active but with a focus on doing more with local documents rather than making GDocs the default environment for all of my document editing which was previously starting to become the case.

Google Groups

I subscribe to just a small handful of Google Groups which are basically listservs with a forum-style frontend. Sadly there does not appear to be a way to subscribe and interact with a group without a Google account. A few searches appeared to show some promise but the information is outdated and none of the things I tried work anymore. I get the digest emails to my new email address since Gmail forwards over but I can’t respond via email and if I want to interact with the thread I have to log in with my Google account. Ugh.

So yeah, a few wins, a few losses. Overall I can honestly say this hasn’t been as painful as I would have expected. Especially for something as big as email which I use so regularly. I think we often get caught up in the fear of the unknown and we let it paralyze us from exploring these alternatives. Google is our comfortable Lord and ruler. It’s been refreshing to notice I’m no longer signed into a Google account all the time and my life is no worse for it (and in many ways better).

Photo Credit: Alan Levine

Roger Ebert’s final words on his blog just 2 days before passing