Side note – I do not like Mysql. Drizzle is cool. Postgres is pretty awesome but Mysql – not a fan.

And now to the snark (note there is cursing)

That last point had me looking at the various cloud solutions. Most of the cloud solutions are pure backup solutions – as in you delete it off your drive, and 30 days later they delete it from their archive. That’s the Mozy and SugarSync way. This was not what I wanted. I want persistent storage!

So now my thoughts turned to Amazon S3 and Google. These are the two heavy weights in the arena. S3 has the aging, but still great, Smugmug front end (for my photos). However, since I shoot raw, Smugmug would end up costing me an arm and a leg. As I looked at a few more S3 front ends , Google came in with some amazing pricing for Google Docs storage. Since I’m already in bed with Google (Google is Mother, Google is Father) this became a no brainer. Then I learned just how painful Google’s file uploader is for massive amounts of files (hundreds of thousands). To put it mildly, Google Docs, with default uploader, is not usable as a “cloud drive” in any way, shape or form.

Enter Memeo. At first I loved the elegance of their solution – however I’d need to subscribe to Google’s business package and I’d need to keep paying Mameo every year. Extra bonus, I couldn’t use my existing account (at least I didn’t see a way). Ok, this was not a solution since I just bought a ton of space on google. Memeo, for all it’s possibilities, lost me on cost and ease.

A few more searches led me to Gladinet! For the most part I could not be happier. Gladinet allows me to map a logical drive to my google docs account – giving me pure drag and drop (woo hoo!). But here’s the cool thing – incremental backup! And I can configure it to not delete the files I’ve deleted! How awesome is this?! It’s a one time price that is more than worth it. Quite simply, Gladinet works and solved my cloud storage problems elegantly.

This isn’t to say that Gladinet is perfect – I have had issues with my incremental backups. If my “pending tasks” hits 60k+ and a problem happens (for example, I’ve deleted queued up files) the program will get confused. Additionally, there isn’t a way (that I know of) to purge or reset a queue. Overall, these aren’t huge issues. I’ve found decent work arounds (I schedule my backups for 2 days a week, I have several scripts instead of one – that seemed to solve 90% of the problems).

Overall though, Gladinet is a killer application for google docs. I can’t imagine using Google docs without it. There is a free trial, and if you use Google Docs, you owe to yourself to try out Gladinet – you will not be disappointed.

The Chocolate Tart makes a darn good cake

In other news; I’m not surprised Oracle sued google over the Dalvik JMV. I am willing to bet that Oracle will lose – 5-10 years from now. Meanwhile I am rethinking my recent embrace of Java. It’s decent, I will use it for my android programming however I’m really not excited about where Sun’s been taking it and, from the looks of it, Oracle isn’t going to take it anyplace better. I think I’ll put my spare cycles onto other fun and interesting languages: like Python or Ruby or G.

The database markets are saturated. There are a ton of companies fighting over the same spaces: spaces that is getting served by a lot of open source projects and/or by the major appliance players. NoSQL and cloud are changing the rules. The appliance companies are trickling down with software solutions to match their screaming hardware (or, in the case of Oracle, trickling up). The landscape is changing and companies not playing with heavy hitters are doing the work themselves. And finding that up to a point, it’s really not horrid. That last part is due, in no small measure, to companies like Facebook and Google releasing their code to the open source community. Additionally programs like Drizzle, I believe, are already creating a compelling argument that open source projects can supply 90%+ of a database user’s needs with lot less pain than a few years ago. Add in memcache and, well, I don’t envy any database company targeting data sents under a few hundred gig.

For any new database company to make it, to me, they need to offer some compelling, game changing ideas. I’ve heard a few bantered about – and some are neat. However game changing? I haven’t seen it. VoltDB? Infobrite? Neat but not game changing.

I really need to stop writing about database companies and start talking more about my awesome new vocational field – robotics.

And now for something completely different. I love lego. I love Ted. Together you get this wonderful talk.

Having worked in a cloud based company, I realized that someone must have a cheap solution to my problem. I quickly started looking at the prime players: Amazon’s S3, Google Docs, Carbonite, and Mozy. I quickly removed Mozy and Carbonite from the list as I wanted incremental backup without deleting files I delete locally (I wanted true persistent storage). This left Google Docs and S3. Actually, Skydrive from Microsoft was also an option, but 25g (even though free) just wasn’t enough for me.

Comparing prices and who I’m already deeply entwined with, Google docs won. For $100 USD I get 400gigs of online storage. That’s a hard price to beat. The only issue left to me was the interface. Google docs is not designed for exactly what I needed but it does have a fairly robust API.

Thankfully I didn’t need to write the app I wanted, gladinet already exists and it was cheap enough to qualify as an impulse buy. So far, it meets all my needs and then some. It took a minute to set up. The interface, while not perfect, is easy enough to figure out. And it’s not a resource hog. What’s not to love?

Boston Common Statues

A happy baby

My son loves to smile. Earlier today we were explaining how much we love him (this much) and as you can see – he really enjoyed it.

Sometimes, all you need is a smile from a wee one to be truly happy

The first is that process has to make the product and the path to product get better. A lot of people share this mantra with me. The trick is to figure out what this means, per company, and how to apply it (or remove it). This is something that I’ve played with at almost every company I’ve been at over the last 10 years or so. There aren’t silver bullets and there is trial and error.

Another thought that I find myself frequently coming back to is that at the end of the day, seeing an actual piece of hardware as your product is more fulfilling, to me, than something in the clouds. Don’t get me wrong, I love all the great work going on in the clouds – and I plan on utilizing it more and more (hello offsite storage!) but for me, at this stage, tripping over robots is pretty damn neat. And more satisfying. I guess you could say my years in database helped me appreciate embedded development.

Now that time is sort of returning to normal, I do plan to return to posting more thoughts on management and a lot more thoughts on robotics and AI/Machine Intelligence. For example, this site has gotten a lot of my free time attention of late – aigamedev.

In fact, I don’t think I would have completed my grad work without him. Prof Schmoltze was always ready with a quick smile, a joke, an insight or an idea. He, more than anyone else, inspired me to concentrate on AI for my masters work. I barely knew him during my undergrad – but I do remember a few hallway and lab conversations with him. He cared.

Goodbye Jim.