On Cloud Repatriation
If you’ve stumbled across this post using a search engine, you might be wondering what this cloud repatriation is all about. Well, to us it’s about trying to build a locally hosted alternative to placing all of our data in the datacenters of a few very large corporations. Why are we spending our time doing this? you might also wonder. Well, there are obvious economical aspects, but I, for one, have always just liked the thought of running my own basic information infrastructure, controlling my own data, etc. It probably boils down basic desires of self-sustainability and to know exactly what is happening with my data. And I’m certainly not the only one feeling this way. Others are sceptical of leaving all their data, and the control of what is done with it, in the hands of corporations that in reality answer to no one. Or they also just like to tinker and run stuff themselves.
We got rid of the server rooms, but do we really want the mainframe back?
The initial promise of the Internet was, to at least some of us, for it to become a global, distributed, decentralized network of independent peers exchanging data using open protocols. What we have ended up with it seems, is exactly the opposite. This most likely will not change substantially for decades, if ever. There is simply no real movement in the opposite direction, no local software offerings of clearly higher quality than the services offered in the cloud, no real regulation in sight, no vocal opponents, in essence no real barriers. This has consequences for the data, information tools and social life of all citizens, but for the purpose of this post, we’ll look at what this means for plain, regular businesses. Following the path we are currently on, the logical terminus will be the transformation of all IT departments into support and call centers for the various SaaS-offerings their company has signed up for. If, for whatever reason, that does not sound attractive to you, read on for a look at some of your options for running stuff yourself.
Anyone studying the history and rise of the centralized Internet, I think would agree that one of the reasons SaaS is winning big over locally installed services in corporate server rooms, is that it is generally just higher quality, which probably just boils down to economy of scale, and the complacency of the enterprise IT industry, which by the way is still earning big bucks. With SaaS, a service can be purchased and put to use instantly, while all the issues of having an IT department installing and running stuff locally just disappears. There really is no way for traditional software to compete with that. I, for one, am not shedding any tears for the likes of Remedy (obviated by Zendesk). One problem of course is, that with SaaS we generally do not end up with a healthy eco-system of competing peers, who agree on some level of standardization, like say the auto industry (fuel standards, interchangeable wheels, tires, etc.), but generally end up with a single provider of a service, and your data and you being stuck with that provider. The Internet has multiplied the problem of the winner-takes-all dynamic of a global economy. However – that’s just the way it goes. Some of us just keep working in the opposite direction of all this massive centralization, just because we can, because it’s interesting and fun, and because a society really should not put all its eggs in one basket, especially not someone else’s basket.
Can’t we just have good, basic, digital infrastructure, without giving up basic freedom?
So, with all that in mind, what basic infrastructure does a regular small to medium-sized business need to operate? The mainstream narrative is that digital natives, only armed with hastily assembled FaceBook and LinkedIN pages, are able to somehow plan and coordinate complicated work, communicate with clients, send out invoices, pay their employees, etc. That is of course not the case – they end up using and paying for a pile of SaaS offerings, which is fine, but are there alternatives, at least for the basic stuff? Well, actually a lot of good options exists for running the information infrastructure of SMB’s with software that is both Open Source and of pretty high quality. Products like Ghost (for running a blog), Discourse (for running a discussion forum), Zimbra (for running your own email, calendar and directory infrastructure), FreeNAS (for running your own file server), Rocket.Chat (for running your own chat/messaging), NextCloud (for running your own file synchronization service), Asterisk (for running your own telephony service), Gitlab (for managing your code) and many others exist, and are generally a joy to set up and use.
So why on earth would you e.g. fork over $480 a year to Squarespace to run a simple e-commerce website, when you could just download and install WordPress and a few plugins for free? Or upload all your precious code to Microsoft-owned Github to the tune of $480 a year to (if you are a team of 10), when you could simply just download and install Gitlab? Or a pay whopping $720 a year to Google for data mining all your sensitive emails, internal and external, when you could just download and install Zimbra and NextCloud? Well, part of the answer to that is of course, that all of this wonderful software requires a higher tecnical skill-level to install and use, than the SaaS offerings. But another important part is, that it also requires somewhere to run and store data. In other words, it requires infrastructure. If this infrastructure is a datacenter run by Amazon, Google or Microsoft, you may rightfully ask, why then bother? Why go to all this trouble when all you have achieved, is replace one data overlord with another?
Well – that is exactly why we built Stabile. SaaS will eat the software industry, IaaS will eat the server industry – we’re not trying to change that. We built Stabile for those who, like us, still like to tinker, hack and run our own basic infrastructure. To try to make it just a little bit easier and perhaps even fun, to set up and run your own infrastructure and even more importantly, have a say in how it is built and what is done with your data.
