Your Data vs. Your Server
This post was originally published on my Postach.io blog on January 13th, 2015.
Most of my blogs over the past few years have not been actually hosted on jeffreykishner.com. The blog you’re reading right now is hosted by Postach.io
, and I am using domain mapping to make it appear at thoughts.jeffreykishner.com. My previous blog, blog.jeffreykishner.com
, is actually hosted at GitHub Pages
, and I am once again mapping the blog
subdomain to GitHub.
The CMS (content management system) for this blog is Evernote
. (Postach.io “reads” any notes that are stored in the Postach.io notebook in my Evernote account, and decides what to do with each note based on the tags I use.) Although Evernote is considered part of the “cloud,” all of my Evernote content is downloaded to my home iMac. The CMS for blog.jeffreykishner.com is Fargo.io
. Fargo stores all of its data (rendered in an outlining markup language called OPML
) in Dropbox
, and renders the blog as HTML files. I publish these HTML files to GitHub Pages, but they can also be published on a subdomain at smallpict.com, or at one’s own domain via fargoPublisher
. As with Evernote, all of the content I store on Dropbox is also downloaded to my home iMac.
In both of these cases, I don’t “own” my server, but I do own my data. I also own my domain name, and I am mapping subdomains to various hosts so that they appear at *.jeffreykishner.com.
If Postach.io were to go out of business, it would not be the end of the world to reconstruct my blog, but it would be inconvenient. I’d probably have to go through my Twitter timeline to reconstruct the permalinks of all my blog posts, then go into Evernote to find the content associated with each blog post. I could more-or-less reconstruct the posts from the HTML files downloaded into my Evernote client, or by rendering the text in my Evernote files as HTML. Then I’d have to create a thoughts subdomain and upload all of those HTML files, so that the permalinks would be the same as they were when Postach.io would have been in business.
In short, as long as I own my domain name and I can access my data on my iMac, I feel secure that I can reconstruct any lost blog posts.
As example of owning one’s data and
server would be a self-hosted WordPress blog. (You can find my Wordpress archives at my Previous Writing
page.) This involves uploading a bunch of PHP files to a server that I rent, creating a mySQL table, and making some changes to a PHP file. Nobody can take this blog down. Even if the company that owns WordPress goes out of business, my WordPress blog would still exist. And my data would still be safe, albeit stored in mySQL tables. (I use WordPress for Sasstrology.com
, and for peace of mind I subscribe to a backup service called VaultPress
. Otherwise I would feel compelled to occasionally download my mySQL tables to my iMac for safekeeping.)
An example of owning neither one’s data nor one’s server would be Twitter, Facebook, Tumblr, etc. Any of these services could disable my account, and I would lose all my data. The same goes with blogs hosted on WordPress.com or Blogspot.com.
Some very brave souls do static blogging at their own domains. They use some kind of script to, for example, convert a text file written in Markdown into an HTML file, and then upload the file to their server. The advantage of this is that there is no reliance on a database – like mySQL – that can be corrupted. fargoPublisher does something like this, in that it renders an OPML file into a bunch of HTML files (including index files for a directory structure).
My current comfort level is with data that I own, even if I don’t own/rent the server, as long as I can use a subdomain of jeffreykishner.com.
How do you feel about having control of your data and/or your server?