I considered changing this sites’ platform for the first time in a decade.

Up until 2006, I ran my own homebrew CMS. Many of my links and older posts still reference this, and I have many, many shims that still hide under the works- even if things such as my software have been unavailable for nearly the same time. I decided to discontinue work on the CMS, as it had slowly become bloated (although I made some great features like “automatic listening-to” blogging, a “things that interest me” bookmark blogging tool, and more), but I made it very difficult to re-theme. This is how it looked without CSS abuse- with CSS intact, it looked quite a bit like the MacOS 10 Finder threw up.

I ended up settling on TextPattern, which not only had a neat new way (as of 2002) of doing things without needing to use inline HTML, called Textile, of which Markdown borrows quite heavily from.

Long story short, my site is really looking dated. I lack the enthusiasm I once had to rebuild it’s powerful design into the oversized pretty blue clouds and huge friendly fonts that adorn pretty much everything, with parallax scrolling stock photos beneath.

I started off this new journey with HTMLy, which didn’t foot the bill as although designed for a flat-file build, it just wasn’t to my liking. I’m looking to simplify and beautify. GRAV is an impressively large YAML parser, but it requires quite a bit to setup as a blog- it has a much better design flow for single pages, and does anything BUT simplify.

I’m sticking with good old TextPattern- it’s served me well, and continues to do everything I want, except it looks dated. Maybe I’ll break down and re-theme it one of these decades.

Everything, and I mean everything is either on the “cloud”, moving to the “cloud”, or promoting the entire concept- even if it has nothing to do with it. As evidenced by Kodos (from way back when the Simpsons was still watchable):

I’ve touched on virtualization a few posts ago, so I won’t go on about the means or implementations here.

The OS movement has so many options available to make virtualized systems at the press of a few buttons. There is OVZ and LXC for Linux, KVM for Linux, NetBSD, and Solaris/IlluminOS/SmartOS, and now, OSV comes up with _an entirely different design. This is a new introduction to the party, based on FreeBSD.

It’s already setup to run on EC2, GCE, or even your current machine. as evidenced here, it’s incredibly fast to create an instance, and the most commonly used HTTP/2 development tools are already setup as containers.

I’ve only played with this for a few hours, but it seems to foot the bill for many purposes:

• No overly complex zone setup, which is what keeps many new developers shy away from SmartOS- it has a simple REST API.
• ZFS support from the get-go. No need to muddle around with the 40,960 different filesystems available for Linux. ZFS it is!
• Small Footprint- I did a simple task deployment, and it added about 20MB of itself to my entire task. That isn’t bad, considering that is the entire virtual machine!

One of the problems plaguing our technology-biased society is that electronics have become disposable. We are at a point that anyone who has a smartphone that is more than a year or two old is viewed poorly in certain circles. I do not run with that crowd.

You may not be aware that the Nexus 4 is a wonderful Android phone which was released less than 5 years ago. It is available with a still-impressive 16GB base, has an 8MP camera on the rear, and a 1280×768px screen.

You can find them virtually anywhere from prices of $40 to $80, depending on your luck and how scuffed they might be, partially in thanks to Google no longer providing an “in house” Android software revision after 5.1.1.

Thanks to the Android core being open sourced, there are Developers and Designers keeping this still-useful phone alive.

My favorite community-built ROM, Pure Nexus, has an unofficial port of Android 7.1.1, with security patches to-date. This brings this phone nearly to the standards of the latest and greatest- at a fraction of the cost.

I’ll be the first to tell you that I’m a big fan of the Open Source movement- I’ve built a career out of supporting “free” software. For a brief flash in the pan of the late 90s and early 2000s, most technology based corporations saw the possible future of Linux, and embraced it wholly. It’s not quite the same today..

What’s my point? Some corporations have assisted Linux development by open sourcing specifications and code samples for kernel and driver writers to add support the product with Linux. Others have offered no assistance at all (requiring a “clean-room” reverse engineering method for support, requiring a lot time and dedication), and yet others have “sparingly” supported Linux, by making their own code and releasing it pre-built to be linked to, uploaded, or otherwise sent to the device.

I am speaking of the infamous “Binary Blobs”. Committing a blob is premature obsolescence.

The first time I ran into this design was when I was stuck with an ATI video card, which forced me to use an ATI build of X11 binary so I could get my video above 800×600, and a half decent rescan rate.

This design is problematic. In the case of my video card, it just won’t work properly anymore (such as evidenced by these instructions to get “blobby” builds of X11 to work with recent versions of Debian). These old, unsupported bits of software are an enormous security risk, as they are no longer updated, supported, and won’t work well (if at all) with recent distributions.

Intel has been guilty of the same- notably with their wireless devices in the past, and formerly, their onboard video. If I had to point a finger at any specific piece of hardware which is overwrought with blobs, it would be wireless devices.

OpenWRT, and LEDE have their hands tied for new device support because of these binary blobs (DD-WRT has no problem blindly throwing blobs at devices, which is why it is so popular, and supports so many more devices).

OpenWRT-cum-LEDE have decided they’d rather have proper support for devices, so they can integrate it in their current and future builds. LEDE is currently based on a recent 4.x kernel, whereas many RealTek, MediaTek, and RALink device blobs were written for (and supported by) Linux 2.6.x to 3.x. Quite a bit has changed under the hood since then. Developers are forced to work around these old designs whenever possible, which aren’t often capable of supporting recent enhancements and security implementations.

When I’m looking to purchase hardware, if I find these blobs available, I’m just as likely to pass on the device whenever possible. I suggest everyone do the same- and support corporations that support our hardware without hindering the device with ineffectual commitments.

I don’t really have much of interest to share at the moment- we’ve got at least 8” of fresh powder outside, so no work has commenced on the various car projects this weekend.

I’ve been following DragonFlyBSD from the sidelines for nearly a decade. It’s a fork of FreeBSD 4, when Matt Dillon decided he didn’t like where things were moving.

His HAMMER filesystem is intruging, and I’ve been looking for an excuse to play with it for awhile- but until recently I haven’t had access to hardware worthy of this hungry filesystem.

I was able to get a promo 2TB drive with a lower-end colo, and have been using it as my offsite non-cold-storage backup. So far, it’s been working great- I rebooted it once in the middle of a larger operation, and it recovered quite nicely. I don’t know if I’d use it for anything super-important, since there aren’t many tools for recovery should an actual disaster occur- but for my silly little toys, it’s more than sufficient.

I promised you pork. It’s in your clam chowder. Enjoy!