Featured Post

RedScarf II Build Log

A 60% board with integrated numpad

More ›

Unlimited, Encrypted, Zero Knowledge Cloud Storage

I’ve been searching for a zero-knowledge cloud storage service for some time. For someone who is privacy and security focused, most of the common options like Dropbox or Google Drive go out the window right way. Both companies of course encrypt your content in transit, but both have at least some level of visibility into one’s content once it is at rest.

There are a few popular secure options out there, notably from SpiderOak. Their product offers zero knowledge storage using local encryption. I did some playing with the service, but I found the desktop software confusing and lacking the functionality I wanted. The mobile software was just terrible. No company gets mobile cloud storage right in my opinion - namely that I want most of the same file-system power that I get on a desktop.

The goal, as I see it, is to find a cloud storage solution that offers competitive pricing, and control over my content - specifically the ability to encrypt it prior to transit. If encrypt my content locally it won’t matter what level of visibility the provider has into the content - without the local keys it will be useless (aside from metadata, but that’s basically impossible to avoid being tracked/mined). It was also important for me to have good mobile clients, and fast sync between devices.

Clearly no single product or service is going to deliver what I wanted. The first step was Boxcryptor. Its functionality is quite simple: encrypt content before it is uploaded. Boxcryptor is compatible with most company’s solutions, be it Dropbox, Google Drive, iCloud, OneDrive, Amazon, etc. It can even encrypt arbitrary folders on a system if the cloud storage isn’t directly compatible. Boxcryptor is free for individuals connected to only 1 service. Business and multi-cloud-provider setups require a license.

In terms of managing one’s private keys, Boxcryptor offers two options: one can create a Boxcryptor account, and Boxcryptor will sync your keys across all your devices, alternatively one can create a local account, and your keys remain locally on your device, and are never transmitted by Boxcryptor to any other drives. Getting Boxcryptor set up on other devices requires copying that private key to the second device, then providing the passphrase (on desktop this simply requires copying the key over a flash drive, on mobile the key can be sent via a secured medium like Airdrop).

After Boxcryptor is set up, it will create an alias in Finder, the Boxcryptor folder behaves like a mounted volume, but provides contextual options for encryption when adding new content.

Encrypted content in the Boxcryptor folder is denoted by a green padlock. If Boxcryptor is not actively running, viewing encrypted files is not possible.

I initially played with Boxcryptor on iCloud Drive and Dropbox - performance was fantastic on using either cloud provider. I ended up deciding that iCloud’s sync was too slow, and that Dropbox’s client and pricing weren’t up-to-snuff.

I ended up settling on Amazon Cloud Drive (will be referred to as ACD going forward). At $60/yr for unlimited storage their product is a steal. The only downside to ACD is that no desktop sync client is available. ACD very one-sided. Amazon does have a desktop client, but its sole purpose is upload. Managing content and downloading it from ACD must be done from the web.


Enter the second step: ODrive. This is really clever software for cloud storage. It changes the entire thought process behind cloud storage in a way that really make sense to me. ODrive talks to numerous cloud providers and is itself a sync client. Most importantly: ODrive gives ACD a true desktop sync experience. Suddenly I’m not relegated to web-based uploads/downloads, I have truly unlimited storage at my fingertips.

What is so clever about Odrive is a very simple sync/unsync model. With typical sync clients all content from the cloud is synced locally as well. Normally that is what I’d prefer, but with truly unlimited ACD I’ll have terabytes in the cloud - I have room for that on my server, but not my laptop.

Obviously other providers client’s offer selective sync functionality, but that’s often buried in the preferences, with Odrive it’s a contextual menu. When folders are synced, their content is downloaded locally, when they are unsynced, a zero byte placeholder is present on the file system. This workflow makes it very simple to have certain folders always syncing everything locally, and others syncing nothing. With Odrive I can have the right content on the right computers, but I can always go back to sync.

ODrive also offers the ability to sync arbitrary folders back to ACD. I have a main repository for files that I manage much like one would manage any cloud storage. I also have setup sync for several folders on several drives on my server. I have the ability to sync all my content from all my machines now. With the content I’ve got, I’m easily reaching into terabytes of synced content.

Technically speaking ODrive is free, but for features like unsync there is a yearly license of $100 - a fair price in my book.

My setup is likely needlessly complicated, but I really like the level of control and simple day-to-day usability. I have Odrive on all my machines and have syncing setup so that each machine has the content it will normally need. Almost all the files are also encrypted with Boxcryptor.

I’ve only been running with this setup for a few weeks now, but I’m really pleased with it. The pricing is reasonable, between ODrive and ACD it’s roughly $13/mo. Obviously this is a lot more than the $8/mo Dropbox charges for 1TB, however, considering the power and amount of storage I’m getting, it’s a fair trade.

It’s good to be back into cloud storage. I avoided it like the plague when I became security conscious, but I feel like I’ve got a setup. With local encryption keys and a powerful sync client, I’m likely as secure and private as I could be while using cloud storage.

Typed on ErgoDox Test Board

iPhone 7 Ad

Great work from Apple’s marketing team. I love this type of work.

Typed on Octopage


John Gruber on the headphone jack:

Few companies other than Apple make decisions that they know will provoke outrage just because they think it’s the right thing to do.

Apple have done this time and time again. Floppy drive. Flash. Macbook.

Apple if anything are a company about convictions. Continually they make decisions on our behalf about what technology ought to be. Apple’s product decisions are patient and deliberate. Often they will hang back and see how the industry plays something put before they make their move, but when they see an area for genuine improvement they will push the technological envelope. They innovate. They make waves. They push the industry forward in ways no one else can - because of their convictions

Removing the headphone jack was the right call. Wireless is the future. In a few years we’ll finally see the headphone jack for how archaic it is.

We, as a species, are hooked up to focus on the short run, and we’re hooked up to seek popularity and avoid criticism. Choosing to do what you know will be unpopular in the short run but you believe will prove correct in the long run takes courage. Courage of one’s convictions, not courage running into a burning building to save a life, but courage nonetheless.

Typed on Octopage

Apple's iPhone 7 Event

Quick Thoughts


iPhone 7 looks great. I can’t wait to get my hands on one.

I’ve been an iPhone guy since 2008. Over the past 9 years (has it really been that long?) I’ve had nearly all of them. I upgraded every year like clockwork. Last year was an exception - I stuck with my 6 Plus. My 6 Plus is the first phone I’ve ever had any longer than a year. I’m looking forward to moving from my gigantor 5.5” phone to a much more reasonable 4.7”. I decided long ago that the 5.5” 6 Plus was just too large, but I still held off for the 7, that said, I’m glad I didn’t upgrade last year, it was worth waiting for this.


Apple of course kept the old standards of silver, gold, and rose gold. The standout color is of course Jet Black. Obviously I haven’t seen one of these phones in person yet, but I’ve spoken to folks who have - they tell me it looks gorgeous. There is no doubt in my mind that Jet Black will be second only to the new ‘black’. Apple was smart to limit Jet Black to the higher capacities (128GB & 256GB). Limiting it creates a bit of exclusivity, but you’re not paying any more than you would for another color in the same capacity. Ever since the 4S I’ve been a white iPhone man, I’ll be changing for the 7. I’m all aboard the Jet Black train.


Part of why I specifically chose the 6 Plus 2 years ago was for OIS camera (optical image stabilization). OIS was limited to the Plus models of the 6 and 6S. With iPhone 7, Apple is bringing OIS to the 4.7” product as well. Great!

The new dual camera system on the 7 Plus is brilliant. Apple is being conservative at first, when 7 Plus ships the OS will be able to use either the telephoto or wide angle lens. I would think a future software update will allow blending of the two images (perhaps to allow retroactive focus adjustment). 2 cameras isn’t incredibly powerful now, but it will be. I just wish this were on the 4.7” product as well.


Apple is once again proving how powerful it is to own the whole stack. They’re doing something interesting with A10 fusion. Not only are they producing a 4 core phone, but they’re actually using those cores intelligently (looking at you Android OEMs). I hope the promise of raw high performance, combined with better battery life for less intensive apps proves successful. This is clever engineering.


I’m glad to see the headphone jack die. I used the headphone jack maybe half a dozen times with my 6S in the past two years, I won’t miss it on iPhone 7.

Apple Watch

Series 2 is a great product name. It’s clearly a revision on the earlier product, but doesn’t discount the original run either.

It’s a bit quicker, it’s got some GPS chips, but it maintains the same overall design as Series 1. Apple has successfully made a better product without making me feel like a peasant for only having a Series 1.

I’m still quite happy with my original Series 1. I’m quite looking forward to watchOS 3, but I have no plans to upgrade to a Series 2 watch.


These are brilliant. Price seems reasonable (by Apple standards), functionality seems great. I love the idea of wireless headphones that just work. No stupid power buttons. No weirdo covers to remove to charge it. No crappy-ass bluetooth pairing. I love that they turn on and pair automatically, I really love that you can use only 1 Airpod if you wish. If Apple can actually deliver on the experience they’re promising, these will sell well. That said, at $159 Airpods are clearly not marketed towards the masses. Airpods will be for people who want a wireless experience and are willing to pay for it.

If I used headphones I’d seriously consider these.

The Keynote Itself

I still don’t understand why Apple didn’t announce new Macs on Wednesday. Sure, they won’t be ready to ship for another couple months (probably), but at least announce the damn things. This was a very odd keynote for me. Apple seemed to be killing time for the first hour with stupid software demos (I’ve hated the third-party software demos since they started doing it years ago) and overall very slow pacing. The latter half (the iPhone half) was rapid-fire, rapid-pace, as if they were rushing against the clock. There were only 3 things to announce, why waste time with filler in the first hour? Odd.

Typed on ErgoDox Test Board

Possible 40% w/Numpad Project

Yesterday, Hisfanator on Reddit posted this idea of a 40% keyboard with numpad. Not all that dissimilar to a RedScarf or a Monarch, just reduced to 40%.

I’ve had a Zenith Data Systems keyboard with glorious Alps SKCL Greens for a while, and I’ve been trying to think of a project for it. I’m over 60% boards for now - I love using them, but I don’t need to make another at this point. I’d considered just modifying the existing board and modernizing it. I’d also considered just harvesting the switches and using them in BlueNalgene’s upcoming Alps TKL.

What I really quite like about the Zenith (other than its switches) is the layout. It’s a QWERTY layout, but not standard ANSI. Even the bigass Enter isn’t standard.

No matter what type of board I make out of this, whether it’s a 60% or a TKL, I wouldn’t be able to reuse the Zenith’s original high quality keycaps as the layout isn’t standard. Alps keycaps are of course very very hard to come by. One either needs vintage keycaps, or inevitably lower quality modern-day ABS keycaps. Typically the modern day Alps keycaps only ever come in ANSI layout, so nothing truly innovative is possible with them.

Thus we get to this 40% concept. The notion I’ve had is to get a set of plates cut, then harvest the green switches and build a board around that. I quite like this idea as it would be a custom board from the ground up, but I could maintain the cool Zenith layout and keycaps. Essentially I would just be removing the F keys and the number row.

I’ve mocked up a version of this on Keyboard Layout Editor and then subsequently fed it through Swill’s plate building tool and generated this.

My thought is that I would do a simple sandwich design. Probably aluminum top and bottom plates with an acrylic middle plate. Nothing too complicated, but still high quality.

Obviously this is simply a concept at this time. I imagine that if I go into production with this that I’ll need to make significant changes to the plate and board design to make it fully compatible with the Zenith’s keycaps, but it’s an interesting start. I’ve slowly learned to like my Golbat and its 40% layout, so I think I could really like this. A 40% with numpad and green alps could be a hell of a combination.

With all the projects I’m working on currently, I won’t even be able to think about this seriously until next year, but I always like to keep my projects planned out months ahead of time. Maybe this one will come to fruition.

Typed on ErgoDox Test Board