Now, it's a reality.
We've been really pleased by the support and assistance that's come forth from the community during this beta testing phase. Many folks have stepped up to help test out network connections from various platforms and using various client-side tools, and we've had sharp-eyed folks catch typos on our website, broken links in the forum, and various rough edges here and there. This kind of engagement from the community is enormously helpful, as there's just no way a small team can do that kind of full-scale test assessment in-house; if we did, it'd take years to build out the network and related systems architecture... which would just be useless.
Additionally, folks have come forward to purchase network access tokens - despite the fact that we're very much in beta testing mode, and things are still far from elegant when it comes to "user interface" features (see below, for more details). That's a tangible form of support that really has helped us keep focused on making this project the best, most secure, most scalable, most durable, most resilient, most member-friendly secure network in existence.
A week into the beta launch, we'd like to share a few observations from the team:
The Beta Testing Decision
The choice to deploy the network in beta testing was made when we reached a comfort level with the overall security architecture and network auth systems. We launched before our network access widget is even ready for beta testing! The launch decision has helped us see ways to further improve the scaling capabilities of our backend systems - which was our goal - without requiring us to be ready with fully-polished client-side applications. That's enabled a multi-track approach to network improvements, and has really helped accelerate our progress overall.
"User Friendly" Tools & UI/UX Prioritisation
Without hesitation, we want to be really clear about something: we consider member-friendly tools to be equally as important as strong, well-implemented cryptographic technologies when it comes to deploying a genuinely useful secure network! We refer to "member-friendly" tools, rather than "user-friendly," because the latter has a derogatory tone, to our ear, that we feel is inappropriate. Rather, member-friendly tools are the gateway that allows serious cryptographic systems to be made available to a wide range of people - not just those with deeper tech expertise. Without that ability to provide protection to a wide audience, even the best "security technology" is useless against state-level surveillance machines and the Orwellian nightmare of NSA dragnet spying worldwide.
This is a topic on which, we as a team, are absolutely unbending: without elegant interfaces to strongly-secure systems, these systems are useless in real-world circumstances. If there isn't a "make it work" button that does exactly what it says - make it work - then the project has failed, period. The best secure network "client application" is one that doesn't exist: it fades into the background, doing what must be done reliably, securely, and with a "fail closed" approach to boundary-state circumstances.
One of the reasons our client access widget isn't ready for testing already is that we've pushed it through countless rounds of iteration in order to make it simpler, more elegant, and as "drama-free" as possible. We argue - with strong support from a host of academic studies and extraordinarily wise security experts - that a tool that is elegant, streamlined, and shorn of useless crap is by definition more secure than a bloatware-laden, complicated, unreliable "bells and whistles" monstrosity of an application. Less is more: do one thing - secure network traffic - and do it very, very well. This is our approach.
So How Does The Circle Square?
There's an interesting juxtaposition between our stated commitment to elegance in member-interface design, and our decision to roll out a beta network with essentially zero member-interface polishing whatsoever! Are we true to our word, or is there a split between what we say and what we do?
The decision we've made is to "proof" the network backend before opening up to wider community availability - this is fundamental, as connecting people to an insecure network is worse than useless. However, at the same time we feel we can best fine-tune our member-interface tools in collaboration with the community, post-launch, rather than trying to do that work entirely in-house. Since interface elegance is, by definition, measured by how it actually plays out in actual member-use scenarios, there's no sense of pretending we can do this ourselves - without extensive community feedback. So we've not pretended.
Further, when providing support for a wide range of platforms, and operating systems, and local configurations, it's all but impossible to preemptively test out every access configuration in advance. One of the powerful benefits of a beta test is that we can listen to feedback from folks with a vast range of local machine setups - and work with them to make whatever adjustments are needed to support the widest possible range of client setups. That's really effective.
So, in a sense, we're building out the member-interface side of things on the fly. This might not be entirely conventional (although lots of "agile" development models, in the deeper tech world, nowadays do exactly this as a core strategic decision), but for us it allows for a better, faster, more broadly-tested deployment. It's quite a bit of work for our team - all the feedback comes in and must be routed, digested, acted upon, and so forth - and it's of course work for our beta testers, as well. But the end result is a better network, with more elegant member-interface tools, for everyone. A big win!
Is Good Security Only For Hackers?
In a word: no.
And that's the whole point, isn't it? Any of us on our team - and many folks reading this post - will know how already to protect themselves with freely-available cryptographic tools. Yay for us. That's not going to do a damned thing to counter NSA-style illegal surveillance - it just leaves the rest of society that much more vulnerable to dragnet spying... and that's unacceptable.
Designing, building, improving, and managing security systems that are both reliably effective against intensive attack vectors and also elegant enough for non-geeks to use isn't easy. Not at all. But it's entirely possible and, franky, it's not the proverbial rocket science. The challenge lies more in taking off the "aren't we cool, we know so much about tech" chip on our shoulders, and humbling ourselves in the face of normal folks. Normal folks are people with lives, who don't want to burn time and energy learning the intricacies of crypto tech. That choice - to do other things than study crypto algorithms or systems-theoretic security models - is a healthy one. Sure, we're geeks - we love this stuff. But that makes us no better than anyone else. Indeed, it forms a sort of duty... a duty for us to use our expertise to create tools that allow other folks to benefit from what we know, as well.
That's our approach. Our job is to do this stuff right, tech-wise, and to do it so it's easy and elegant for everyone. This isn't a "geeks only" project. Well, ok, let's say this: during beta testing it's pretty much requiring some expertise with tech that's not widespread in society. That's, as we said above, a part of our testing/rollout methodology and most assuredly not some kind of acquiescence to a long-term "geeks only" status. Absolutely not!
We're geeks, and we're proud to be geeks. As are many of our members, our supporters, and those in the community who have helped this project come into existence. We're taking those skills, and experiences, and wisdoms, and we're baking them into a system that embodies them in a way that doesn't require geek status to operate. The "make it work" button, in real life.
The Next Steps
This first week of beta testing has been wonderful, honestly so. Not a figure of speech, but actually wonderful. Exciting, as something we've worked hard in behind the scenes births into a real system in the real world. Feedback from smart, dedicated, careful beta testers that's allowed us to zero in on dozens of improvements, in a flurry of fine-tuning. And, yes, some security-level holes (we're hunting a few DNS leaks in some client configurations; that's the big one right now) that we're quickly plugging. This is when things go from theoretical discussions to something... well, something real.
Next week, we're moving into an "official" launch - no longer testing, but a real rollout. The widget will be released on Monday, albeit with some fine-tuning to go as we move forward. We've got a couple of token resellers ready to go, hopefully Monday, to expand the options beyond Bitcoins and payment via Paypal (those two are already in place now). And we're going to make some wider announcements of the network; so far, we've only mentioned things to our existing customers, and to folks who follow us on twitter. Next week, we spread the word a bit further.
There will be lots of things to improve, to add, and to adjust as we go forward - that's true for any healthy tech project. We don't see a "code freeze" for the network, ever. Rather, our model is intrinsically dynamic - an ever-evolving approach that reacts to new threats, improves security against known threats, and all along works to improve member-interface elegance every step of the way. That's the "business" we're in, at core.
Shiny, Pretty, New..?
Oh, yeah... we've got some cool shit coming down the development pipeline, too. Protocol obfuscation. Leakblock integration. Local routing-table (& metric) modifications. Fast-flux based exitnode-auth redundancy to protect against Great Firewall-style DNS blocks on auth resources. VM-bound, OS-integrated network connection integration. USB image, VM-derived ephemeral secure network environments available as downloadable, self-installing iso's. Widget installers pre-loaded with auth tokens & disposable once the tokens expire, for maximal decoupling of temporal sessions. Token "wallet" functionality in the widget, to enable automated loading of new tokens without fiddly intermediate steps. Multi-layer token distribution structures to act as "token tumblers" and provide zero-knowledge proofs of token anonymity for buyers. Bitmessage-based token wallet loading tools. Blockchain-based public auth systems decentralisation. Opensourcing of exitnode management framework to enable broad community oversight & attack the "evil exit node" challenge. Token "chaining" for zero-knowledge proofs of physical IP nexus...
We've got a whole roadmap of stuff to do, as we move forward. Some of it's very, very interesting - and will change how we all think about 'network security' and what is possible in protecting against the most persistent, advanced, well-resourced threat vectors in existence. Plus, of course, additional activist outreach and additional support for others working hard - and taking real risks - to secure a free, open, diverse future for us all.
Then, there's the cleanphone project...
The shiny, pretty, new things are something we don't talk too much about publicly; it can seem a curse - "Duke Nukem" syndrome, vaporware 101. But sometimes it's ok to lay out a larger roadmap, a sense of where things are heading. In that regard, protocol obfuscation is really the neutron-star core of our future development. It's a big deal.
- - -
We don't have a "marketing department." We don't do "affiliate links," and we don't bribe "VPN review" websites to lie about how amazingly cool we are. To hell with all that. We also don't pull punches when "journalists" puff up advertisers who are deploying encraption: cryptographic systems with so many implementation flaws that they're nothing but security theatre.
Instead, we focus on our members, the community, and the technology.
Let someone else go out and fiddle SEO gimmicks to trick Google into thinking they're a credible security company (that means you, Hide My Ass aka Snitch My Ass). That's not us. We'd rather do something genuinely better, than waste time and effort trying to trick the world into think we're doing that when in fact we're just spinning bullshit. it's surprising how rare that approach is, but that's how things play out. "So it goes..." as Kurt Vonnegut succinctly put it. So it goes.
We're out to change the world.
That means we're doing things unlike others do things - that's not a bug, it's a feature. We might make mistakes - heck, we'll certainly make mistakes and already have - but we'll correct them, learn from them, and get better every step. With community engagement, community critique, and public review of what we're doing, we stand a fighting chance to make things better for everyone.
And that's our goal. We want to see these tools in the hands of millions of people - not a few thousand, thanks. Millions. There's a few billion people using the internet every day; how many have protection from surveillance... real protection? A few hundred thousand Tor users, perhaps? That's, what, less than 0.01% of internet users? That's pathetic. Sorry, but it is.
If we want to turn the tide against the global surveillance machine, we need to put these tools in the hands of a few hundred million folks... for a start. Shoot for a billion, go from there. That's the big picture, and we're in it to impact that picture - not to play in the kiddie pool. There's enough kids out there splashing around in the yellow-tinged shallows. Someone's gotta go out into the sharky, deep waters and face the ocean.
Thank you, again, for your support and your encouragement and your sharp critique and your "it's broken" missives, and your suggestions and above all else your patience as we make this thing come alive. We're mortals, the team here: just ignorant monkeys doing our best to manifest a soupçon of wisdom, and perhaps be of some use to the rest of our world along the way. We're imperfect - but we're honest enough to reach out and lean on others to help lessen those imperfections and, together, make us stronger as a result.
Stronger we shall become, day by day. That's how we're going to change the world: together, one step at a time, sights set high.
- ~ cryptostorm_team