Since the initial launch of Qubes OS back in April 2010, work on Qubes has been funded in several different ways. Originally a pet project, it was first supported by Invisible Things Lab (ITL) out of the money we earned on various R&D and consulting contracts. Later, we decided that we should try to commercialize it. Our idea, back then, was to commercialize Windows AppVM support. Unlike the rest of Qubes OS, which is licensed under GPLv2, we thought we would offer Windows AppVM support under a proprietary license. Even though we made a lot of progress on both the business and technical sides of this endeavor, it ultimately failed.
Luckily, we got a helping hand from the Open Technology Fund (OTF), which has supported the project for the past two years. While not a large sum of money in itself, it did help us a lot, especially with all the work necessary to improve Qubes’ user interface, documentation, and outreach to new communities. Indeed, the (estimated) Qubes user base has grown significantly over that period. Thank you, OTF!
But Qubes is more than just a nice UI: it’s an entirely new, complex system — a system that aims to change the game of endpoint security. Consequently, it requires expertise covering a wide spectrum of topics: from understanding low-level aspects of hardware and firmware (and how they translate to the security of a desktop system), to UI design, documentation writing, and community outreach. Even if we consider only the “security research” aspect of Qubes, this area alone easily scales beyond the capabilities of a single human being.
In order to continue to deliver on its promise of strong desktop security, Qubes must retain and expand its core team, and this requires substantial funding. At this point, we believe the only realistic way to achieve this is through commercialization, supplemented by community funding.
We’re taking a different approach to commercialization this time. Building on the success of the recent Qubes 3.2 release, which has been praised by users for its stability and overall usability, we will begin offering commercial editions (licenses) of Qubes OS to corporate customers. We believe that the maturity of Qubes, combined with its powerful new management stack, makes it ripe for adoption by any corporation with significant security needs.
Commercial editions of Qubes OS will be customized to meet special corporate requirements. For example, two features that might be particularly attractive to corporate customers are (1) “locking down” dom0 in order to separate the user and administrator roles and (2) integrating our local management stack with a corporation’s remote management infrastructure. These are both examples of features that our developers are capable of implementing now, on Qubes 3.2.
We plan to partner with one to three corporate clients in order to run a pilot program throughout the first half of 2017. After it has been successfully completed, we’ll then widen our offer to more corporate customers and, ultimately, to small business customers. Our main constraint is the scalability required to cover each additional client. Hence, we plan to focus on larger customers first.
Let there be no misunderstanding: Qubes OS will always remain open source. We anticipate that the majority of our commercialization efforts will involve the creation of custom Salt configurations, and perhaps writing a few additional apps and integration code. In the event that any corporate features require reworking the core Qubes code, that new code will remain open source.
We considered many other ways of attempting to commercialize Qubes before arriving at this model. One possibility that some of our users have inquired about is that we sell dedicated Qubes hardware (i.e. laptops). However, there are a number of challenges here, both in terms of making the hardware trustworthy enough to merit our “seal of approval”, and from a business and logistics perspective. For these reasons, we don’t plan to pursue this option in the immediate future.
Unfortunately, the financial necessity of shifting our priorities to commercial clients will mean that we have less time to work on features that benefit the wider, security-minded open source community, which has been our focus for the past seven years. This deeply saddens us. (We all use Qubes on our personal computers too!) However, the reality is that ITL can’t afford to sustain the open source development of Qubes for much longer. We’re running out of time.
In an attempt to keep the open source development of Qubes going, we’ve teamed up with Open Collective, which makes it easier to donate to the Qubes project. Now, in addition to our Bitcoin fund, we can also accept donations via credit card. ITL will not benefit from any of the money donated through Open Collective. Instead, the funds will be paid directly to individual developers who have been hired to work on the open source edition of Qubes. With the help of our community, we hope eventually to build a nonprofit organization that will ensure the long-term future of Qubes as an open source operating system that is freely available to all — one of the few operating systems that places the security of its users above all else.
If you are a user of Qubes and want to help us continue working on it, please donate now. Those who have contributed will be publicly recognized on our Open Collective page (if they so choose). Organizations that support the Qubes project will be publicly recognized on our Partners page (again, if they so choose). If you are interested in supporting Qubes with significant resources, whether as an individual or on behalf of an organization, we ask that you please contact us directly, since donating through Open Collective entails significant administrative overhead.
Thank you for your continued support. Together, we can ensure that Qubes is around to secure our digital lives for many years to come.
Q & A
Here are some of the questions we’ve received from the community regarding this announcement.
Q: I’m concerned that large donors might weild undue influence over the direction of the Qubes project. Is this something users should worry about?
A: No. Our position has always been — and continues to be — that donors do not have any special influence over the direction of the project. Making a donation does not entitle anyone to decide or vote on development priorities.
Q: Will the process of receiving and spending donations be transparent?
A: Yes. All donations received through Open Collective will be publicly visible, as will the way this money is spent. (You can view this on our Open Collective page.)
Q: Can you give me an idea of how much money the project needs?
A: I think some kind of measurement is that we’ve received $410k from OTF over the last year. During that time, we’ve got also received some much smaller donations, but that doesn’t change the above figure much. This was enough to survive and release new versions, but not enough to implement everything we’ve planned for. Some examples:
- Qubes 4.0 is delayed more and more,
- We haven’t managed to add Gnome support
- Live USB is still in alpha phase and basically unmaintained
- A lot of bugs are not fixed (see GitHub issues)
This is a lot of money. And we don’t think it’s realistic to collect it just from public donations. This is why we’re introducing a commercial edition. But that doesn’t mean we don’t need your support. On the contrary! We’d love to continue work on the open source edition as much as possible! Not only as a base for commercial product, but also as a fully usable and functional system on its own!
Q: The announcement says that if I’m interested in supporting Qubes with “significant resources,” I should contact you directly. How much is “significant”?
A: I think anything >= $10k. We also have some internal overhead of handling individual donations manually (mostly related to the number of such donations, not necessarily the amount), so it’s not really worth it for smaller amounts. That’s why we have Open Collective to handle these.
Q: Will any/all of the current ITL devs continue to work on the open source edition of Qubes?
A: Probably, but it depends on how much of their time is demanded by corporate clients. (Remember, though, that any changes to core code will remain open source, so it’s quite likely that the core devs will continue to contribute to the open source edition, at least indirectly.)
Q: Will any/all of the current non-ITL devs continue to work on the open source edition of Qubes?
A: Probably, but it depends on what they want to do, how much gets donated, and therefore how much money is available to pay them.
Q: If new developers are hired, will the quality of their code be reviewed to ensure the high standards of the Qubes project are maintained?
A: Yes, absolutely.