'Widespread adoption' is not the current sort-term aim. While we all think that fast, universal, anonymous internet access would be a good thing, we simply can't support that right now. The volunteer network of relays isn't that big. Even now, Tor has trouble dealing with the network load. If Tor were to become more popular with users, without seeing a commensurate increase in the relay capacity, that would massively reduce the functionality of the network. Right now, It's important to make tor available for the most at-risk users: those in oppressive regimes, whistle-blowers, undercover journalists and so on. After that, the priority is on improving the structure of the system, not on further usability.<div>
<br></div><div>While I can see where you're coming from in suggesting that a kernel driver is better, the reality is more complex. Because the primary development effort is on the core Tor software, a VM requires a relatively small development effort at present. One barely even needs to be able to code to construct a VM that uses existing software: it's mostly an exercise in implementing best-practices. Moreover, a VM is actually easier to support right now than a driver. A driver would need a coder dedicated to maintaining it. A VM on the other hand needs only a geek, and can easily update to the latest versions of Tor (and supporting applications) when they are released.</div>
<div><br><div><br><div class="gmail_quote">On Thu, Aug 13, 2009 at 01:23, Peter <span dir="ltr"><<a href="mailto:necedemalis@gmail.com">necedemalis@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
I'm not complaining about it, I'm just saying, if you want widespread<br>
adoption, a kernel driver is the way to go. And moreover, a kernel<br>
driver is easier to write and support than a VM.<br>
<div><div></div><div class="h5"><br>
On Thu, Aug 13, 2009 at 1:21 AM, Flamsmark<<a href="mailto:flamsmark@gmail.com">flamsmark@gmail.com</a>> wrote:<br>
> Then perhaps complaining about the direction of the work that many others<br>
> have done pro bono is a little premature, no?<br>
><br>
> On Thu, Aug 13, 2009 at 01:18, Peter <<a href="mailto:necedemalis@gmail.com">necedemalis@gmail.com</a>> wrote:<br>
>><br>
>> Heh, well, I tell you what. You send me a hundred thousand dollars,<br>
>> and after the check clears I'll write you a great windows kernel<br>
>> driver. Otherwise, I'm broke, my life is a living hell, and I already<br>
>> have several projects I work on out of the goodness of my heart, so,<br>
>> I'm sorry.<br>
>><br>
>> Thanks and good luck.<br>
>><br>
>> On Thu, Aug 13, 2009 at 1:13 AM, Flamsmark<<a href="mailto:flamsmark@gmail.com">flamsmark@gmail.com</a>> wrote:<br>
>> > Why not just a Windows kernel driver? Because it hasn't been written<br>
>> > yet.<br>
>> > You're welcome to help write a kernel driver, or a VPN host or whatever<br>
>> > else<br>
>> > you think is the next logical step to improving Tor. However, remember<br>
>> > the<br>
>> > version number: 0.2.1.*. Tor is not a 'finished' piece of software. It<br>
>> > is<br>
>> > not feature-complete; it does not implement everything that's either<br>
>> > desired<br>
>> > or required for ideal use. However, right now, much of the development<br>
>> > effort is not spent making it easier for clients to use. There's a<br>
>> > feeling<br>
>> > that it's currently 'good enough' that those who really need to use Tor<br>
>> > will<br>
>> > be able to follow the instructions and get it working. If you don't<br>
>> > agree<br>
>> > with that emphasis, again, it's your prerogative to build those feature<br>
>> > that<br>
>> > you think are most important.<br>
>> > We all look forward to seeing your contributions!<br>
><br>
><br>
</div></div></blockquote></div><br></div></div>