On 11/17/2014 02:14 AM, Florent Daigniere wrote:
> On Sun, 2014-11-16 at 18:26 +0000, Matthew Toseland wrote:
>> On 16/11/14 17:30, Matthew Toseland wrote:
>>> On 16/11/14 16:36, Ian Clarke wrote:
>>>> We're in an interesting situation. The world finally appears to really
>>>> care about the things that Freenet has been about from the very beginning a
>>>> decade and a half ago (most of the publicity back then viewed Freenet
>>>> through the prism of Napster and copyright infringement). People finally
>>>> care about anonymity, privacy, government monitoring, etc. We should be
>>>> able to capitalize on this but it will take work.
>>> And in the meantime every wannabe clone project gets all the funding,
>>> and we don't, because we're old news. Yeah.
>>>
>>> You need to look at Arne's work on a funding proposal ("Freenet for
>>> journalists"). IMHO to make a big difference, make some decent progress
>>> towards something we could call 1.0, we need $1M+, i.e. a team of paid
>>> devs for at least a year. As you mentioned earlier, Kickstarter is out -
>>> it takes a lot of design resources up front, we don't have goodies to
>>> give donors, and we're excluded anyway because of being a social network.
>>>
>>> http://127.0.0.1:8888/USK@s9sxY2cTJWHKRsTuBTkjrXW4HfzrdUlwFqft1mzV0Gs,2E4DOMYy-~zOdp8-5OQH2IcmLfey0AOIkms-73Mx2tI,AQACAAE/freenet-funding/40/
>>> (Is there an open gateway at the moment???)
>>>> Thoughts?
>>> Well, this summer we've made significant progress, some of which has
>>> been deployed:
>>> - The client layer rewrite. This will be deployed soon and will have a
>>> substantial positive impact on usability for serious users, as well as
>>> solving a number of long term problems like big freesite uploads.
>>> - The link length changes. These are already deployed and have greatly
>>> improved opennet performance, doubling the average data persistence time.
>>> - Xor is on the verge of solving two of the biggest
>>> usability-for-real-users issues with WoT.
>>>
>>> I agree that we need a new UI. I don't think I'm the person to do it,
>>> but it sounds like the Winterface project is making progress on this.
>>>
>>> Personally I think we should put some significant effort into security;
>>> any tweaks we do around the edges to boost opennet performance can
>>> probably be sabotaged fairly easily, and there are real enemies. That
>>> means 1) making darknet easy, 2) making darknet fast, 3) solving the
>>> Pitch Black attack and 4) implementing tunnels, both on opennet and
>>> darknet. A good tunnel implementation could give us better anonymity
>>> than Tor, at least in theory. There's lots we could do on performance
>>> but little point if we're relying on an easily-sabotaged opennet model.
>>>
>>> We have code for transport plugins, and it's very popular among the
>>> community, but it probably needs rewriting and would be a big project...
>>> I'm not convinced it's worthwhile given that it provides no practical
>>> benefit except for a few users with UDP blocked, and no real security
>>> benefit (even with darknet, p2p networks are detectable regardless of
>>> protocol).
>>>
>>> Sone needs some problems sorting out, making into an official plugin and
>>> so on. There's lots of stuff we could do to improve performance of chat
>>> apps and so on; IMHO making chat work consistently well over Freenet is
>>> a long-term research project, like search. :|
>>>
>>> The existing fred changes, the two main WoT changes (event pushing and
>>> progressive recalculation) and fixing Sone and making it official would
>>> be enough for a reasonably interesting 0.8 IMHO, certainly an
>>> improvement on 0.7.5; a bit better if it includes Winterface. But I have
>>> no idea what the timescale is for Winterface, the WoT work or Sone.
>>>
>>> This is IMHO. I've been out of it really for some time; I was fairly
>>> single-minded over summer. You should look at other people's roadmaps.
>>> But it's worth discussing.
>> Lets separate this out into another thread:
>>
>> I propose that we ship 0.8 with:
>> - The existing client layer rewrite (the next build) and opennet
>> improvements.
>> - Progressive recalculation and event-pushing in WoT.
>> - Winterface.
>> - Sone as an official (but not bundled) plugin.
>> - Any bug fixes we have time for.
>>
>> What needs to happen for this to be possible? What is the status of each
>> of these components?
>> (Of course there is a connection between this and the other issues -
>> we'd like to ship 0.8 *and* milk it for donations by having a clear
>> vision to propose to potential donors)
>
> We're making over and over the same mistakes... None of the components
> you're talking about above are in a released build (they haven't been
> tested in a meaningful way... and there's over 100kloc of diff!)...
> Don't set a feature goal, set a time-goal; at least that won't
> disappoint. Answer the "when do you need it for?" question, then the
> feature list will become obvious.With this in mind I think Winterface and Sone are dropped from the 0.8 list. They'd certainly be nice to have, but are too big to hold up a release. It seems really hard to set a time-goal when we only have volunteers to work on Fred - does anything come to mind? > From the list above, there's several critical items missing IMO: > - a new OSX installer (we currently install through JNLP; that's known > to be broken on anything even remotely modern) What about dropping support for OS X unless an OS X maintainer joins us? > - a "better" windows installer (various bugs are in deer need of fixing) Oh? Are there any of these [0] you categorize as such? - Steve [0] https://bugs.freenetproject.org/view_all_bug_page.php
signature.asc
Description: OpenPGP digital signature
_______________________________________________ Devl mailing list Devl@freenetproject.org https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
