> Maybe so, but it means a full rewrite every few years, which we are
> unlikely to have the resources for. Even if we did, it would mean
> throwing out years of hard-won expertise.
That's not necessarily a bad thing. It forces the accumulation of new
expertise, keeps bringing in new generations of developers and makes sure
the project takes advantage of whatever language advancements have taken
place. Encourages forking and pits the different implementations against
one another so that the fittest may survive. OTOH having a single
platform/toolset dictated discourages contributions. Yes, java is the most
popular language by many metrics, but that popularity does not translate to
a large pool of passionate volunteers.
> But there are other trends that might favour us, e.g. cheap but powerful
router boxes, Raspberry Pi /
> Arduino hobbyist stuff etc.
Every trend can bring more volunteers passionate about it, so all venues
are worth pursuing :)
zab/topiltzin
On Thu, Oct 15, 2015 at 8:57 PM, Matthew Toseland <mj...@cam.ac.uk> wrote:
> On 15/10/15 20:40, Zlatin Balevsky wrote:
> > I first got involved in Freenet when it was 0.2. At the time it was
> using
> > cutting edge technologies and an contributing was an opportunity to learn
> > valuable skills. Contributing was fun and that was the driving factor.
> >
> > If Freenet was to start fresh, it should do whatever it takes to regain
> the
> > coolness factor. That means embracing new tools and technologies even if
> > there is no strict technological advantage in doing so. For better or
> > worse Java will never be hip with the open-source crowd, and personally,
> > after 10 hours of looking at Java code for my day job the last thing I
> want
> > is to look at more Java code in my free time. Some exotic new language
> > like Scala or Go or whatever the $COOL_LANGUAGE_DU_JOUR is would be a
> > different story.
> Maybe so, but it means a full rewrite every few years, which we are
> unlikely to have the resources for. Even if we did, it would mean
> throwing out years of hard-won expertise.
>
> Can we make it more attractive to new devs without needing to take such
> a drastic step?
> > Yes this can lead to fragmentation as various contributors veer off each
> > into their own direction; it's the job of the leader to keep things
> > coherent and aligned with the project vision. It's very easy to
> > underestimate how difficult the job of the leader is.
> Agreed.
> > Lastly, I'd like to point out that mobile is the future - not that I like
> > that a single bit.
> If mobile is the future, we're stuffed. Mobile simply can't do p2p. The
> networks will do everything necessary to stop it, and it drains power,
> storage and above all scarce bandwidth. The only realistic options for
> mobile are pure client nodes ("transient mode"), which is what mobile is
> designed for, or variants on Sneakernet. But there are other trends that
> might favour us, e.g. cheap but powerful router boxes, Raspberry Pi /
> Arduino hobbyist stuff etc.
> > zab/topiltzin
>
>
> _______________________________________________
> Devl mailing list
> Devl@freenetproject.org
> https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
>
_______________________________________________
Devl mailing list
Devl@freenetproject.org
https://emu.freenetproject.org/cgi-bin/mailman/listinfo/devl
