Fixing or at least improving the bloomfilters would be very important IMO. 
I got shocked when I was reading the research papers regarding the bloom 
filters.
I started to work on that in my branch but did not had time yet for a more 
in depth investigation.
But fixing the most critical issues was pretty easy (though it needs a 
review and more time to be sure that it covers all edge cases).

Here is a summary of my changes as well as a call for devs to help to 
improve that:
https://github.com/bitsquare/bitsquare/issues/414

Br,
Manfred


Am Dienstag, 17. Mai 2016 10:21:24 UTC+2 schrieb Andreas Schildbach:
>
> Two more ideas: 
>
> - Proper reject message handling, feeding into transaction confidence. 
>
> - Addr message handling, to reduce dependency on the DNS and HTTP seeds. 
>
>
> On 05/14/2016 02:38 PM, Andreas Schildbach wrote: 
> > It's time to think about what we'd like to see in bitcoinj 0.15. 
> > 
> > Here are some initial ideas... 
> > 
> > - Finish the context refactoring. I recently (re-)discovered the plan 
> > laid out by Mike for this: 
> > https://github.com/bitcoinj/bitcoinj/blob/master/designdocs/Contexts.md 
> > 
> > - Migrate the build to Gradle. The main compile should be pretty 
> > straightforward; but the gritty details are a bit hidden: protobuf 
> > generation, deploying to binary repos, authenticating dependencies 
> > 
> > - Segregated Witness. Nicolas Dorier started a branch for this, 
> > unfortunately it lost a bit of steam in the past weeks: 
> > https://github.com/bitcoinj/bitcoinj/pull/1171 
> > 
> > - Introduce a Payments API that takes a different view on transactions. 
> > See this (preliminary) gist: 
> > https://gist.github.com/schildbach/a1c0c05e205baa229077 
> > 
> > What do you think? 
> > 
>
>
>

-- 
You received this message because you are subscribed to the Google Groups 
"bitcoinj" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to