[Libre-soc-dev] new NLnet grant funds

[djac at redsemiconductor.com]

Yes, bring it up tomorrow please.

 

 

From: Luke Kenneth Casson Leighton <lkcl at lkcl.net> 
Sent: 01 January 2024 20:07
To: Libre-Soc General Development <libre-soc-dev at lists.libre-soc.org>; David Calderwood <djac at redsemiconductor.com>; James Lewis <james.lewis at redsemiconductor.com>
Subject: new NLnet grant funds

 

https://nlnet.nl/taler/

this is one of the new NLnet Grants, co-sponsored by the EU
and the Swiss State Secretariat for Education, Research and
Innovation (SERI).

it would be perfect for a continuation of the cryptoprimitives
grant, focussing on algorithms utilised by GNU Taler. see
https://www.google.com/search?q=gnu+taler+cryptographic+algorithms

section 4.1.2 page 62.

the algorithms are:

* SHA512 (Power ISA but VSX - TODO, add scalar variants)
* ed25519 (already in the crypto grant)
* curve25519 (uses ed25519 underneath)
* HKDF (trivial)
* FDH-RSA (also very simple)

ED25519:

* https://bugs.libre-soc.org/show_bug.cgi?id=773#c3

HKDF https://github.com/casebeer/python-hkdf/blob/master/hkdf.py

FDH-RSA:

* https://github.com/taabishm2/FDH-RSA-Signature/blob/master/fdh.py
* https://github.com/taabishm2/FDH-RSA-Signature/blob/master/signRSA.py

FDH-RSA is basically multiple SHA256s back to back on the same
data with a sequence number 0 1 2 3 ... appended, then
pow(x,y,mod) on that.

with SHA-256 and SHA-512 having been added to Power ISA
(likely as VSX so we will need to add Scalar variants)
it should be quite a straightforward grant application
that gives us easy funding for not a huge amount of work,
with the added benefit of "putting feet on the ground"
of the previous pure research work (cryptoprimitives grant)
into an ecosystem that is gaining a lot of traction in
the banking arena.

thoughts?

l.


-- 

---
geometry: without it life is pointless

the fibonacci series: easy as 1 1 2 3