18:07 <rjo_> sb0: ehsm over? how was it?

18:50 <sb0> rjo_, was good :) LEGO Photonics was one of my favorites. and of course, the DIY vacuum stuff from the Polish crew. we had some live demos of electron beam welding in the hall...

18:52 <sb0> oh, and I met one of your 'colleagues' from PTB Braunschweig

18:54 <sb0> unfortunately, the Bell test hackathon did not work out due to lack of equipment. but people talked a lot about it - maybe it'll work at the next edition (the DESY folks seem quite motivated to organize it)

19:17 <rjo_> ;) sounds good!

19:19 <rjo_> i would love to do my optics with lego if the thermal drift and backlash could be fixed.

19:22 <sb0> did you see the presentation's livestream? it seems the stability is rather good ("even with a vacuum pump running on the next table")

19:23 <sb0> if not, we're going to upload the videos soon (with much improved quality compared to ehsm-2012)

19:31 <rjo_> no i didn't see the stream. if they can stably couple light into a single mode fiber with that i will believe them.

19:36 <rjo_> what we have nowadays is already very lego-like. just terribly overpriced. otoh lego itself is also overpriced.

19:39 <mumptai> did they do optical interferometry with it?

19:41 <sb0> their main demo (on the booth) was a Michelson interferometer. not the hardest thing to do actually...

19:41 <mumptai> i know. was it thermally stable?

19:44 <mumptai> we did that in school a long time ago, and letting in some cold air from the outside (on purpose) had a huuuuge effect

19:54 <aeris> sb0 c'était vraiment super

19:58 <sb0> from #openrisc: http://pastie.org/pastes/9341373/text (after the last 2 commits)

20:17 <_florent_> hi

20:18 <_florent_> not sure if it's working, but with this I have a gain of 60 registers / 20 Luts on MiniSoC with openrisc

20:18 <_florent_> http://pastie.org/9341430

20:19 <_florent_> I'm also pretty sure we can save lots of registers on mor1kx_store_buffer.v

20:20 <_florent_> and lots of luts on the barrel shifter

20:20 <mumptai> no more lm32?

20:21 <_florent_> on lm32, the barrel shifter is done only on one direction and bit reversed depending the direction

20:24 <_florent_> mumptai : on my side I was just comparing the code of the lm32 and openrisc and try to do some optimizations

20:27 <ysionneau> _florent_: it seems sb0 has seen a cut in the size of openrisc core after the last 2 commits : 21:58 < sb0> from #openrisc: http://pastie.org/pastes/9341373/text (after the last 2 commits)

20:28 <_florent_> yes I see :)

20:28 <_florent_> I'm also playing with the code to try to reduce the logic

20:29 <ysionneau> cool!

20:30 <_florent_> but I'm just trying some simple things

22:45 <stekern> I missed _florent_, but interesting, especially about the barrel shifter