13:17 <wpwrak> i just saw the the mailing list moves from http://lists.milkymist.org/pipermail/devel-milkymist.org/ to https://ssl.serverraum.org/lists/listinfo/devel/pipermail/devel-milkymist.org is this move permanent ?

13:19 <ysionneau> isn't it related to the emails Sebastien sent about migrating the mailing list?

13:19 <ysionneau> and related to the errors mwalle got (bouncing email errors)

13:23 <wpwrak> ah, there it is

13:26 <wpwrak> the old url gets redirected, but not perfectly (e.g., a url to the archive page gets sent to the main page). so i wonder what the proper name is.

13:26 <wpwrak> and the transition is http://lists.milkymist.org/pipermail/devel-milkymist.org/ -> https://ssl.serverraum.org/lists-archive/devel/

13:27 <wpwrak> that's the archive. the other one was the main page

13:28 <ysionneau> yes

13:28 <lekernel> yes, the move is permanent. one problem was the website was hosted on the tmplab account and this organization is dying.

13:30 <lekernel> I wouldn't have been surprised if dreamhost terminated the account because of billing problems, and then I'd be stuck without access to the latest data, and no one responsible

13:30 <lekernel> so it's hosted on mwalle's server now

13:30 <wpwrak> so the "official" URL to the archive would now be https://ssl.serverraum.org/lists-archive/devel/ ?

13:30 <lekernel> yes

13:31 <wpwrak> kewl, thanks

13:31 <wpwrak> oh, and "tmplab" kinda suggests that it may be temporary. there is truth in advertising after all ;-)

15:04 <lekernel> is there 2.5V DDR400?

15:48 <wpwrak> like this ? http://www.issi.com/WW/pdf/43R32800B.pdf

15:52 <lekernel> ah yes

15:52 <lekernel> so it exists

15:52 <lekernel> thanks

15:52 <wpwrak> or this one. wider VDD range. http://www.winbond.com/NR/rdonlyres/F0007FC8-698E-4379-AB3A-DD9235B32BC6/0/W9464G6JH.pdf

15:52 <wpwrak> there's a lot of them

15:52 <lekernel> usually DDR400 requires 2.6V

15:52 <lekernel> oh wait

15:52 <lekernel> actually we can have 2.6V on the mixxeo

15:53 <lekernel> since we switched to VCCAUX=3.3V

15:53 <wpwrak> in digi-key, restrict the search by speed and voltage. then you get them.

15:53 <wpwrak> heh :)

15:53 <lekernel> (the upper tolerance for 2.5V VCCAUX was 2.625V, which prevented it on the M1)

15:55 <lekernel> and VCCO is specd to 2.7V. that should work.

15:55 <wpwrak> what capacity are you looking for ?

15:55 <lekernel> I guess we'll stick to the same DRAM chips as on the M1

15:56 <lekernel> except running at 2.6V

15:56 <lekernel> and if they act up, going back to 2.5V is just a resistor replacement

15:56 <wpwrak> so they can run faster ? or safer ? :)

15:56 <lekernel> they can run faster

15:56 <lekernel> and

15:56 <lekernel> 2.6V is the normal spec for DDR at 400Mbps

15:57 <lekernel> I'm aiming for a 100MHz SoC in -3 speed grade

15:57 <lekernel> with 200MHz memory clock and 400Mbps/pin

16:00 <lekernel> 128MB will be a lot of RAM for the Mixxeo...

16:00 <wpwrak> you can never have too much RAM ;-)

16:02 <wpwrak> and in any case, you should search by bracket. from minimum to the maximum you could handle. sometimes, smaller capacity is more expensive/exotic than higher capacity because the market has moved on.

16:02 <wpwrak> of course, you can still find the older and smaller stuff, but the search gets more involved

16:02 <lekernel> yes

16:02 <lekernel> better stick to a mass produced chip for the PC market

16:03 <wpwrak> yeah. if there's one thing i've learned about component selection is that you really really don't want to paint yourself into a corner when it comes to sourcing

16:03 <lekernel> sometimes I wish I had put DDR2 on the M1... in hindsight it wouldn't have been much more difficult than DDR

16:04 <wpwrak> well, there's a plan for mixxeo v2 ;-)

16:04 <lekernel> plus, doesn't require those series resistors and runs faster