Comments for technomasochism

Comment on Financial Site Password Policies by Tom Dilatush

Tom Dilatush — Thu, 21 Feb 2013 20:00:20 +0000

I just went through a very similar experience, starting exactly the same way – moving my retirement accounts to Fidelity. I was annoyed by the password restrictions, but used a randomly-generated 12 character string and felt ok about it. Then I ran into the fact that their web site accepts passwords by touch-tone keyboard and I was floored. It’s actually a little worse than you stated in your post – the touchtone “character space” is only ten characters (when you press the “5″ key, it can’t tell whether you meant “5″, “j”, “k”, or “l”)!

Good grief…

Comment on Measured Launch on OE core by Philip

Philip — Sun, 20 Jan 2013 04:02:42 +0000

Thanks JP. Your link to PCR 17 calculation is much better than the one I tracked down. That will be very helpful. Still thinking about how to integrate this into the OE build but I’ll start playing around with the parsing and calculations first.

The method you posted is a bit cryptic but only in that it uses some scripts without providing their sources … got a github I can pull from to checkout the additional code?

Comment on Measured Launch on OE core by jpb

jpb — Sun, 13 Jan 2013 20:04:06 +0000

Nice work!

Slightly more recent discussion of legacy pcr 17 calculation here:
http://www.mail-archive.com/tboot-devel@lists.sourceforge.net/msg00332.html

Re:time for tpm to hash rootfs
My guess is days. TCG doesn’t specify min specs. I’d be interested in a toolset that uses regular hardware (maybe with help of fairly new Intel AVX instructions) to crank out hashes faster.

Comment on Comparison of Catalyst Logo by Philip

Philip — Sat, 08 Sep 2012 14:47:09 +0000

Jeez it’s been nearly two years since I put this one up. The SPI never responded to my email. Your point about infringement is well taken but unless the copyright holder is willing and has the funds to enforce the copyright it’s moot I guess.

Comment on Comparison of Catalyst Logo by Carsten

Carsten — Sat, 08 Sep 2012 13:07:41 +0000

Hi, I just spotted this similarity myself. It was a low-resolution photo in a press article, and the logo was small somewhere in the corner, and still it was immediately apparent to me that they copied the Debian logo. Then I googled it and found this blog entry.

Did you receive any answer yet? This is a clear copyright violation, and I think the Debian people should act on it. In a time where people are being sued for downloading copyrighted stuff for private use, I think its outrageous that this company can get away with commercial use.

Comment on TXT Capable Desktop Virtualization System by Philip

Philip — Sat, 01 Sep 2012 20:41:28 +0000

In the later 4.X Xen releases it’s actually quite good. Passing through stuff like a NIC “just works”. GPUs can get tricky but it’s usually the fault of the desktop. Dell has a tendency to treat their “primary” PCIe slot special and the BIOS reconfigures a bunch of stuff automagically when there’s a graphics card in it (disable intel graphics hardware etc). This will break PCI passthrough on Xen last time I checked.

A good work around on the Dell is to use the 4xPCIe slot instead of the 16x. With this config on a Dell 990 I was able to pass through an ATI card to a guest with no problems. This left the Intel graphics in dom0. Nvidia cards have a tendency to be more difficult but I’ve heard of people having success passing through large SLI cards to guests for CUDA processing.

I haven’t tried any of this on the Intel board from this post though. If you get adventurous and start playing around let me know what you come up with.

Comment on TXT Capable Desktop Virtualization System by Reid

Reid — Sat, 01 Sep 2012 02:18:36 +0000

How good is PCI passthrough nowadays? I’m assuming you can share most any card with one guest and do some fancy networking footwork through the host?

I’ve long dreamt of building a gaming/video rig where I could run an IDS/firewall on dom0 (networking to the guest via a virtual adapter) and have good graphics speeds to boot. But then I’m paranoid, maybe?

Reid

Comment on Openembedded Yocto Native Hello World: Take 2 by technomasochism - openembedded yocto native hello world

technomasochism - openembedded yocto native hello world — Sun, 19 Aug 2012 18:29:34 +0000

[...] I took the time to get to the bottom of the issue discussed in this post. There’s a new post here that explains the “right way” to use Makefiles with yocto. As always, the error in this [...]

Comment on openembedded yocto native hello world by technomasochism - Openembedded Yocto Native Hello World: Take 2

technomasochism - Openembedded Yocto Native Hello World: Take 2 — Sun, 19 Aug 2012 18:28:43 +0000

[...] packages with raw Makefiles (no autotools). I wrote up the problem and the work around I was using here. I got some feedback pointing out what I was doing wrong but I guess my brain just didn’t [...]

Comment on Using tmpfs to Minimize Disk IO by PiServ.nl » Reducing SD card I/O activity on the Raspberry Pi

PiServ.nl » Reducing SD card I/O activity on the Raspberry Pi — Mon, 13 Aug 2012 16:49:09 +0000

[...] http://www.cyrius.com/debian/nslu2/linux-on-flash.html http://twobit.us/blog/2010/01/using-tmpfs-to-minimize-disk-io/ http://www.linuxquestions.org/questions/suse-novell-60/interpreting-proc-diskstats-360350/ Tags: [...]