GNOME 3 Stupidity

FFS can somebody already fork this stuff away from the people so enamored with giving me a shitty tablet experience on my primary development workstation already?

Seriously. I wonder why I have any faith what-so-ever left in the Linux desktop experience not going entirely to shit over the next few years.

On a dual head display of Fedora 17 (double 1900x1200) I get this scroll bar and a footprint usage of the system panel on < 1/4 of one display.

Add to this, the fact that I don't have nor ever will own an over-priced Wacom Graphics Tablet and this desktop does not have Bluetooth.

Frankly, this is pure genius.

Then there is today's minor gripe, the fact that the notify area is forced to the bottom of one display, nicely taking up real estate on my primary display of course.

Why not add the current randomization of the full-screen location? Whenever I full-screen an app, it seems to be a mystery as to which display it's going to go fullscreen on at any particular time, and certainly not the display the app is currently situated.

If the developers of GNOME 3 and the integration people of Fedora are trying to keep me entertained annoyed while they continue to pilot the Linux GUI experience into the future down the toilet, they are doing a good job. 

Xilinx ISE Webpack on 64-bit Ubuntu 12.04

It's been a bit of a boring waste of effort sorting out the dogs breakfast to make this work... here's some stuff I did that will hopefully let you not loose time for the same reasons:

a) download the webpack blob, and install it.
b) to get a license, the install dialog didn't send me to their website so i logged in manually and created the free webpack license, downloaded it and then selected it
c) had to install portmap service
d) had to install libstdc++5
e) had to find libmotif3 64-bit deb and install it

floorplanner lauch script is taking forever to load and fail at this point. still trying to sort it out.

f) installed xfonts-100dpi and xfonts-75dpi as recommended in some forum

receiving RPC auth errors on the console,
receiving Wind/U Error (122) Function GetTextMetrics errors on the console

see this other guy's blog post here: http://wienker.org/blog/?p=188 for more issues/solutions

[update] got floorplanner running by following this: http://forums.xilinx.com/t5/Archived-ISE-issues/XPower-in-Linux/td-p/761 where you add some fallback fonts.

portmap rpcbind errors go away if you run rpcbind with "-wi" options instead of default "-w" option, but you'll have to lock yourself down with iptables afterwards.

RAMPS 1.4 Schematic Errors

I've been working with the RAMPS 1.4 board design, in particular modifying it a bit and as a result, I've noticed a couple errors...

In the top left corner, RX/TX are numbered in correctly, they are really digital 0 and digital 1, not 1 and 2.

In the middle bottom, A-OUT and SERIAL connectors are basically swapped. SERIAL is really D0, D1 (the RX/TX mentioned above) and not A3, A4. And therefore A-OUT must really be A3, A4.

On the PCB, it's hard to tell which is A-OUT and which is SERIAL on AUX-1. The serial pins have tiny letters "s" next to them. This is the row of 4 towards the center of the board. A-OUT is the row towards the board edge.