node: No such file or directory

TL;DR: When installing the nodejs package on Debian, you'll probably need to install nodejs-legacy along with it.

If you are trying to use Node.js on Debian, you might get a node: No such file or directory error. I did. The root of the problem is that, 'Both LinuxNode (package "node") and node.js (package "nodejs") are designed to be accessed through the command name "node"' (bug #614907). Debian's solution is to make Node.js access happen through the nodejs command, move the command that previously used the node command to ax25-node, and (it seems) to let nobody use the node command.

Of course this breaks everything that expects the node command to invoke Node.js, and that's why you see the error. Fortunately, the Debian solution also has a workaround: package nodejs-legacy installs a symlink from node to nodejs.

Quicktile

 
Quicktile is an nifty Python script by Stephan Sokolow that lets you tile windows in non-tiling window managers. I am currently using it with Xfce and Fluxbox (with and without xcompmgr). So far it has been pretty sweet.

The default key-bindings assume you have a numeric keypad, which my main lappy lacks. So, I cobbled together the following quicktile.cfg. It uses the alphabetic keys on the right side of a QWERTY keyboard, with key locations mapping to tile location. It also uses Windows+Alt to mask the commands instead of the original Ctrl+Windows.

[general]
cfg_schema = 1
UseWorkarea = True
ModMask = Mod1 Mod4

[keys]
C = move-to-center
H = horizontal-maximize
V = vertical-maximize
0 = maximize
b = bottom-left
n = bottom
m = bottom-right
j = left
k = middle
l = right
i = top-left
o = top
p = top-right
KP_Enter = monitor-switch

Controlling thumbnails in Thunar

Maybe I'm weird, but I don't like seeing thumbnails of my PDF documents in my file manager. This post pointed the way toward fixing that in Thunar. More concisely, Thunar uses a package called tumbler to generate thumbnails, and this is controlled by a config file. So, begin by making a local copy of the config file:

$ mkdir ~/.config/tumbler/
$ cp /etc/xdg/tumbler/tumbler.rc ~/.config/tumbler/

and then edit the local tumbler.rc file as needed to disable the unwanted thumbnailers:

# PDF/PS thumbnailer
[PopplerThumbnailer]
Disabled=true
Priority=1
Locations=
MaxFileSize=0

You'll also have to delete the unwanted thumbnails from ~/.thumbnails/normal. I have a ~/.thumbnails/large as well, and those might need to be deleted too.

Wheezy!

Debian 7.0 Wheezy is out!

Big Data wean?

Google's recent announcement to end Google Reader came as bad news to a lot of people, including yours truly. However, it might actually be precipitating my weaning off Big Data.

For a long time, it's been my intention to divest from or at least distribute my interests in Big Data. I am a heavy user of Google's search, Calendar, Gmail, Reader, and, of course, Blogger. I've repeatedly tried to find alternatives to all of the preceding before, and every attempt has ended with it just being too much bother and/or my not being able to find a workable alternative.

That's changed.

The impending demise of Reader got me off my butt and into finding something, anything (Big Data or not), that would work for me. And I think I've succeeded. It also just turned out that the best solution I found, Tiny Tiny RSS, is a self-hosted FOSS Web app. It's not perfect, mind you--but it does what I need it to.

Now this has got me going. Next up for me is Calendar--possibly using ownCloud, which could also replace Google Drive (which I barely use at all). Then probably my blogs--for which there are a lot of options.

Email and search? Those will be harder to find alternatives for, but you never know.

I intend to go with self-hosted or at least self-hostable solutions for all the preceding. The added burden is just part of the cost of maintaining a healthy technology ecology.

Totally reminds me of Linux

Qt bug report gets some love

A Qt bug report I submitted almost two years ago finally got a little attention and a status promotion from a Qt dev. Yay! That is all.

Firefox OS falling into the OEM trap?

Marty Eich over on RootzWiki deals with an issue that I mentioned in my previous post--namely that "open" in the mobile OS space has so far mostly been implemented as "open" for handset makers and carriers but not for end-users. Marty points out is that Firefox OS might have started pursuing this model as well, and that makes him (and me) uncomfortable.

I understand that without manufacturer and carrier interest any new mobile OS will be dead in the water, but let's not lose sight of the need to have a platform that's open for the end user. The two don't have to be mutually exclusive!

Why we need another smartphone OS

I'm going to step a little out of character in this post and respond to what might be nothing more than trolling. And in so doing I will probably engage in some myself. I am referring to a piece run on CNET today that asserts, "No, we don't really need another smartphone OS."

The work reads not unlike a piece of corporate protectionist propaganda, serving up the familiar, "It will be impossible for any newcomer to the mobile OS space to gain critical mass so they just better not try," stuff we've heard before. So, there's really nothing new in the content. What is new, or at least noteworthy, is the claim to address whether we need another smartphone OS--rather than how difficult a task it would be to build one--something that it pretty much fails to do. It entirely avoids the megaton gorilla that is the reason why we do need another OS: we currently lack a credible, open mobile platform. iOS is nowhere close to having any semblance of open; Android is effectively open only for vendors (not users)--and even then the OHA limits what you can do; and to the best of my knowledge Microsoft's and RIM's efforts are completely closed.

The rebuttal to the above is that, "Users don't actually care about open/free/libre/whateverness." And indeed this might be true. But it doesn't diminish the need for it. Honestly, I wonder if the titans of tech journalism are functionally concerned with the requirements of a healthy computing sociology or whether they are content acting merely as industrial spokespeople.