I think he's probably right in terms of what a government research program should have as its goals. IMO, the purpose of government research on this scale is to drive forward technological development and give the private sector a kick in the pants.

We've already been to the Moon, that technology was developed during the 1960s. We could probably do it better now, but the advancements wouldn't be nearly as significant as what is required for a manned mission to Mars. Leave the moon to the private sector, we should expect to see a private company touching down there within a decade or maybe two. Mars is still a pie-in-the-sky target, let's point NASA at that.

Joke option should have been "It's complex".

Sitting in my living room I can pick up 19 APs, and they're all encrypted. I'm so proud of my neighbors.

I'm pretty sure they already have.

Plato thought that the best form of government would be rule by philosopher-kings.

I'm with Plato: the general population is too stupid for a democratic system. Unfortunately we have not yet reached Culture levels of technology so it's the best option we have at the moment.

Can you imagine having to sit through SIX YEARS of a case? The last time I was summoned for jury duty it took three days until I was dismissed from selection and I was already pissed at everyone involved for wasting my time.

RTFS. Cell signal, not wifi.

lololol RT @adamofgreyskull /dev/sda5 /tmp nosuid,noexec,nodev 0 0

If I were working I would not be reading Slashdot, would I?

I am *at* work, though.

Supporting distros is easy. Most package management boils down to "./configure ; make ; make install" then you create a glorified tarball of the results. For a complex application you may break it into multiple glorified tarballs then stick them into a repository (yum, apt, whatever) and generate some metadata on the packages using a simple command (it's literally "createrepo ./" for RPM distros). This is all basic stuff and can be done in an afternoon.

It's also important to know your market. You say "supporting 80% of the linux user base", but for most apps that is not the goal. If you're writing a game, for instance, you're only interested in desktop users. You can hit 80% coverage by targeting Fedora and Ubuntu alone. If you're writing server side software to be used in datacenters then you can hit 80% by supporting RHEL, CentOS, and Debian/Ubuntu.

In this particular market, RHEL and CentOS are all we need. Our customers are already using them. Adding support for additional distros is possible, of course, and fairly trivial to do. I have worked on other products which supported RHEL, CentOS, Fedora, Debian, Ubuntu, Slackware, and Solaris simultaneously.

I am writing an installer for an enterprise telco carrier application today. It's actually very simple. The installer logic is as follows:

1. Verify that the OS we're running on is a supported one (RHEL or CentOS)
2. Verify that some required hardware is present and configured correctly
3. Configure a yum repository (5 line text file listing where to find RPMs) for our application RPMs present on the CD
4. yum --enablerepo=companyname install productname
5. From this point on yum and RPM work out all of the dependencies and our application gets installed. We can even pull updates from our repositories connected to the Internet (access restricted based on known client source IPs). This is all done using the default distro package management routines.

The real problem in Windows is that all of these software packages have their own independent (and potentially broken) update mechanisms. One thing that modern Linux distros get right is centralized software updates. My Ubuntu laptop has a dialog box waiting for me most mornings that details any software updates it would like to install, and whether or not they are security related. I could tell it to do it all automatically but I like reviewing the changes before I install them.

Oh, and not to start a flame-war, but ASL is the "Ebonics" of sign language. Learn Signed English, instead. Yes, they're different.

Sorry, but I'm actually completely not interested in languages. I took ASL because it was considered a foreign language by my high school (this was 15 years ago) and I didn't want to bother with anything that had a complicated grammar. You have interested me in Signed English somewhat, though. Even with my passing study I'd noticed that ASL was somewhat simplistic (the English equivalent of "me go mall" rather than "I went to the mall") and I'd always wondered if that was a symptom of the language itself or the teacher.

According to my ASL instructor, lip readers are rarely more than 50% accurate. Which makes me wonder about the alleged capabilities of this software, honestly.