It sounds like homebrew provides a good solution. Perl (and Ruby and Python) already have mature packaging systems and they really don't need to interact with each other. So homebrew is a smart packaging system that plays nice with others.

Use unison (free) or Super Flexible File Synchronizer (better). Keeping several hosts synchronized is tough because of conflicting changes, temporary files, large stuff you don't want to transfer and moving files. You need a good UI and smart change tracking, which is what these programs provide.

I agree with most of your post, but Python, C, Lua, and Java are so similar, you shouldn't have any trouble reading through python code if you know one of the others. Especially with a searchable manual by your side.

Hacking on emerge would be harder, but all these langs are basically interchangable. If only they shared a reasonable ABI and VM so they could be mixed freely. But we are stuck with bloated crap like Java and Mono.

I see a dozen comments applauding your idea, but I think we need some counter arguments. Unless your hashing function is time consuming or you're a robot, it must be simple enough that a few examples and brute forcing would allowing guessing other passwords.

Just use the same set of passwords for all sites - it's simpler and as secure, more or less.

I also had a good experience like this in college. My iBook screen broke. I mailed it in and they fixed it. After a few weeks it stopped powering on. So they fixed it again.

Then I was sitting cross-legged with my notebook and the angle put too much strain on the ethernet, snapping it off the motherboard.

I tried to fix it myself with a soldering iron and after I put it back together it didn't work at all.

So I called and said it was broke - no questions asked, they sent me the latest model. They even let me downgrade from the 14 inch (which was just too big) to the 12 inch and they let me use the difference to get RAM and hard-disk upgrades.

if (!flag) // looks like I'm testing a pointer??
if (!count) // looks like I'm testing a flag??
if (!ptr) // looks like I'm testing a counter??

Not only should the name give a clue, but the type of the variable should be available within the past 50 lines or so. I have a bigger problem with the readability of complex conditionals and the !ptr syntax helps. Actually I'm interested if you can give an example of good code where ptr == NULL is more clear than !ptr.

Hi. Thanks for your comment - the replies make this one of the most insightful threads I've read on /.. People get so angry at just the mention of raising gas taxes without even considering the evidence for it.

I support raising gas prices because I don't trust anyone, especially government, to micromanage production without the help of a free market. Small government means simple rules.

You might want to look up a book called 'Natural Capitalism'. It reasons that government is responsible for (1) protecting the environment and (2) raising capitol. Both these goals can be achieved by taxing natural resources, like oil or clean air.

Now when government taxes resources they can reduce income tax. Corporations that now choose to high one guy to run a oil-burning machine will be able to hire two workers instead, since lower income tax means workers can make the same, but cost the company less. This kind of affect cannot be found by mandating car efficiency, but it's exactly the thing we need.

You seem to have ignored your own reasoning that we shouldn't think of "Linux" as a whole, but just one distro at a time. Oracle doesn't support "Linux", they support one distribution: "Oracle Unbreakable Linux". Now the market penetration and package availability of their distro is not comparable to RedHat. Hence, Solaris has a much larger "commercial support base" for people who care about Oracle.

Is this a joke? I've just finished reading the DOI FAQ and from what I can tell it doesn't provide any benefit over URLs.

Take your example: keeping track of articles. Couldn't the person assigning those DOI's just as easily assign a URL,
and at that url place a metadata and a link to the article's content?

In either case it still depends on the link to the content being kept up to date.