What we also really need is for people to understand that not only children play video games.

Yes, but the value in sites like /. is *EDITING*. They should filter shit like this so their s/n ratio is high, and thus their feed is valuable. This will bring subscribers and thus attention and thus ad impressions. At present, slashdot posts 90% crap found on other sites 4-9 days ago, and 7% crap like this and 3% interesting stuff, mostly hard-science stuff I'd find if I was still in school or in a non-CS/Real-Science job.

I don't know where you come from, but the typical users I encounter manage to put all what you listed on their desktop.

They tend to absolutely resist to learn more then the minimal basics, no matter how easy or accessible you make the UI.

Guess it's the 7 things rule: people can remember about 7 things at once / put them in context. Everything else is logical abstraction and training, and most users are not capable of the first and very reluctant to the second.

Go do volunteer basic computer literacy session for your local senior center. Don't try to convert them to linux or get them using Firefox or anything dumb like that. Just ask what their problems are, and how you can help. You will quickly understand how broken and unintuitive computer software is.

I don't think this will, but there's a much more compelling reason: Office Web Apps, which is mostly HTML/JS, with a few Silverlight bits (and even those have HTML/JS fallback). I've tried them now that they're in limited beta, and they work noticeably faster in Chrome and even Firefox than they do in IE8 - and Chrome/Firefox version isn't missing any features, either, so it's plainly better. I have no doubts that relevant teams in Microsoft are well aware of this, and understand how embarrassing it is, so I'd imagine there's a lot of pressure on IE team now to significantly improve performance - specially for JS - in the next release. Now that they have acceptable level of standard conformance (CSS 2.1 is finally fully supported, thank God), focusing on performance is the next logical step.

Ideally, that is true, but most users ARE used to MS style software, and packages like open office are very similar to word in their GUI at least. When software looks similar on the surface, but behaves differently, it confuses users. On the other hand, I disagree with the assertion that in general users see differences as faults. Some are though, and the big faults can easily dissuade users from even making a change.

I haven't actually used open office for a while, but a few years ago I TA'd an intro computer course for non-computer students. It's was typical easy-ish course (word, excel, basic HTML/CSS, some basic command line/ftp stuff in windows and linux), but we crammed in some harder stuff (some lectures on binary addressing, ram, caching) and we make them do some more obscure stuff in word (styles, sections, captions, table generation, cross referencing etc.) As part of one of the assignments, the prof asked them to check out open office and try one or two of the things we covered for Word and write a couple of paragraphs on them. There was a general agreement that the way OO handled captions was vastly superior to word, they were split on features like styles and somewhat indifferent to most of the regular word-processing features (most of which are basically identical). BUT, the first time my lab started up OO, there was a general sort of confusion because when the class double clicked the icon as instructed, they were greeted with a giant, blank grey screen. Once I told them they needed to go to the tiny menu in the corner and select to create a new word-processing document they were fine, but if they were on their own how many would have downloaded OO, seen the blank screen and through "hmmm...nothing opened...looks broken" and then promptly deleted it? Probably most. The article is correct - Linux and most of the mature OSS projects have very solid internals, they just need some non-developers to look at them to polish the externals up a bit.

When you think high tech, well networked states, you don't tend to think of there.

You evidently don't know how big the Moon is, or how much momentum is in its orbit around the Earth. Indeed, the Moon doesn't quite orbit the Earth, but rather the Moon and the Earth orbit one another around a center quite a ways away from the Earth's center. Or you just don't know how much energy can be produced by a nuke plant - a very tiny amount compared to what's needed to push the Moon out of orbit into the Earth in any appreciable amount of time.

But if you want to keep carrying on about some fact free paranoia, that's your business. Lunacy, but your business.

There is a difference between observing users and listening to users. The way to do usability testing is to watch lots of users work with the product and pay attention to the most common problems they have, but not necessarily to listen to what they say. If they say "I don't understand feature X", then fine. If they say "You know what would make this better, you should add feature Y", then you should probably ignore them. Users know what they hate, and they sometimes know what they don't understand, but they hardly every know how to design good software.

TFA is absolutely correct that the developer should watch and stay quiet during the process. (If you've ever been a developer in this situation, you know how incredibly painful and incredibly useful it is.) But the goal of the testing process isn't for the user to give you solutions, it is for the user to shine a spotlight on the problems. Once the problems are clearly understood, the developers (and designers) have to go back to work to figure out solutions.

JAVA will be removed from the *NASDAQ-100 composite index*, but will continue to trade as normal until the company is actually acquired. This point was even mentioned in the press release, so extra points for getting it so (so!) basically wrong.

(Man, /. just continues to accumulate fail. I wonder when it'll implode.)

Funny, it continues to work really well for me every week as I get incremental ugprades that keep my system functional, secure and quite current. I'm more regularly up to date than my big-bang-distro friends, certainly with respect to security issues. I don't have to devote weekends to "upgrading the server", crossing my fingers to make sure it works. And I have multiple options with respect to handling libraries and development packages, and a clear path to mixing stable and unstable packages, and tweaking them as need be. Oh, if I want to add (or remove) support for a feature (e.g., bluetooth), I can easily rebuild every package that provides support for that feature ... without getting the cruft of it when I don't need it.

Quite a "failure", indeed.

(To be clear: Gentoo does have social and architectural problems. But it does have some very compelling points, and the distro as a whole does keep moving on, quite usefully.)

udpeq attempts to solve exactly the latency problem. I've never used it, so I don't know how well it works, but the scheme seems to be this: you have a machine with a fast connection that uses UDP to balance traffic to and from the machine with several connections using UDP streams that get reconstituted on the end box.

You don't use one connection for one purpose. You use all connections for all purposes. Thus you use modem connections to dilute the high latency of the satellite connection toward the relatively lower latency of the modem connections.

I'm not sure how many low quality connections you would need to make online gaming enjoyable under this sort of scheme, but SSH should be usable with a modem or two. I'm not saying this isn't expensive. In addition to the satellite you need a dedicated phone line or two, plus an IP on a friend's machine with true broadband, but for people who live in remote areas, this is a workable solution till they get WiFi.