Note that the summary stated one of those shell games up front. It defines wealth as assets minus debts. The truly wealthy everywhere tend to have a lot of "debts", because they can borrow money easily and at lower rates than the rest of us. Here in the US, part of the regular stories about political corruption is corporations that give interest-free loans to politicians and to their top officers. Such loans are counted as "debt" and this study thus subtracts them from the recipients' assets. But an interest-free loan of, say, $100,000 - even if just stuck in a savings account - generates several percent interest that the "borrower" keeps. Those with inside access to money markets can make a lot more money. It doesn't take too many such loans to wipe out your apparent income and/or assets, making you look not terribly well off in such studies, while spinning off a lot of non-income money that you can use.
I've read a couple of interesting articles about the trouble this sort of thing causes for the eternal attempts to rate towns and county by wealth and/or income. The trouble comes from the fact that the truly wealthy hardly show up in the statistics at all. Just west of where I live are two of the wealthiest towns in the state, but on paper most of the people living there have little or no income and large debts. This study would probably rank both of them as middle-class towns, though the cheapest properties in both sell for several million. When they are sold at all, that is. Many of the properties are owned by paper corporations and "leased" by the residents (who coincidentally are the corporations' officers).
There are a lot of shell games in accounting when you're dealing with the truly wealthy.
I also recently had the fun of downgrading one of my machines to Redhat 7.2, so that I could easily work at home on a system that matched the "standard" for a project.
Of course, in the linux world, there's no salescritter pressure to constantly upgrade, mostly as a way of spending money. You only have to resist the jeers from the Fedora Core geeks.
But I'm not working on that project any more, so I think I'll install ubuntu on it instead. Time for something a little different. And I'm getting tired of all the hassles in getting things in non-Roman alphabets to work; I think the ubuntu folks are a bit farther along in that than most others. After all, they take the subject seriously.;-)
I'm not sure that many business folks would understand this argument. If a computer is doing what you need it to do, how can you be forced to do an upgrade? Microsoft can't order you to upgrade anything, and neither can Dell or whoever you bought it from.
The whole point here is that most "business" machines are never upgraded. No software inside is ever upgraded if it's doing the job you need. The only time a change will ever be made is if the hardware breaks and you can't get a replacement. OK, sometimes you fill the disk and need to get a bigger one, but you can get disks from small suppliers who won't demand an upgrade to anything else. When you find that you need something new done, and the old machine can't do it, then it's time to buy a completely new machine.
Upgrading something just because a salesman tells you you need to is hardly ever done. That costs money and time, it's a pure expense that brings in no money. The money is always better spent on something that effects the business. And even the most naive computer user knows well that upgrades always result in lost time and frustration as you try to learn how to get the new thingie to work as well as the old one did.
One way I've heard it expressed is "Which would you rather do, continue working around the bugs that you're familiar with, or spend time learning how to work around a changing set of bugs that is constantly surprising you with new failures?" This sort of question makes it obvious why so many people are highly resistant to upgrades to a machine that's sitting there doing its job.
Yet another reason everyone should be migrating to Unicode. Characters U+534D and U+5350 are swastikas. Why they're in the CJK Unified Ideograph section isn't too obvious, but they're there.
So, remind me which corporations of notable size are known to be early adopters?
Well, I recently finished a project at a rather large corporation (which I'll mercifully not name here) that hasn't quite finished upgrading all its W95 machines to W98. They also have a few NT machines, mostly in the IT dept.
No, I'm not joking. And this isn't the first case like this that I've seen.
Funny thing was that the project I worked on involved migrating software from a big IBM mainframe to a flock of distributed unix servers. Talk about having one foot in each world.
I'm not aware of any content that is available in [WMV] format only.
You obviously haven't been looking at much video on the web. There are lots of web sites, including some of the big-name "channels", that provide video clips only in WMV format.
For example, it used to be that comedycentral.com had clips in both WMV and Real formats; some months ago they dropped Real and went with WMV only.
Gaybuntu is not a distribution, but rather a forum
Ah. I was under the impression that the goal was a separate "gay" distro. I wasn't too clear on what that might entail, aside from the usual jokes about gays' aesthetic and fashion senses. But that didn't matter, as I'd argue that if a particular group thinks that they need a special distro tailored for them, then they probably do need it, and the rest of us shouldn't question their need. I mostly make this argument about the needs of disabled users, but there's no inherent reason to reject any group's claimed needs. I also use knoppix, which could be considered a specialized distro tailored for support and admin users, and it contains things that would baffle your average user.
I suppose we could extend the running gay jokes by suggesting that the gaybuntu people be encouraged to take the lead in UI development. Or have those jokes been overdone already? I know the Mac crowd is getting a bit tired of them.
And actually, the linux crowd has long since separated off UI details into the "themes" concept. So if you wanted some special "gay" themes, it's easy enough to do it. This would imply that the gaybuntu forum is aimed at something larger than the stereotypical UI stuff.
Programming is an example of a task that requires concentration. You have to simultaneously think about a list of things while writing in a language where the exact placement of commas, semicolons and parentheses can be critical. Any distraction at all kills the task you're working on, and forces you to start over when the distraction finally goes away. So it's not surprising that programmers might want and need solitude.
Add that to the fact that when they need to communicate with a human, the communication usually can't be handled verbally, because spoken languages don't contain the detail and precision needed. So verbal interchanges are mostly just social bonding, and the real communication takes place via email and comments in the code.
All this is why successful software projects tend to be done by distributed teams that rarely meet face to face.
Of course, this only applies to projects where success is measured by correctly-working code. If success is defined as a pretty UI or flashy stuff that helps make sales but doesn't need to give "correct" output, then it helps if people are working together in the same room.
It's telecommuting that's the dumb idea, and the managers all know it. Email and IM simply do not have the bandwidth of face-to-face communication.
Well, yes and no. As with the claim that cubicles and open-plan offices are always bad, this also depends on the task.
Historically, technical people have often collaborated very effectively via print media. The reason is well understood: There are a lot of technical concepts that can't be expressed easily in English or any other "human" language. To communicate effectively, you need to use a blackboard or a piece of paper - or email. Things like equations, diagrams and software can't be communicated effectively via a speech medium; they can only be expressed in writing.
I've seen this on a lot of projects. Very often, I end up just listening quietly in meetings, because it's obvious that people aren't communicating very well. Afterwards, I'll type up my analysis and suggestions, and email them. That's where the actual communication takes place. Then management wants a meeting to discuss things, and we have another meeting where people are talking past each other, and again I mostly sit and listen.
Note that I'm not claiming that this is always true. Some topics can be discussed verbally. And if the group's problems are mostly personal, verbal interactions can be the fastest way to get to the crux of the problems.
But saying that telecommuting is a dumb idea is itself a dumb idea, as bad as claiming that open office plans are always wrong. Some of the most effective computing projects have been done by groups that never meet face to face. I've done some successful projects with people that I've never met. And I've seen group meetups that were quite enjoyable and successful social occasions, but which didn't contribute at all to the project's progress.
It all depends on what you need to communicate, and what's the best language for that communication.
Attn gay Linux users: please join us at gaybuntu.com
As an aside, it seems to me that it could be useful to have similarly tailored releases that cater to assorted special groups. I'd bet there are other groups doing similar things. Is there a more general effort to coordinate such tailoring? I'd think that people working on such special distros could benefit greatly by talking to each other and developing general tools to support such efforts.
I've been getting tempted to switch my RedHat linux server over to ubuntu, just to get familiar with it. Getting mixed up in such an effort could be a motivation to sit down and do it.
That example I'd interpret as indicating incompetent programmers. In all languages I know, it's quite easy to write code that recognizes all of them (and also "1";-), mapping them into whatever internal code is needed. The fact that most programmers don't do this says a lot about any claims of user-friendliness.
The review itself states that sections of the book are likely to give/.'ers fits because it tries to dispel common viewpoints in the IT industry of Microsoft being an underhanded company with a grossly inferior product.
Well, I understood the review as saying that this is quite true, but it doesn't much matter because the other companies are also underhanded and have inferior products. They just aren't as good at underhanded marketing of inferior products as Microsoft is.
Did I read the review wrong?
(And linux really hasn't been much of a competitor mostly because it wasn't intended as a competitor in the underhanded, inferior-product market. It was designed to be useful to people who know what they're doing and want good tools for their jobs. Linux seems to do rather well in that admittedly small market. What's disappointing is that part of the linux crowd keeps trying to compete in the Microsoft/IBM market, leading them to adopt underhanded tactics and inferior [GUI] tools. I don't want something that mimics the MS/IBM approach, dammit; I want quality tools.;-)
It is so much easier to just have plain text config. files out there that you can simply edit with a text editor,...
It may be worth mentioning that when the guys at Bell Labs developed unix back in the 1970s, this was one of their significant technical advances. At the time, there were lots of OSs on the market, and the hot new UI technology was "full screen" displays. They were character displays at the time, with the charsets augmented with a few things for drawing boxes, but it was flashy new stuff at the time. Every application package typically had its own config editor (or often, several of them), and each had a complex UI that had to be learned. The actual config files were binary files, incomprehensible without the source code, and if you screwed something up, you usually just erased the config files and started over.
The unix guys came up with this radical new concept: plain-text config files. You could configure anything with a single tool: a text editor. The config files usually had a comment syntax (any line starting with '#' is ignored), so you could document each config setting right in the sample config file. You could present a list of options, all commented out, and a user could just delete the '#' for the option they wanted. It was fast, slick, and easy to use. It didn't require mastering a flock of flashy config tools, each of them different. You didn't have to wade through a complex mess of config screens ("windows" in today's terminology), hunting around for the screen that handled the thing you were trying to change. Just edit the config file, search for the name of the thing, and there it was.
Nowadays, even unix systems are going with the idea that users are such idiots that they can't handle editing a plain-text file. To make something commercially successful, you have to provide a special-purpose config editor for every app, and you have to present the user with a complex maze of config windows that they will never remember. But users demand this, it seems, because we keep telling them how difficult it is to edit a plain-text file. So we give them something that's much more difficult to use, but looks prettier on the screen, and pretend that we're helping them.
We've gone back 30 years in this topic, and are failing to learn some of the lessons that we learned back then.
I've long contended that the main reason apache is so successful is that it provides a plain-text config file that's full of documentation and useful examples. I can install apache in 15 to 20 minutes on a new machine, and it usually works the first time. If not, I don't have to trash the config file and start over; the error_log usually tells me the problem (often giving the line number). Another quick edit and a "restart" and it's up and running.
You'd think that other developers would have picked up on this, but this doesn't seem to happen often. I guess it's good for apache.
Interesting... I really dislike chocolate... apparently there's a genetic basis for this with regards to being able to pickup the taste of a normally tasteless chemical (couldn't pin down the name) found in chocolate.
There are a number of chemical taste differences like this known and explained by researchers. One is that there's a chemical in cooked cabbage that only about 1% of the population can taste, and to them it's really delicious. To the rest of us, cooked cabbage just tastes like, uh, cooked cabbage, and isn't particularly good or bad tasting. I think raw cabbage tastes better than cooked, so I'm obvious not one of the 1%.
Another funny example was an explanation I read once of the "New Coke" fiasco. They had done a randomized taste test and found that most people liked the new formulation better than the old Coca Cola. So they released it, and after the marketing disaster, they figured out why. The old Coke contained vanilla and was lower in sugar (to make up for vanilla's sweetness). The new formulation had vanillin, or "artificial vanilla". It turns out that between 5% and 10% of the population can taste vanilla. To the rest, "plain vanilla" pretty much means "bland and tasteless", but to those sensitive to it, vanilla is one of the world's best flavors (and is fairly sweet). It seems that Coca Cola's market wasn't random, but included mostly the vanilla-sensitive people. To everyone else, it was just a somewhat harsh cola, because of the lower sugar content. But to their market, the New Coke was merely sweet, and didn't include any vanilla, which was why they liked the old Coke. They went back to the old formula with real vanilla, and recovered their market share (which also includes vanilla insensitives who just like their cola less sweet).
I found this interesting because, despite my cynical attitude toward such things, I had always found Coke better tasting than the other colas. This story explained why. I'm obviously one of the vanilla-sensitive types. And I have a daughter who's among the vanilla super-sensitives. She has always puzzled her friends by chosing vanilla ice cream over chocolate, which is an insane choice to most people. She can detect vanilla at a hundred paces, and knows when she enters a room if there's vanilla ice cream present. But it has to be real vanilla, because the artificial stuff doesn't taste or smell much like vanilla.
Anyway, there's a growing list of such taste sensitivities that have been documented and explained by medical research. In some cases, the genetics behind the sensitivity is known, but that's mostly future research.
I've got four of the little bastards and I've never noticed them doing that. Now I'm gonna have to watch 'em for a while and try see it happen.
We found that it's easiest if you approach from the sleeping side, and walk quietly. Then sometimes you'll see the other eye wide open. If you start on the awake side, or make a noise, that side is likely to wake up the other side. Also, I don't know if all cockatiels do it, but ours seem to.
Are they doing this when they sleep on one foot with the other one tucked away?
Yeah, but I think that's how they always sleep. It's sort of a general bird thing. I don't think anyone knows why.
I know a number of people who don't consume caffeine, and I've yet to hear anyone call them "uncool" because of this.
One is my wife, who years ago learned that she was overly sensitive to caffeine, so she just avoids it. OTOH, she sees me drink a cup of coffee and then go off and take a nap. I once explained that it's because I'm just an insensitive male, and she agreed that that must be the explanation.
Her main complaint is that she can't eat chocolate with getting a strong reaction, and she loves chocolate. But she has learned to avoid it. This is difficult in restaurants, where most desserts contain chocolate.
Yeah, but there was the recent research showing that many birds sleep on just one side of their brain at a time.;-)
One thing I thought interesting about the report was that some of their test subjects were cockatiels. We have 'tiels, and we've often seen them apparently asleep, but if you can move to see their other eye, you find it open and alert (while the first one is still closed). Apparently part of what triggered the research was people reporting this sort of thing in their pet birds.
The researchers instrumented the birds' brains (with very light-weight instruments;-), and found that when the birds were in this state, with one eye open and one closed, their brain activity matched the pattern, with one hemisphere quiet and the other one active.
A curious aspect to this is that birds' eyes are, like ours, wired into both sides of the brain. But the "asleep on one side only" pattern exists, and matches the eyes.
The hypothesis is that birds generally don't need as much sleep as they get, so they stay half-alert to watch out for predators, maintain their grip on their perch, etc. And the alert side can send a wake-up signal to the sleeping side if anything interesting happens.
The reason ICANN wants to do lots of testing (after having dragged their feet for years before getting started) is that IDNs fundamentally change how DNS works, and it's really important not to break too much when you do that
Well, I've found that almost all of my software works just fine with UTF-8, without any patches at all. The explanation is simple: Most software that does anything at all with text does so by looking at specific special chars; all others are just carried along as "unrecognized" text that is only compared with other text strings. Usually the special chars are all 7-bit ASCII, so any char with the 8th bit set is "unrecognized". It turns out that usually this is equivalent to treating them as "letters", in the sense that all unrecognized strings of letters are just unprocessed text. The only problem is with code that uses the 8th bit for some special purpose, and even mediocre programmers know not to do that.
I did an instructive experiment a while back, in which I had some Chinese text in a bunch of HTML files, and I defined a CSS class whose name was the obvious Chinese character. (For the curious, it's ""="zhong1", which I've found that slashdot garbles somehow so it probably won't appear correctly on your screen.) I used this class to increase the font size of Chinese text to 160%. I was curious how many browsers would accept a CSS class with a Chinese name. I was unable to find a single browser that didn't accept it handle it correctly.
I was only slightly impressed by this. All it really meant was that the browsers didn't care what the "letters" in CSS class names are. They just took whatever was there that wasn't a special char in CSS, and used it as the name. So the 8-bit chars making up that Chinese char were just three "letters" to the browsers' parsers, with nothing special about them, and it all worked without even a warning message.
As far as I can tell, if you use UTF-8, there's no real reason any software except the low-level library code that draws characters on a screen or paper should ever have any problem with names in any language. If a program has a problem, it means that it's doing something clever (read: dumb) with the 8th bit.
So really, what problems would be caused for DNS if ICANN were to decree that the encoding is UTF-8, and all chars from 0x80 0xFF are to be treated as "letters"? I don't know of anything that DNS does that should be bothered by this. If there's a problem, is there a coherent explanation of it somewhere? Does some obscure DNS rule require using bit 8 for some purpose? They couldn't have been that dumb, could they?
It always kills me how people presume that if they can't think of a fix to a problem that there must not be a fix to a problem.
It's not so much that people can't think of solutions, but rather to so often people's idea of their own writing system is so entrenched that they refuse to solve problems.
You can see this even in the limited English writing system. The computer field has been plagued from the start with the O/0 and I/l/1 problems, and there isn't the slightest chance that any solution will ever be accepted. We also keep using fonts that confuse "d" with "cl" (as in clear vs dear) and "m" with "rn" (as in modem vs modern), even when we realize that the font causes a problem.
If we can't straighten out the ongoing screwups caused by these charset problems in English, how can we preach to the rest of the world about how easy their problems are?
Of course, the Unicode gang made the east Asian scripts even worse in this regard than even their writing system. Many characters have 2 or 3 different codes, and I think I saw one with the same glyph for 4 codes (but I could just be dreaming this;-). At least ASCII doesn't have 3 different codes for the letter 'b', for example.
So can someone explain, in a way that makes sense to a programmer who works with email all the time, how email and IM differ?
I mean, I've written a lot of email software, some of which does IM, and I have no idea what distinction is being made here. I've always thought that IM was just a marketing term for yet another implementation of email. It's a somewhat limited sort of email, of course, due to the limitations of the tiny little gadgets that it's implemented on. But that's not materially different from the limitations of the machines that we did email on back in the 1980s.
Why are people trying to pretend that IM is something different? What properties are they noticing that I'm failing to see? It can't be just the size limits; that's too stupid a distinction for even politicians and lawyers to attempt to make.
It can't the the rapid deliver or notification by a beep; we did that on computers before 1980 and we called it email (usually with a hyphen back then;-).
What is the distinction? Curious readers want to know...
You compose an email message and send it to your email server. The email server then figures out which server it needs to be delivered to based on the recipient....
From this, I'd conclude that you understand little of email in general, and completely misunderstand SMTP.
The RFCs that define SMTP don't talk about email servers. The primary intended implementation would attempt first to make a direct TCP link to the recipient machine, and if successful, the message would go directly from source to destination machine with no intermediate "server" machines.
The primary reason that email servers exist is that Microsoft's DOS systems at first couldn't do direct TCP connections to each other, because they couldn't run a background task to listen on an IP port. Or even if they could, the machines usually had only a modem internet connection, so most of they time they weren't connected to the internet at all, and attempting to connect to them would fail. So the server approach was added to SMTP to accommodate machines with such intermittent network connections.
Even now that many home users have always-on internet connections, there are still many who don't, so the server system is kept alive. And ISPs do like it, because storing all messages on their server lets them do commercially-useful things like scanning the messages for keywords, for use in targeted advertising campaigns. (And it also means that they can comply with government access requirements if necessary.)
But the idea that email always works by bouncing messages off servers is flat wrong. I routinely run a number of email agents (some of which I wrote myself as tools to diagnose network problems) that deliver email by connecting directly to the machine in the address, and hunt around for servers if that fails. If I were to send you a message from the machines that I work on most, you'd see only one "Received:" line in the headers, indicating that it reached you in one hop with no intermediate servers involved. Unless you're on a Microsoft system, of course, in which case you're still probably not running an SMTP listener, so my machines can't connect to your port 25. (People knowledgable in SMTP will now explain why you still might see only one "Received:" line.;-)
I'd go into more detail, but I can hear the readers falling asleep already...
It's doubtful that Christians will happen upon Gaybuntu and suddenly decide that Linux is simulataneously satanic, communist, Iranian, atheist, humanist, and Buddhist, and therefore needs to be banned by the government.
It's not? Then what am I doing using it?
Actually, I'm not Iranian, so I guess I wouldn't be interested anyway.
How dare you support Kubuntu! Ubuntu is supposed to be a Linux distribution, not a platform to push the agenda for a particular desktop. After all we don't have a FvwmBuntu, a FluxboxBuntu or a TwmBuntu either!
Yes, we do. But they're still in the closet. This is out of fear of attacks from the Gnome and KDE crowds. It appears that this fear is justified.
This very problem was, in effect, a major motive for the long development of the SGML standard, and its special cases like HTML and XML. It's also part of the thought behind acronyms like ASN-1 and UTF-8.
The difficulty of decoding computer files even a few years old is a problem that dates from the earliest days of computers. Programmers have been battling this problem all along. And we know a lot of solutions.
The only problem is getting people to use the solutions. This means fighting the natural tendency of management to discourage anything not aimed at improving this quarter's bottom line.
Slashdot Mirror
Shell games. ...
Note that the summary stated one of those shell games up front. It defines wealth as assets minus debts. The truly wealthy everywhere tend to have a lot of "debts", because they can borrow money easily and at lower rates than the rest of us. Here in the US, part of the regular stories about political corruption is corporations that give interest-free loans to politicians and to their top officers. Such loans are counted as "debt" and this study thus subtracts them from the recipients' assets. But an interest-free loan of, say, $100,000 - even if just stuck in a savings account - generates several percent interest that the "borrower" keeps. Those with inside access to money markets can make a lot more money. It doesn't take too many such loans to wipe out your apparent income and/or assets, making you look not terribly well off in such studies, while spinning off a lot of non-income money that you can use.
I've read a couple of interesting articles about the trouble this sort of thing causes for the eternal attempts to rate towns and county by wealth and/or income. The trouble comes from the fact that the truly wealthy hardly show up in the statistics at all. Just west of where I live are two of the wealthiest towns in the state, but on paper most of the people living there have little or no income and large debts. This study would probably rank both of them as middle-class towns, though the cheapest properties in both sell for several million. When they are sold at all, that is. Many of the properties are owned by paper corporations and "leased" by the residents (who coincidentally are the corporations' officers).
There are a lot of shell games in accounting when you're dealing with the truly wealthy.
I also recently had the fun of downgrading one of my machines to Redhat 7.2, so that I could easily work at home on a system that matched the "standard" for a project.
;-)
Of course, in the linux world, there's no salescritter pressure to constantly upgrade, mostly as a way of spending money. You only have to resist the jeers from the Fedora Core geeks.
But I'm not working on that project any more, so I think I'll install ubuntu on it instead. Time for something a little different. And I'm getting tired of all the hassles in getting things in non-Roman alphabets to work; I think the ubuntu folks are a bit farther along in that than most others. After all, they take the subject seriously.
I'm not sure that many business folks would understand this argument. If a computer is doing what you need it to do, how can you be forced to do an upgrade? Microsoft can't order you to upgrade anything, and neither can Dell or whoever you bought it from.
The whole point here is that most "business" machines are never upgraded. No software inside is ever upgraded if it's doing the job you need. The only time a change will ever be made is if the hardware breaks and you can't get a replacement. OK, sometimes you fill the disk and need to get a bigger one, but you can get disks from small suppliers who won't demand an upgrade to anything else. When you find that you need something new done, and the old machine can't do it, then it's time to buy a completely new machine.
Upgrading something just because a salesman tells you you need to is hardly ever done. That costs money and time, it's a pure expense that brings in no money. The money is always better spent on something that effects the business. And even the most naive computer user knows well that upgrades always result in lost time and frustration as you try to learn how to get the new thingie to work as well as the old one did.
One way I've heard it expressed is "Which would you rather do, continue working around the bugs that you're familiar with, or spend time learning how to work around a changing set of bugs that is constantly surprising you with new failures?" This sort of question makes it obvious why so many people are highly resistant to upgrades to a machine that's sitting there doing its job.
[T]here's no schwastika in most standard fonts.
Yet another reason everyone should be migrating to Unicode. Characters U+534D and U+5350 are swastikas. Why they're in the CJK Unified Ideograph section isn't too obvious, but they're there.
So, remind me which corporations of notable size are known to be early adopters?
Well, I recently finished a project at a rather large corporation (which I'll mercifully not name here) that hasn't quite finished upgrading all its W95 machines to W98. They also have a few NT machines, mostly in the IT dept.
No, I'm not joking. And this isn't the first case like this that I've seen.
Funny thing was that the project I worked on involved migrating software from a big IBM mainframe to a flock of distributed unix servers. Talk about having one foot in each world.
I'm not aware of any content that is available in [WMV] format only.
You obviously haven't been looking at much video on the web. There are lots of web sites, including some of the big-name "channels", that provide video clips only in WMV format.
For example, it used to be that comedycentral.com had clips in both WMV and Real formats; some months ago they dropped Real and went with WMV only.
As another poster said, you do NOT want to screw around with your visa in Japan. They are unbelievably unforgiving.
What? Are you saying that the opening scenes of megatokyo were wrong?
I was expecting a followup story describing his Mortal Kombat session with a ninja who works for Immigration Control.
Gaybuntu is not a distribution, but rather a forum
Ah. I was under the impression that the goal was a separate "gay" distro. I wasn't too clear on what that might entail, aside from the usual jokes about gays' aesthetic and fashion senses. But that didn't matter, as I'd argue that if a particular group thinks that they need a special distro tailored for them, then they probably do need it, and the rest of us shouldn't question their need. I mostly make this argument about the needs of disabled users, but there's no inherent reason to reject any group's claimed needs. I also use knoppix, which could be considered a specialized distro tailored for support and admin users, and it contains things that would baffle your average user.
I suppose we could extend the running gay jokes by suggesting that the gaybuntu people be encouraged to take the lead in UI development. Or have those jokes been overdone already? I know the Mac crowd is getting a bit tired of them.
And actually, the linux crowd has long since separated off UI details into the "themes" concept. So if you wanted some special "gay" themes, it's easy enough to do it. This would imply that the gaybuntu forum is aimed at something larger than the stereotypical UI stuff.
Programming is an example of a task that requires concentration. You have to simultaneously think about a list of things while writing in a language where the exact placement of commas, semicolons and parentheses can be critical. Any distraction at all kills the task you're working on, and forces you to start over when the distraction finally goes away. So it's not surprising that programmers might want and need solitude.
Add that to the fact that when they need to communicate with a human, the communication usually can't be handled verbally, because spoken languages don't contain the detail and precision needed. So verbal interchanges are mostly just social bonding, and the real communication takes place via email and comments in the code.
All this is why successful software projects tend to be done by distributed teams that rarely meet face to face.
Of course, this only applies to projects where success is measured by correctly-working code. If success is defined as a pretty UI or flashy stuff that helps make sales but doesn't need to give "correct" output, then it helps if people are working together in the same room.
It's telecommuting that's the dumb idea, and the managers all know it. Email and IM simply do not have the bandwidth of face-to-face communication.
Well, yes and no. As with the claim that cubicles and open-plan offices are always bad, this also depends on the task.
Historically, technical people have often collaborated very effectively via print media. The reason is well understood: There are a lot of technical concepts that can't be expressed easily in English or any other "human" language. To communicate effectively, you need to use a blackboard or a piece of paper - or email. Things like equations, diagrams and software can't be communicated effectively via a speech medium; they can only be expressed in writing.
I've seen this on a lot of projects. Very often, I end up just listening quietly in meetings, because it's obvious that people aren't communicating very well. Afterwards, I'll type up my analysis and suggestions, and email them. That's where the actual communication takes place. Then management wants a meeting to discuss things, and we have another meeting where people are talking past each other, and again I mostly sit and listen.
Note that I'm not claiming that this is always true. Some topics can be discussed verbally. And if the group's problems are mostly personal, verbal interactions can be the fastest way to get to the crux of the problems.
But saying that telecommuting is a dumb idea is itself a dumb idea, as bad as claiming that open office plans are always wrong. Some of the most effective computing projects have been done by groups that never meet face to face. I've done some successful projects with people that I've never met. And I've seen group meetups that were quite enjoyable and successful social occasions, but which didn't contribute at all to the project's progress.
It all depends on what you need to communicate, and what's the best language for that communication.
Attn gay Linux users: please join us at gaybuntu.com
As an aside, it seems to me that it could be useful to have similarly tailored releases that cater to assorted special groups. I'd bet there are other groups doing similar things. Is there a more general effort to coordinate such tailoring? I'd think that people working on such special distros could benefit greatly by talking to each other and developing general tools to support such efforts.
I've been getting tempted to switch my RedHat linux server over to ubuntu, just to get familiar with it. Getting mixed up in such an effort could be a motivation to sit down and do it.
That example I'd interpret as indicating incompetent programmers. In all languages I know, it's quite easy to write code that recognizes all of them (and also "1" ;-), mapping them into whatever internal code is needed. The fact that most programmers don't do this says a lot about any claims of user-friendliness.
The review itself states that sections of the book are likely to give /.'ers fits because it tries to dispel common viewpoints in the IT industry of Microsoft being an underhanded company with a grossly inferior product.
;-)
Well, I understood the review as saying that this is quite true, but it doesn't much matter because the other companies are also underhanded and have inferior products. They just aren't as good at underhanded marketing of inferior products as Microsoft is.
Did I read the review wrong?
(And linux really hasn't been much of a competitor mostly because it wasn't intended as a competitor in the underhanded, inferior-product market. It was designed to be useful to people who know what they're doing and want good tools for their jobs. Linux seems to do rather well in that admittedly small market. What's disappointing is that part of the linux crowd keeps trying to compete in the Microsoft/IBM market, leading them to adopt underhanded tactics and inferior [GUI] tools. I don't want something that mimics the MS/IBM approach, dammit; I want quality tools.
It is so much easier to just have plain text config. files out there that you can simply edit with a text editor, ...
It may be worth mentioning that when the guys at Bell Labs developed unix back in the 1970s, this was one of their significant technical advances. At the time, there were lots of OSs on the market, and the hot new UI technology was "full screen" displays. They were character displays at the time, with the charsets augmented with a few things for drawing boxes, but it was flashy new stuff at the time. Every application package typically had its own config editor (or often, several of them), and each had a complex UI that had to be learned. The actual config files were binary files, incomprehensible without the source code, and if you screwed something up, you usually just erased the config files and started over.
The unix guys came up with this radical new concept: plain-text config files. You could configure anything with a single tool: a text editor. The config files usually had a comment syntax (any line starting with '#' is ignored), so you could document each config setting right in the sample config file. You could present a list of options, all commented out, and a user could just delete the '#' for the option they wanted. It was fast, slick, and easy to use. It didn't require mastering a flock of flashy config tools, each of them different. You didn't have to wade through a complex mess of config screens ("windows" in today's terminology), hunting around for the screen that handled the thing you were trying to change. Just edit the config file, search for the name of the thing, and there it was.
Nowadays, even unix systems are going with the idea that users are such idiots that they can't handle editing a plain-text file. To make something commercially successful, you have to provide a special-purpose config editor for every app, and you have to present the user with a complex maze of config windows that they will never remember. But users demand this, it seems, because we keep telling them how difficult it is to edit a plain-text file. So we give them something that's much more difficult to use, but looks prettier on the screen, and pretend that we're helping them.
We've gone back 30 years in this topic, and are failing to learn some of the lessons that we learned back then.
I've long contended that the main reason apache is so successful is that it provides a plain-text config file that's full of documentation and useful examples. I can install apache in 15 to 20 minutes on a new machine, and it usually works the first time. If not, I don't have to trash the config file and start over; the error_log usually tells me the problem (often giving the line number). Another quick edit and a "restart" and it's up and running.
You'd think that other developers would have picked up on this, but this doesn't seem to happen often. I guess it's good for apache.
Interesting ... I really dislike chocolate ... apparently there's a genetic basis for this with regards to being able to pickup the taste of a normally tasteless chemical (couldn't pin down the name) found in chocolate.
There are a number of chemical taste differences like this known and explained by researchers. One is that there's a chemical in cooked cabbage that only about 1% of the population can taste, and to them it's really delicious. To the rest of us, cooked cabbage just tastes like, uh, cooked cabbage, and isn't particularly good or bad tasting. I think raw cabbage tastes better than cooked, so I'm obvious not one of the 1%.
Another funny example was an explanation I read once of the "New Coke" fiasco. They had done a randomized taste test and found that most people liked the new formulation better than the old Coca Cola. So they released it, and after the marketing disaster, they figured out why. The old Coke contained vanilla and was lower in sugar (to make up for vanilla's sweetness). The new formulation had vanillin, or "artificial vanilla". It turns out that between 5% and 10% of the population can taste vanilla. To the rest, "plain vanilla" pretty much means "bland and tasteless", but to those sensitive to it, vanilla is one of the world's best flavors (and is fairly sweet). It seems that Coca Cola's market wasn't random, but included mostly the vanilla-sensitive people. To everyone else, it was just a somewhat harsh cola, because of the lower sugar content. But to their market, the New Coke was merely sweet, and didn't include any vanilla, which was why they liked the old Coke. They went back to the old formula with real vanilla, and recovered their market share (which also includes vanilla insensitives who just like their cola less sweet).
I found this interesting because, despite my cynical attitude toward such things, I had always found Coke better tasting than the other colas. This story explained why. I'm obviously one of the vanilla-sensitive types. And I have a daughter who's among the vanilla super-sensitives. She has always puzzled her friends by chosing vanilla ice cream over chocolate, which is an insane choice to most people. She can detect vanilla at a hundred paces, and knows when she enters a room if there's vanilla ice cream present. But it has to be real vanilla, because the artificial stuff doesn't taste or smell much like vanilla.
Anyway, there's a growing list of such taste sensitivities that have been documented and explained by medical research. In some cases, the genetics behind the sensitivity is known, but that's mostly future research.
I've got four of the little bastards and I've never noticed them doing that. Now I'm gonna have to watch 'em for a while and try see it happen.
We found that it's easiest if you approach from the sleeping side, and walk quietly. Then sometimes you'll see the other eye wide open. If you start on the awake side, or make a noise, that side is likely to wake up the other side. Also, I don't know if all cockatiels do it, but ours seem to.
Are they doing this when they sleep on one foot with the other one tucked away?
Yeah, but I think that's how they always sleep. It's sort of a general bird thing. I don't think anyone knows why.
I know a number of people who don't consume caffeine, and I've yet to hear anyone call them "uncool" because of this.
One is my wife, who years ago learned that she was overly sensitive to caffeine, so she just avoids it. OTOH, she sees me drink a cup of coffee and then go off and take a nap. I once explained that it's because I'm just an insensitive male, and she agreed that that must be the explanation.
Her main complaint is that she can't eat chocolate with getting a strong reaction, and she loves chocolate. But she has learned to avoid it. This is difficult in restaurants, where most desserts contain chocolate.
Uh birds sleep too.
;-)
;-), and found that when the birds were in this state, with one eye open and one closed, their brain activity matched the pattern, with one hemisphere quiet and the other one active.
Yeah, but there was the recent research showing that many birds sleep on just one side of their brain at a time.
One thing I thought interesting about the report was that some of their test subjects were cockatiels. We have 'tiels, and we've often seen them apparently asleep, but if you can move to see their other eye, you find it open and alert (while the first one is still closed). Apparently part of what triggered the research was people reporting this sort of thing in their pet birds.
The researchers instrumented the birds' brains (with very light-weight instruments
A curious aspect to this is that birds' eyes are, like ours, wired into both sides of the brain. But the "asleep on one side only" pattern exists, and matches the eyes.
The hypothesis is that birds generally don't need as much sleep as they get, so they stay half-alert to watch out for predators, maintain their grip on their perch, etc. And the alert side can send a wake-up signal to the sleeping side if anything interesting happens.
The reason ICANN wants to do lots of testing (after having dragged their feet for years before getting started) is that IDNs fundamentally change how DNS works, and it's really important not to break too much when you do that
Well, I've found that almost all of my software works just fine with UTF-8, without any patches at all. The explanation is simple: Most software that does anything at all with text does so by looking at specific special chars; all others are just carried along as "unrecognized" text that is only compared with other text strings. Usually the special chars are all 7-bit ASCII, so any char with the 8th bit set is "unrecognized". It turns out that usually this is equivalent to treating them as "letters", in the sense that all unrecognized strings of letters are just unprocessed text. The only problem is with code that uses the 8th bit for some special purpose, and even mediocre programmers know not to do that.
I did an instructive experiment a while back, in which I had some Chinese text in a bunch of HTML files, and I defined a CSS class whose name was the obvious Chinese character. (For the curious, it's ""="zhong1", which I've found that slashdot garbles somehow so it probably won't appear correctly on your screen.) I used this class to increase the font size of Chinese text to 160%. I was curious how many browsers would accept a CSS class with a Chinese name. I was unable to find a single browser that didn't accept it handle it correctly.
I was only slightly impressed by this. All it really meant was that the browsers didn't care what the "letters" in CSS class names are. They just took whatever was there that wasn't a special char in CSS, and used it as the name. So the 8-bit chars making up that Chinese char were just three "letters" to the browsers' parsers, with nothing special about them, and it all worked without even a warning message.
As far as I can tell, if you use UTF-8, there's no real reason any software except the low-level library code that draws characters on a screen or paper should ever have any problem with names in any language. If a program has a problem, it means that it's doing something clever (read: dumb) with the 8th bit.
So really, what problems would be caused for DNS if ICANN were to decree that the encoding is UTF-8, and all chars from 0x80 0xFF are to be treated as "letters"? I don't know of anything that DNS does that should be bothered by this. If there's a problem, is there a coherent explanation of it somewhere? Does some obscure DNS rule require using bit 8 for some purpose? They couldn't have been that dumb, could they?
It always kills me how people presume that if they can't think of a fix to a problem that there must not be a fix to a problem.
;-). At least ASCII doesn't have 3 different codes for the letter 'b', for example.
It's not so much that people can't think of solutions, but rather to so often people's idea of their own writing system is so entrenched that they refuse to solve problems.
You can see this even in the limited English writing system. The computer field has been plagued from the start with the O/0 and I/l/1 problems, and there isn't the slightest chance that any solution will ever be accepted. We also keep using fonts that confuse "d" with "cl" (as in clear vs dear) and "m" with "rn" (as in modem vs modern), even when we realize that the font causes a problem.
If we can't straighten out the ongoing screwups caused by these charset problems in English, how can we preach to the rest of the world about how easy their problems are?
Of course, the Unicode gang made the east Asian scripts even worse in this regard than even their writing system. Many characters have 2 or 3 different codes, and I think I saw one with the same glyph for 4 codes (but I could just be dreaming this
So can someone explain, in a way that makes sense to a programmer who works with email all the time, how email and IM differ?
;-).
...
I mean, I've written a lot of email software, some of which does IM, and I have no idea what distinction is being made here. I've always thought that IM was just a marketing term for yet another implementation of email. It's a somewhat limited sort of email, of course, due to the limitations of the tiny little gadgets that it's implemented on. But that's not materially different from the limitations of the machines that we did email on back in the 1980s.
Why are people trying to pretend that IM is something different? What properties are they noticing that I'm failing to see? It can't be just the size limits; that's too stupid a distinction for even politicians and lawyers to attempt to make.
It can't the the rapid deliver or notification by a beep; we did that on computers before 1980 and we called it email (usually with a hyphen back then
What is the distinction? Curious readers want to know
You compose an email message and send it to your email server. The email server then figures out which server it needs to be delivered to based on the recipient. ...
;-)
...
From this, I'd conclude that you understand little of email in general, and completely misunderstand SMTP.
The RFCs that define SMTP don't talk about email servers. The primary intended implementation would attempt first to make a direct TCP link to the recipient machine, and if successful, the message would go directly from source to destination machine with no intermediate "server" machines.
The primary reason that email servers exist is that Microsoft's DOS systems at first couldn't do direct TCP connections to each other, because they couldn't run a background task to listen on an IP port. Or even if they could, the machines usually had only a modem internet connection, so most of they time they weren't connected to the internet at all, and attempting to connect to them would fail. So the server approach was added to SMTP to accommodate machines with such intermittent network connections.
Even now that many home users have always-on internet connections, there are still many who don't, so the server system is kept alive. And ISPs do like it, because storing all messages on their server lets them do commercially-useful things like scanning the messages for keywords, for use in targeted advertising campaigns. (And it also means that they can comply with government access requirements if necessary.)
But the idea that email always works by bouncing messages off servers is flat wrong. I routinely run a number of email agents (some of which I wrote myself as tools to diagnose network problems) that deliver email by connecting directly to the machine in the address, and hunt around for servers if that fails. If I were to send you a message from the machines that I work on most, you'd see only one "Received:" line in the headers, indicating that it reached you in one hop with no intermediate servers involved. Unless you're on a Microsoft system, of course, in which case you're still probably not running an SMTP listener, so my machines can't connect to your port 25. (People knowledgable in SMTP will now explain why you still might see only one "Received:" line.
I'd go into more detail, but I can hear the readers falling asleep already
It's doubtful that Christians will happen upon Gaybuntu and suddenly decide that Linux is simulataneously satanic, communist, Iranian, atheist, humanist, and Buddhist, and therefore needs to be banned by the government.
It's not? Then what am I doing using it?
Actually, I'm not Iranian, so I guess I wouldn't be interested anyway.
How dare you support Kubuntu! Ubuntu is supposed to be a Linux distribution, not a platform to push the agenda for a particular desktop. After all we don't have a FvwmBuntu, a FluxboxBuntu or a TwmBuntu either!
;-) here?)
Yes, we do. But they're still in the closet. This is out of fear of attacks from the Gnome and KDE crowds. It appears that this fear is justified.
(Lessee; do I need a
This very problem was, in effect, a major motive for the long development of the SGML standard, and its special cases like HTML and XML. It's also part of the thought behind acronyms like ASN-1 and UTF-8.
The difficulty of decoding computer files even a few years old is a problem that dates from the earliest days of computers. Programmers have been battling this problem all along. And we know a lot of solutions.
The only problem is getting people to use the solutions. This means fighting the natural tendency of management to discourage anything not aimed at improving this quarter's bottom line.