I'm pretty happy about where/. sits in the quality/cost continuum. Sure, it has typos, and sometimes bad posts get highly moderated, but OTOH its ads are pretty unobtrusive and it usually loads pretty fast. And new content appears really often and is usally interesting.
Another site I read often, Salon, has higher quality, more professional content, but its ads are also a lot more annoying and it doesn't get new content as fast as/..
To a first approximation, you get what you pay for.
I agree. I think another big problem is measurement. It's easy to count features and relatively easy to measure speed. But it's really hard, to measure ease of use or aesthetics. So the people involved in selling software hype the things they can measure, and the buyers compare those numbers because it's the only data they've got.
I'd like to amplify your comment a bit. Not only should the QA team have HCI expertise, so should the design team. It's really time-consuming and expensive, if not outright impossible, to add usability to a product that wasn't designed to be usable. It's much cheaper, more efficient, and more effective to think about usability from the start.
According to modern usage, you're right that the BBC is uploading spyware to people's computers. However, historically this would have been called downloading because it's (in most cases) from big computers at BBC to small desktop computers.
I would have said download. I guess I'm older than I thought.
I've seen several comments asking "Why do I have to use special paper?" so I'll reply at the top level.
I think the technical reason is that the special paper gives the pen absolute positioning on the page. If I take my optical mouse, lift it up, and put it down somewhere else on the desk, the mouse doesn't know it's at a different part of the desk. The pen does know where is is on the page if you do this to it, because of the special paper.
Of course, as someone else pointed out, that's not to say that there is no business reason for this as well...
That's interesting. The way you've described it, it sounds more like distributing a derivative work than an exercise of fair use.
It brings up lots of questions, like: Should be allowed to hire a 3rd party to edit my legitimite copy of copyrighted material? Presumably I am allowed to use tools to edit my copy, even tools that a 3rd party made. What about "smart" tools that automatically redact certain things? We couldn't build an automatic tool to reliably remove objectionable material today, but someday we probably will be able to. Should that be allowed?
It's all political posturing, and it's all bullshit. We must attack Al Qaeda in order to preserve our national security.
Yup, political posturing aplenty.
It has nothing to do with good vs. evil.
I don't think that's true. Yes, it has to do with a lot more things than just that, but I think that is a factor.
Good vs. evil is a psuedo-religious sham.
In what context do you mean? If you mean in the context of the "War on Terrorism", then I think you're partly right. Politicians are invoking Good and Evil to further their own agendas, as usual.
If you mean that good and evil are meaningless concepts, then I strongly disagree with you. But I'm getting kind of OT here, so I won't pursue that. (If you want to pursue it, followup.)
I'm not sure how much more conservation is realistic in California, considering that, per capita, it already uses less energy than almost every other state. (Exactly where it falls depends on which website you look at. According to the SF Chronicle, it's
#2 after Rhode Island. Lawrence Berekeley National Lab says it's
#4 overall, and #2 among states with more than 5M people (based on 1997 data).)
HCI is a broad field, and its practicioners have a a wide range of degrees, backgrounds, expertise, etc. Most either have degrees in psychology or in computer science (e.g., me), but some have degrees in art or design. Although HCI is not viewed well in some CS departments/schools, there are some where it's well-supported, such as U.C. Berkeley, Virginia Tech, Georgia Tech, and Maryland (College Park) (to name some I can think of off the top of my head). Also, there are a small but increasing number of schools that offer degrees in HCI. Carnegie Mellon offers a professional Master's and PhDs in HCI, for example, at the HCI Institute. (Full disclosure: I currently work at the HCII.) As another comment said, schools or departments of information science/technology are becoming more prevalent, and would provide a suitable background for HCI.
Then again, you don't necessarily need a degree in HCI, CS, or psych at all. For example, if you're coming from the programming side (as I suspect many here on/. are:) ), you could get a job building user interfaces, which is mostly programming with some HCI component. Then you could migrate pretty smoothly to doing higher-level, design type work, which would be more HCIish and less CSish.
As far as books, here are a few I like:
The Design of Everyday Things, by Don Norman.
Programming As If People Mattered, by Nathaniel Borenstein.
Dan Olsen and Ben Shneiderman have written good HCI/UI (user interface) books, too.
Almost everything you said is wrong.
Linux has some of the best desktops. I use WindowMaker on every machine and I install it as the default on every machine. Even people new to computers settle into it within a few minutes. It is far better than KDE/GNOME/Windows/MacOSX. I've never edited a single WM config file by hand either.
I think it's great that WindowMaker is working out well for you and for those around you.
Unfortunately, they are the absolute worst kind of people you could
ever sent to do desktop stuff.
What I think you really mean is that they are too interested in porductivity and not enough in interesting little icons. Well, most secretaries are interested in productivity too and they don't give a shit about GUI theories that spout all kind of ways to "interface with the user": they want a clean simple fast method of telling their computer what they want it to do next.
You're right: end users are interested in productivity and not in GUI
theories. But anyone who wants their software to be used by the
general public ought to be interested in GUI theories and in little
icons, because those are some of the tools you need to build
interfaces that are easy for people to use and that make people
productive. If you don't know how things like main memory, disk
drives, and networks work, good luck writing OS code. Similarly, if
you want to build good UIs, you need to know about people. And if
you're a developer developing software for the general public, you
can't just use introspection, because The User Is Not Like You.
There is a lot less anti-newbie feeling than there is a dislike of being told that useful and productive tools that need some time to master are less important than pandering to simpletons that can't handle difficult words like "directory". Explain again why "folder" makes more sense; particularly the bit where I open a folder and find more folders inside. Which metaphor are we using here?
Your claim that anti-newbie feeling isn't a problem is contradicted
when you refer to newbies as "simpletons that can't handle difficult
words like 'directory'." This is a perfect example of the attitude
many people have in the OSS community that inibits adoption of OSS by
the mainstream public. I don't know that people find the
term "folder" easier to use than "directory", but you don't either.
It seems to me that they would, since even though "folder" isn't a
perfect analogy, as you point out, it's still a more direct physical
metaphor than "directory".
But the point is that if you want to make your interface easy to use,
you shouldn't assume that just because you don't understand why
someone would find A easier than B, that they don't or they shouldn't.
The primary goal of good UI design is not to convince the user to do
things the "right" way, the goal is to build the interface so that
what the user does naturally is the right thing.
Assuming they wanted to get to that point, where their market is
shrinking and the hardware they use is grossly overpriced for the
performance and there hasn't been a new application of any note for a
decade.
Ok, if you don't care if the general public uses your software, you
can ignore everything I've written. But the original article was
about mainstream adoption of Linux, and I think many people in the
Linux (and OSS) community want it to be adopted more widely. It's
those people I'm addressing.
How true. It is much harder to get people to try thinking instead of just following the latest pronouncements of the Gates and Jobs of this world. Imagine if people using computers felt they had a chance of arranging their desktop to suit themselves instead of some expert with a joke degree in Human-Computer-Interfaces. Or even, Jobs-forbid! an actual choice in which desktop to use! Jesus Christ! The sky is falling, the users have choice; the unified user interface is under attack!
I think you drastically overestimate the number of people who care
about rearranging their desktop or who could possibly be bothered with
choosing among multiple types of desktop. You make the point before
that people want to be productive, and that's exactly right. People
do not want to spend time configuring their desktop, they want to get
their work done.
Personally, I vastly prefer X Windows over the alternatives because I
am an expert user and I love having a highly configurable desktop.
But for the vast majority of computer users, a much, much
more important feature is that a system come with good defaults out of
the box, and require no configuration.
Basically, to hell with you and to hell with people that want their
users to be good little sheep. Linux on the desktop does every work
related task I've had for four years now ranging from graphics to web
design to large document preparation to programming and if you want to
pretend it's not happening it's no skin off my nose. I'm not depending
on a financially insecure company with a terrble track record for
supporting its users when things get tough.
I think it's great that Linux is working well for you and I have no
desire to see you stop using it, or use it differently than you have
been. But Linux is not ready (yet) for most people to use it. And if
it's ever going to be, its developers need to get away from the
attitude "It's good enough for me, it should be good enough for
everyone" and look at why other people aren't using it and work on
changing to technology to fit the people and their needs.
- Khelder
P.S. I really want Linux and OSS to succeed, and I think this is a
huge stumbling block in the path of that success. I'd like to discuss
this more, so if anybody knows of fora where this is discussed,
follow-up to this post and let me know. Thanks.
Re:When will the real evolution of RTS arive?
on
HIstory of RTS Games
·
· Score: 1
Kohan: Immortal Sovereigns was also somewhat different from other RTSs. Units were grouped into companies that had internal structure (i.e., 4 front-line troops and 2 back-line). Units had morale and could retreat or rout. Perhaps most importantly, it cost resources not just to build units, but to maintain them.
It was too easy, except for the last mission, which was impossible, but otherwise it was very fun.
I know it's not exactly what you're after, but transparent process migration was part of the
Sprite operating system (developed between '84 and '92). Anyone planning to implement single-process hibernation/suspension/whatever-you-want-to-call-i t might want to check out their papers.
Other cool features of Sprite included a log-structured file system (yeah, everybody has one now, but they didn't 10+ years ago) and RAID.
They've been using a cheating detector for CS classes at Berkeley for over 10 years now. It builds an abstract syntax tree of programs and compares them. Its use was common knowledge. I think it caught some people every term.
I'm not positive, but I think the usual penalty for cheating was a zero on the assignment/exam or an F in the course.
So by extension, RMS should be considered a significant contributor to any program that's typically (or solely) compiled with gcc? That's absurd. It's like saying Leslie Lamport and Don Knuth are significant contributors to documents I compile with LaTeX.
Should the authors of theses great tools be given recognition for their efforts? Of course! That doesn't automatically make them "contributors" to projects that use their tools.
For more about why programs are getting less secure as time goes by, not more (and other interesting security-related topics, too), I highly recommend Secrets & Lies: Digital Security in a Networked World by Bruce Schneier. (/. had a review
of it last year.)
You're right that OSS advocates don't address these, but often you have the exact same issues with commercial software (at least if it's shrink-wrapped, off-the-shelf).
1. There's no support built into the product. Yes, you can hire people to support it for you, but it's a seperate cost.
2. Similarly, There is no warrantee of any kind. If it breaks, you have no one to complain to: "you get what you pay for."
With commercial software, this is mostly true. Some companies still offer free call-in tech support, but it seems like more and more companies are moving to model where you have to pay if you use up one of their employee's time. Many commercial vendors do provide web sites with searchable tech support info, but the same kind of thing is available for lots of OSS, too.
3. The programmers may suddenly decide they have no vested interest in continuing the project, or development may slow to a crawl (eg, mozilla), and there's nothing you can do about it.
The same thing can happen with commercial software, too. Not for Micros~1 products, maybe, but there are commercial software vendors that go out of business or discontinue products.
If you're talking about custom commercial software, it's a whole different ball game. None of the above 3 issues are problems, but you pay through the nose. In a lot of these cases, though, I would think you'd be better off with an OSS-based custom system, since then you're not tied to the people who did the customizing.
For some companies, for some applications, OSS makes sense. For others, they're better off with commercial software.
I think iabervon's comment had nothing to do with the difference
between file type and creator. Either or both of those can be used to
determine what application to use on a file. I think his main point
was right on: How useful file metadata is to you depends on how you
use your files.
As iabervon said, CLI (e.g., Unix) users normally specify the program
to use on a file explicitly. Metadata is not required. If you always
use the same application on the same file types, this is not efficient
for the user, but if you use many different applications on the same
file, then it isn't a big deal (and may even be more efficient than the GUI strategy).
OTOH, GUI (e.g., Mac) users most often double-click on a file to
operate on it, and let the system choose the right application to
use. In this case, the system needs metadata to figure out the right
application (e.g., type, creator, or some combination of the two).
[from iabervon]
Incidentally, has anyone else noticed that the MacOS scheme is
equivalent to having 4 character extensions which aren't displayed,
with the corresponding problem of having malicious executables named
README.txt (or even README)?
I'm not sure about this since I rarely use Macs, but doesn't the Mac
show different icons depending on the file type? A Mac user would not
mistake a non-text file for a text file since it wouldn't have the
"text file" icon.
I guess whether you think something removes or adds restrictions depends on what you think the default state of the world is. It seems to me that the default state is that any two parties can enter into any contract they both agree to. We as a society have decided that certain types of contracts are not legally binding. For example, a contract to kill someone is not binding, nor is a contract to give up all personal freedom and become a slave.
If you tell two parties that they may not enter into a legally binding contract where party A gives party B money (for example) in exchange for using something party B owns, possibly with usage limitations that both parties agree to, that is a restriction on parties A and B.
I'm not saying that all possible contracts between A and B are ethical, good for society, etc., but I think we should be clear that if we forbid certain types of software licenses (for example), that we are restricting people's actions.
The subject of my post is not the concept of a scholarly journal devoted to computer games--I think it's a great idea. What I want to criticize is this implementation of the idea.
I read several of the papers on the web site, and was very unimpressed. They're about on the level of/. Articles: one person's opinion of some aspect of games. It's possible that my problem with it is that I don't know filmspeak or whatever jargon they're using, but I don't see the point in some of them, and the theses that I can find seem to be obviously false if you've actually played a lot of games.
One article lauds The Sims and bashes fantasy games because The Sims is about people and fantasy games are not (so the author says). It's nonsense to say that no fantasy games are about people. Planescape: Torment, for example, raises issues of mortality, ethics, and identity.
Another article claims that nodramatically compelling games have yet been written. It never explicitly defines dramatically compelling, but says that no games "offer captivating narrative". Considering that I know people (and have been one) who sometimes watch people play some games (Final Fantasy, Baldur's Gate II, Fallout, Zork: Nemesis, just to name a few) for the purpose of seeing how the plot (or narrative) will unfold, I'd say this proposition is clearly false.
I would like to see a high-quality academic journal about games, because I think there's a lot more to them than they are given credit for. Unforutnately, this publication isn't it.
I was impressed with the quick thinking of a friend of mine years ago during a demo. He was demoing a system he'd built and been working with a long time. Near the end of the demo, there was a delay in the system's response, and he could smell that it was crashing. So he moved the mouse cursor over to the "Quit" button on the screen, and when the program quietly died everyone thought he'd pressed "Quit" instead.
I think it's interesting how often people on/. say things like the poster of this article:
There is no end to what people will pirate(and I personally don't mind that one bit).
At least, when the material pirated is owned by someone they don't like (e.g., TV studios).
But when it comes to material copyrighted by someone they do like (e.g., pretty much any GPLed software), they scream like stuck pigs when there's even a suggestion that it has been pirated.
Slashdot Mirror
I'm pretty happy about where /. sits in the quality/cost continuum. Sure, it has typos, and sometimes bad posts get highly moderated, but OTOH its ads are pretty unobtrusive and it usually loads pretty fast. And new content appears really often and is usally interesting.
/..
Another site I read often, Salon, has higher quality, more professional content, but its ads are also a lot more annoying and it doesn't get new content as fast as
To a first approximation, you get what you pay for.
I agree. I think another big problem is measurement. It's easy to count features and relatively easy to measure speed. But it's really hard, to measure ease of use or aesthetics. So the people involved in selling software hype the things they can measure, and the buyers compare those numbers because it's the only data they've got.
I'd like to amplify your comment a bit. Not only should the QA team have HCI expertise, so should the design team. It's really time-consuming and expensive, if not outright impossible, to add usability to a product that wasn't designed to be usable. It's much cheaper, more efficient, and more effective to think about usability from the start.
Does anyone besides me not equate legal with ethical, moral, righteous, virtous, etc.?
By Cringley's logic, if I do something vile that happens not to be illegal, I am not a villian. This is absurd.
</rant>
Good point, but I think you mean Fred Brooks. Great book, BTW.
Then it's just fun.
According to modern usage, you're right that the BBC is uploading spyware to people's computers. However, historically this would have been called downloading because it's (in most cases) from big computers at BBC to small desktop computers.
I would have said download. I guess I'm older than I thought.
I've seen several comments asking "Why do I have to use special paper?" so I'll reply at the top level.
I think the technical reason is that the special paper gives the pen absolute positioning on the page. If I take my optical mouse, lift it up, and put it down somewhere else on the desk, the mouse doesn't know it's at a different part of the desk. The pen does know where is is on the page if you do this to it, because of the special paper.
Of course, as someone else pointed out, that's not to say that there is no business reason for this as well...
That's interesting. The way you've described it, it sounds more like distributing a derivative work than an exercise of fair use.
It brings up lots of questions, like: Should be allowed to hire a 3rd party to edit my legitimite copy of copyrighted material? Presumably I am allowed to use tools to edit my copy, even tools that a 3rd party made. What about "smart" tools that automatically redact certain things? We couldn't build an automatic tool to reliably remove objectionable material today, but someday we probably will be able to. Should that be allowed?
Yup, political posturing aplenty.
It has nothing to do with good vs. evil.
I don't think that's true. Yes, it has to do with a lot more things than just that, but I think that is a factor.
Good vs. evil is a psuedo-religious sham.
In what context do you mean? If you mean in the context of the "War on Terrorism", then I think you're partly right. Politicians are invoking Good and Evil to further their own agendas, as usual.
If you mean that good and evil are meaningless concepts, then I strongly disagree with you. But I'm getting kind of OT here, so I won't pursue that. (If you want to pursue it, followup.)
Other info: State Energy Data Report for 1999 from the DOE. (The 2000 report isn't due out until Dec. 2002.)
Then again, you don't necessarily need a degree in HCI, CS, or psych at all. For example, if you're coming from the programming side (as I suspect many here on /. are :) ), you could get a job building user interfaces, which is mostly programming with some HCI component. Then you could migrate pretty smoothly to doing higher-level, design type work, which would be more HCIish and less CSish.
As far as books, here are a few I like:
- The Design of Everyday Things, by Don Norman.
- Programming As If People Mattered, by Nathaniel Borenstein.
Dan Olsen and Ben Shneiderman have written good HCI/UI (user interface) books, too.If you want to see what the cutting edge of HCI is, check out proceedings and journals, such as the ACM conference on HCI (Human Factors in Computing Systems, a.k.a. SIGCHI) or the ACM Symposium on User Interfaces Software and Technology (UIST).
But the point is that if you want to make your interface easy to use, you shouldn't assume that just because you don't understand why someone would find A easier than B, that they don't or they shouldn't. The primary goal of good UI design is not to convince the user to do things the "right" way, the goal is to build the interface so that what the user does naturally is the right thing.
Ok, if you don't care if the general public uses your software, you can ignore everything I've written. But the original article was about mainstream adoption of Linux, and I think many people in the Linux (and OSS) community want it to be adopted more widely. It's those people I'm addressing. I think you drastically overestimate the number of people who care about rearranging their desktop or who could possibly be bothered with choosing among multiple types of desktop. You make the point before that people want to be productive, and that's exactly right. People do not want to spend time configuring their desktop, they want to get their work done. Personally, I vastly prefer X Windows over the alternatives because I am an expert user and I love having a highly configurable desktop. But for the vast majority of computer users, a much, much more important feature is that a system come with good defaults out of the box, and require no configuration. I think it's great that Linux is working well for you and I have no desire to see you stop using it, or use it differently than you have been. But Linux is not ready (yet) for most people to use it. And if it's ever going to be, its developers need to get away from the attitude "It's good enough for me, it should be good enough for everyone" and look at why other people aren't using it and work on changing to technology to fit the people and their needs.- Khelder
P.S. I really want Linux and OSS to succeed, and I think this is a huge stumbling block in the path of that success. I'd like to discuss this more, so if anybody knows of fora where this is discussed, follow-up to this post and let me know. Thanks.
It was too easy, except for the last mission, which was impossible, but otherwise it was very fun.
Other cool features of Sprite included a log-structured file system (yeah, everybody has one now, but they didn't 10+ years ago) and RAID.
I'm not positive, but I think the usual penalty for cheating was a zero on the assignment/exam or an F in the course.
So by extension, RMS should be considered a significant contributor to any program that's typically (or solely) compiled with gcc? That's absurd. It's like saying Leslie Lamport and Don Knuth are significant contributors to documents I compile with LaTeX.
Should the authors of theses great tools be given recognition for their efforts? Of course! That doesn't automatically make them "contributors" to projects that use their tools.
For more about why programs are getting less secure as time goes by, not more (and other interesting security-related topics, too), I highly recommend Secrets & Lies: Digital Security in a Networked World by Bruce Schneier. (/. had a review of it last year.)
1. There's no support built into the product. Yes, you can hire people to support it for you, but it's a seperate cost.
2. Similarly, There is no warrantee of any kind. If it breaks, you have no one to complain to: "you get what you pay for."
With commercial software, this is mostly true. Some companies still offer free call-in tech support, but it seems like more and more companies are moving to model where you have to pay if you use up one of their employee's time. Many commercial vendors do provide web sites with searchable tech support info, but the same kind of thing is available for lots of OSS, too.
3. The programmers may suddenly decide they have no vested interest in continuing the project, or development may slow to a crawl (eg, mozilla), and there's nothing you can do about it.
The same thing can happen with commercial software, too. Not for Micros~1 products, maybe, but there are commercial software vendors that go out of business or discontinue products.
If you're talking about custom commercial software, it's a whole different ball game. None of the above 3 issues are problems, but you pay through the nose. In a lot of these cases, though, I would think you'd be better off with an OSS-based custom system, since then you're not tied to the people who did the customizing.
For some companies, for some applications, OSS makes sense. For others, they're better off with commercial software.
http://safe.millennium.berkeley.edu/
The page also has links to other similar sites.
As iabervon said, CLI (e.g., Unix) users normally specify the program to use on a file explicitly. Metadata is not required. If you always use the same application on the same file types, this is not efficient for the user, but if you use many different applications on the same file, then it isn't a big deal (and may even be more efficient than the GUI strategy).
OTOH, GUI (e.g., Mac) users most often double-click on a file to operate on it, and let the system choose the right application to use. In this case, the system needs metadata to figure out the right application (e.g., type, creator, or some combination of the two).
[from iabervon] Incidentally, has anyone else noticed that the MacOS scheme is equivalent to having 4 character extensions which aren't displayed, with the corresponding problem of having malicious executables named README.txt (or even README)?
I'm not sure about this since I rarely use Macs, but doesn't the Mac show different icons depending on the file type? A Mac user would not mistake a non-text file for a text file since it wouldn't have the "text file" icon.
If you tell two parties that they may not enter into a legally binding contract where party A gives party B money (for example) in exchange for using something party B owns, possibly with usage limitations that both parties agree to, that is a restriction on parties A and B.
I'm not saying that all possible contracts between A and B are ethical, good for society, etc., but I think we should be clear that if we forbid certain types of software licenses (for example), that we are restricting people's actions.
I read several of the papers on the web site, and was very unimpressed. They're about on the level of /. Articles: one person's opinion of some aspect of games. It's possible that my problem with it is that I don't know filmspeak or whatever jargon they're using, but I don't see the point in some of them, and the theses that I can find seem to be obviously false if you've actually played a lot of games.
One article lauds The Sims and bashes fantasy games because The Sims is about people and fantasy games are not (so the author says). It's nonsense to say that no fantasy games are about people. Planescape: Torment, for example, raises issues of mortality, ethics, and identity.
Another article claims that no dramatically compelling games have yet been written. It never explicitly defines dramatically compelling, but says that no games "offer captivating narrative". Considering that I know people (and have been one) who sometimes watch people play some games (Final Fantasy, Baldur's Gate II, Fallout, Zork: Nemesis, just to name a few) for the purpose of seeing how the plot (or narrative) will unfold, I'd say this proposition is clearly false.
I would like to see a high-quality academic journal about games, because I think there's a lot more to them than they are given credit for. Unforutnately, this publication isn't it.
I was impressed with the quick thinking of a friend of mine years ago during a demo. He was demoing a system he'd built and been working with a long time. Near the end of the demo, there was a delay in the system's response, and he could smell that it was crashing. So he moved the mouse cursor over to the "Quit" button on the screen, and when the program quietly died everyone thought he'd pressed "Quit" instead.
But when it comes to material copyrighted by someone they do like (e.g., pretty much any GPLed software), they scream like stuck pigs when there's even a suggestion that it has been pirated.
Clearly, consistency is overrated...