I can't rightly claim that I read the article, I can say that I tried though. Between the cluelessness and and the rambling, it was pretty tough going.
Dilger (a writing teacher) sounds like a typical customer who senses the need for an engineering solution.
He correctly senses that something's wrong.
Rather than explain what he sees as the symptoms of the problem, he diagnoses the problem and offers solutions. The broad premise is that users of modern GUIs don't understand enough about computer/network system structure to make effective use of current systems. His diagnoses and treatments, on the other hand, seem way off the mark.
For instance, he says that "ease of use" is the problem. Clearly silly. A well designed system should provide both an easy path to get the simple task done, and a flexible path, perhaps one that requires more dexterity, to perform more complex tasks. Such system designs go quite far back into our history.
Consider the text editor.
A short digression:
25 or 30 years ago, any text editor was considered a luxury, most users poked holes in cards or ribbons of paper (and before that, they juggled patch cables). Eventually, we had the soft luxury of keying our input to magnetic media, disk or tape, still in batch mode. Any ability to edit a line of text, without retyping a card, was considered ease of use.
In time, terminals were connected directly to computers, and people could actually interact with them.
Early interactive editors were designed
for paper-output ttys or for "glass ttys" which may not have provided any random access to screen locations, and thus, no
wysiwyg. In those days, wysiwyg was considered "ease of use" as GUIs are today, and many hardcore hackers considered wysiwyg users as soft. This spirit of ease of use is reflected in the following classic tale: (Ken Thompson is a father of UNIX, and its original editor, ed. For a long time, ed would respond any errant input with a simple "?" - in fact, many users found this sufficient.)
Ken Thompson has an automobile which he helped design. Unlike most
automobiles, it has neither speedometer, nor gas gage, nor any of the
numerous idiot lights which plague the modern driver. Rather, if the
driver makes any mistake, a giant "?" lights up in the center of the
dashboard. "The experienced driver", he says, "will usually know
what's wrong."
Anyway, back to the point. In time, powerful editors emerged, teco, ed, and variants. Eventually, xy-addressable crt's arose, and folks wrote extensions of these editors to provide user-friendly wysiwyg (emacs for teco, vi for ed).
The point of this digression: If you look at the design of emacs or vi, you will find that they provide a wysiwyg interface and a command line interface as well! A user-friendly interface and a separate more powerful interface with iterative control and so forth. We can have our cake and eat it too. Software should be powerful and easy to use. What is infuriating about most modern software isn't that it makes simple tasks easy, it's that it makes complex tasks terribly difficult.
Well, I started hacking UNIX at Bell Labs in 1978, when I completed my undergrad studies at the tender age of 19, which might make me slightly less of an old fart than you might imagine, but probably not.
Note that in those days, the UNIX community was very much like today's free software community. UNIX was closed source in a strict sense, but there really wasn't much of a commercial market for it, so most users were in the research and academic community, and so had source access, and there was lots of sharing of code and fixes.
The equivalent of today's Linux vs. Windows battle was either UNIX vs. IBM or UNIX vs. VMS.
Around that time, there was also the unfortunate splintering battle within UNIX, AT&T vs. BSD,
not to mention the battles within AT&T, between Research, USG, PWB, and Columbus.
Sort of like Red Hat, Debian, SuSE, Slackware, Mandrake, et al. Such splits (among allies) have always been a bad thing, and I believe they shall remain so.
No, that wasn't Andrew Tanenbaum, the Minix guy. That was me, I spell my name Tannenbaum. Tanenbaum was around at that time, of course. He would sometimes come to Bell Labs in the summers to work with the CS research folks and we'd get each other's mail.
I was an enthusiastic poster of netnews in those days, when you could read all the netnews, and recognise the names of most everyone who posted.
trb (formerly floyd!trb, harpo!trb, trb@mit-mc, et al).
As multics was a large bloated behemoth, unix (eunuchs) was intended for smaller networks (at the time)
UNIX was born in 1970. There were really no "networks" at that time (that is, no computers linked by what we think of as networks). For what networks there were, see this atlas. (Or perhaps you simply meant smaller machines?)
I'm wit that. I could never figure out the point of a GUI file manager. The text is the important part! The stupid little icons are useless noise. If you want to see an intelligently designed UI where you you can actually do powerful things with mouse and text, see Rob Pike's Acme UI from Plan 9. Other Plan 9 papers here.
I was also disturbed by the term "General Public License." I'm surprised this thread is being moderated down, I find expurgation pretty offensive. Why not just go and edit an author's licensing terms out of his sources? Moderate me down, there's a difference between karma points and karma.
They delivered 335,000 names on 60,000 pages? That's fewer than six names per page. Seems pretty inefficient to me, that should have fit on a few floppies, certainly on a single zip disk.
If you want to move your desktop workspace from place to place, it's the file system that you want to carry, not the processor package. Such portable disk solutions already exist.
Try a different approach. Instead of fretting about whether or not the uninformed teacher is treating you with respect, find a software engiineer in your neighborhood who is willing to come to your school and lecture your class on software licences, open source, and everything. Teachers like when professionals from their community come to tell the kiddies about the real world. What the hell, maybe you can invite the local press too, and make it an event.
The press release, dated Thursday, 27-January, says: "It will take several days to complete the processing and the researchers do not expect to have confirmation of a signal until some time next week." This is not an open-ended quest, like the search for strange new worlds, this is a week-long data reduction task. I don't see why you would need a large coordinated effort like distributed.net.
You guys are missing the point. The feds don't want the cleartext contents of his disk, they want the key! It's probably something like DoJsux or the name of one of his old girlfriends, and he doesn't want to embarrass himself. Haven't you ever had a password like that?
It seems that most hackers are quite clueless about this stuff. I tried to post a fat response to slashdot but it killed my browser (oops). I fished the data out of the cache and stuck it in a web page. Here's some of the fruit of my experience with incentive stock options. I hope this helps.
It doesn't make sense to tweak your kernel unless your tweaks are going to make a difference. Or, to put it in more nerdly terms, it's pointless to optimize unless you know where your bottlenecks are. At the price of memory these days, it's just dumb to be tossing out a feature that only takes up 3kb of memory.
At the suggestion of this article, I took a look at my config, and I found that almost everything that's optional was dynamically linked anyway.
I have a PII processor. Do I want to change my processor type? I don't know. What does it buy me to change from 386 to PPro? Is it worth the trouble? I assume that 99% of people are running the default 386. Might there be some subtle kink in the PPro kernel that isn't as well tested? Hmmm.
There's all these little features that I can turn off. How much memory am I saving with each one? It would be nice to know. Is it worth it to turn off a dozen features to save 64k on my 64Mb box? Nope. What kind of gains can a typical user expect from a reconfig?
Last lick. Of course, It's useful to have a kitchen sink kernel so that people can boot it on random hardware, but once you've booted, Linux knows what's there, and writes a/var/log/dmesg file with all that useful and mysterious info about your very own system. How about a prog that takes this custom data and builds a lean Linux kernel config from it?
It may be that this modem is "spoofing" some protocols, that is, looking at the data contents and doing something smart with it. This was originally done, I think, by Telebit 9600 baud modems, which did UUCP spoofing, and is also done by current fax modems, which may spoof certain handshaking. This typically requires pairs of modems that act as a black box, looking normal on the outside, and spoofing on the inside.
While I'm not a software design patterns maven, I am a big fan of Christopher Alexander (the source of many ideas about software design patterns) and his design ideas.
Alexander suggests that the designs you produce should be whole and harmonious and appropriate, and they should fit in their environment. So it seems ironic to me that advocates of such a wholistic ideal should tempt people with a reward that people find so irritating. This indicates that something is wrong with the design of the contest.
I don't think a political party is an integral part of government, it is a temporary body that occupies government. Government is a container, and it's occupied by party members. If the current parties disappear, the government survives quite nicely.
If gop.gov is allowed, then so should every political party, great and small, be allowed a foo.gov domain. The biggest parties are not cast in stone, neither their number nor their identification. There is nothing special about them, except that they are big and rich. They were not created with the nation, they were late comers. As such, why do they have power to demand special treatment (like a.gov) any more than libertarian, reform, communist, socialist, green, et al?
I can't see how this would happen. If there would be a cross polination, and the terminator gene was transferred, well, then the product would not survive and the gene would not spread. To me it sounds like a terminator gene is much safer than any other kind of gene manipulation.
Let's say some mutation of the terminator seed is only 75% effective, and it weakens the ability of crops to reproduce instead of terminating them completely. This bug gets mixed into the normal strains. Get the picture now?
It's either 1280x768 (16:9) or 1024x768 (4:3), see the Pioneer web site. As for a 1600x200 21" computer monitor being better or worse, note that there is a big difference between TV monitors and computer monitors. The phosphors are formulated with different goals in mind. You sit 18 inches (.5 meters) away from your computer screen all day. You sit 10 feet (3 meters) away from your TV. If your computer screen was as bright as your TV screen, your eyes would be fried in short order. TV's are 60 Hz (or 50) interlaced. Computer screens are typically 60 Hz or more, non-interlaced - this all affects the amount of flicker. On your computer, you are looking mostly at stationary text and images. In the olden days (before windows), there used to be lots of text scrolling, and long persistence phosphors would leave trails. This would be a problem with lots of motion as in modern games.
Plasma monitors can't be compared head to head with CRTs based simply on resolution, there are lots of other variables.
This is from the Journal of Irreproducible Results, (www.jir.com), Volume 36, Number 1. I think it was published about ten years ago, and is reproduced widely on the net, and as such, isn't terribly newsworthy.
Slashdot Mirror
Dilger (a writing teacher) sounds like a typical customer who senses the need for an engineering solution. He correctly senses that something's wrong. Rather than explain what he sees as the symptoms of the problem, he diagnoses the problem and offers solutions. The broad premise is that users of modern GUIs don't understand enough about computer/network system structure to make effective use of current systems. His diagnoses and treatments, on the other hand, seem way off the mark.
For instance, he says that "ease of use" is the problem. Clearly silly. A well designed system should provide both an easy path to get the simple task done, and a flexible path, perhaps one that requires more dexterity, to perform more complex tasks. Such system designs go quite far back into our history.
Consider the text editor. A short digression: 25 or 30 years ago, any text editor was considered a luxury, most users poked holes in cards or ribbons of paper (and before that, they juggled patch cables). Eventually, we had the soft luxury of keying our input to magnetic media, disk or tape, still in batch mode. Any ability to edit a line of text, without retyping a card, was considered ease of use. In time, terminals were connected directly to computers, and people could actually interact with them.
Early interactive editors were designed for paper-output ttys or for "glass ttys" which may not have provided any random access to screen locations, and thus, no wysiwyg. In those days, wysiwyg was considered "ease of use" as GUIs are today, and many hardcore hackers considered wysiwyg users as soft. This spirit of ease of use is reflected in the following classic tale: (Ken Thompson is a father of UNIX, and its original editor, ed. For a long time, ed would respond any errant input with a simple "?" - in fact, many users found this sufficient.)
Anyway, back to the point. In time, powerful editors emerged, teco, ed, and variants. Eventually, xy-addressable crt's arose, and folks wrote extensions of these editors to provide user-friendly wysiwyg (emacs for teco, vi for ed).
The point of this digression: If you look at the design of emacs or vi, you will find that they provide a wysiwyg interface and a command line interface as well! A user-friendly interface and a separate more powerful interface with iterative control and so forth. We can have our cake and eat it too. Software should be powerful and easy to use. What is infuriating about most modern software isn't that it makes simple tasks easy, it's that it makes complex tasks terribly difficult.
I think Dilger totally misses this point.
Note that in those days, the UNIX community was very much like today's free software community. UNIX was closed source in a strict sense, but there really wasn't much of a commercial market for it, so most users were in the research and academic community, and so had source access, and there was lots of sharing of code and fixes.
The equivalent of today's Linux vs. Windows battle was either UNIX vs. IBM or UNIX vs. VMS.
Around that time, there was also the unfortunate splintering battle within UNIX, AT&T vs. BSD, not to mention the battles within AT&T, between Research, USG, PWB, and Columbus. Sort of like Red Hat, Debian, SuSE, Slackware, Mandrake, et al. Such splits (among allies) have always been a bad thing, and I believe they shall remain so.
I was an enthusiastic poster of netnews in those days, when you could read all the netnews, and recognise the names of most everyone who posted.
trb (formerly floyd!trb, harpo!trb, trb@mit-mc, et al).
If you think that your time isn't worthless (free), then maybe you shouldn't be trusting worthless (free) reviews.
I'm wit that. I could never figure out the point of a GUI file manager. The text is the important part! The stupid little icons are useless noise. If you want to see an intelligently designed UI where you you can actually do powerful things with mouse and text, see Rob Pike's Acme UI from Plan 9. Other Plan 9 papers here.
prostitution/constitution
protestant/contestant
professor/confessor
what, exactly, was your point?
I was also disturbed by the term "General Public License." I'm surprised this thread is being moderated down, I find expurgation pretty offensive. Why not just go and edit an author's licensing terms out of his sources? Moderate me down, there's a difference between karma points and karma.
They delivered 335,000 names on 60,000 pages? That's fewer than six names per page. Seems pretty inefficient to me, that should have fit on a few floppies, certainly on a single zip disk.
If you want to move your desktop workspace from place to place, it's the file system that you want to carry, not the processor package. Such portable disk solutions already exist.
Try a different approach. Instead of fretting about whether or not the uninformed teacher is treating you with respect, find a software engiineer in your neighborhood who is willing to come to your school and lecture your class on software licences, open source, and everything. Teachers like when professionals from their community come to tell the kiddies about the real world. What the hell, maybe you can invite the local press too, and make it an event.
The press release, dated Thursday, 27-January, says: "It will take several days to complete the processing and the researchers do not expect to have confirmation of a signal until some time next week." This is not an open-ended quest, like the search for strange new worlds, this is a week-long data reduction task. I don't see why you would need a large coordinated effort like distributed.net.
You guys are missing the point. The feds don't want the cleartext contents of his disk, they want the key! It's probably something like DoJsux or the name of one of his old girlfriends, and he doesn't want to embarrass himself. Haven't you ever had a password like that?
See the WAP Forum What is WAP page.
It seems that most hackers are quite clueless about this stuff. I tried to post a fat response to slashdot but it killed my browser (oops). I fished the data out of the cache and stuck it in a web page. Here's some of the fruit of my experience with incentive stock options. I hope this helps.
If those aren't enough, here's the Scientific American Higgs link.
Here's the better article.
tupelos should have been: tuples
Symbolix should have been: Symbolics
address base should have been: address space
I would suggest proofreading before publishing. (That said, it was an interesting interview.)
At the suggestion of this article, I took a look at my config, and I found that almost everything that's optional was dynamically linked anyway.
I have a PII processor. Do I want to change my processor type? I don't know. What does it buy me to change from 386 to PPro? Is it worth the trouble? I assume that 99% of people are running the default 386. Might there be some subtle kink in the PPro kernel that isn't as well tested? Hmmm.
There's all these little features that I can turn off. How much memory am I saving with each one? It would be nice to know. Is it worth it to turn off a dozen features to save 64k on my 64Mb box? Nope. What kind of gains can a typical user expect from a reconfig?
Last lick. Of course, It's useful to have a kitchen sink kernel so that people can boot it on random hardware, but once you've booted, Linux knows what's there, and writes a /var/log/dmesg file with all that useful and mysterious info about your very own system. How about a prog that takes this custom data and builds a lean Linux kernel config from it?
It may be that this modem is "spoofing" some protocols, that is, looking at the data contents and doing something smart with it. This was originally done, I think, by Telebit 9600 baud modems, which did UUCP spoofing, and is also done by current fax modems, which may spoof certain handshaking. This typically requires pairs of modems that act as a black box, looking normal on the outside, and spoofing on the inside.
Alexander suggests that the designs you produce should be whole and harmonious and appropriate, and they should fit in their environment. So it seems ironic to me that advocates of such a wholistic ideal should tempt people with a reward that people find so irritating. This indicates that something is wrong with the design of the contest.
If gop.gov is allowed, then so should every political party, great and small, be allowed a foo.gov domain. The biggest parties are not cast in stone, neither their number nor their identification. There is nothing special about them, except that they are big and rich. They were not created with the nation, they were late comers. As such, why do they have power to demand special treatment (like a .gov) any more than libertarian, reform, communist, socialist, green, et al?
Plasma monitors can't be compared head to head with CRTs based simply on resolution, there are lots of other variables.
This is from the Journal of Irreproducible Results, (www.jir.com), Volume 36, Number 1. I think it was published about ten years ago, and is reproduced widely on the net, and as such, isn't terribly newsworthy.