If they're in a computer room, then such a scheme might work. But, if they're on user's desks, you don't really have control. They're subject to filling up, being shut off, being knocked about, crashing, etc. I don't think in this case you would really get the reliability that the diversity and independence would suggest.
I didn't see the first article when it came out, but just read it. It had far too many factual errors to be taken seriously. Examples: OS/360 was not the first general-purpose OS; UNIX was not underfunded by Bell Labs (Thompson was a full-time employee whose job was to do self-guided computing research).
The principal error in the new article is mixing up great software with a great product/service. Craig's List is a good example: The concept and execution were so great that all it needed was workable software, and that's probably all it actually had. It wasn't great software. Google, on the other hand, was and is great software.
But, the articles are certainly enjoyable. Too bad they aren't better researched.
Yes, I took the default... expanding disks. I didn't try any alternatives to see if performance would be different. However, my tests were run on a fresh system, so I doubt that fragmentation would have played much of a role. My test (running ImageIngester to copy several hundred ~5MB raw image files) involved mostly sequential I/O, so my guess is that Windows (or any OS) running on real hardware would employ various techniques (large disk blocks, read ahead, disk optimization etc.) that might not be working as effectively, if at all, on an emulated disk. Maybe there's a difference there as well between fixed-size and expanding.
The answer is "No" for the Mac version, but there's no way to know if the problem is with Parallels.
A virtualized Parallels machine is a vanilla PC, not an iMac, and OS X won't run on a vanilla PC. Since Mac Parallels is an OS X app, there's no way to run Mac Parallels inside of Parallels.
However, Parallels is available for other OSes, too (e.g., Linux), so someone out there may have a different answer for the non-Mac versions. Also, one could run a non-OS X Parallels inside the Mac OS X Parallels, and that might yield a different answer also.
Whatever is done, I'm sure the virtualized machine doesn't support hardware virtualization, so that feature would be gone. (Parallels still works without it.)
Hardware virtualization does indeed provide good CPU speeds (ignoring that there's only a single-core, which the review does cover), but the disk drive is not virtualized--it's emulated in an OS X file, and therefore Parallels can be significantly slower than Boot Camp (or any equivalent actual hardware). I came to somewhat different conclusions than did the review in a recent entry on my own blog (http://basepath.com/index-real.php?url=blogentry/ 2006-06-28.htm).
Some other points: I found the mouse a bit jerky, and no machine in Parallels can have close to the amount of physical RAM (e.g., on my 1.5GB iMac, I could have at most 1GB for a Parallels machine).
Nonetheless, Parallels works really, really well, and it's trememdously convenient. But the article goes overboard in its assessment of its performance.
1. The Report on the language used a formal syntax specification, one of the first, if not the first, to do so. Semantics were specfied with prose, however. 2. There was a distinction between the publication language and the implementation language (those probably aren't the right terms). Among other things, it got around differences such as whether to use decimal points or commas in numeric constants. 3. Designed by a committee, rather than a private company or government agency. 4. Archetype of the so-called "Algol-like languages," examples of which are (were?) Pascal, PL./I, Algol68, Ada, C, and Java. (The term Algol-like languages is hardly used any more, since we have few examples of contemporary non-Algol-like languages.)
However, as someone who actually programmed in it (on a Univac 1108 in 1972 or 1973), I can say that Algol60 was extremely difficult to use for anything real, since it lacked string processing, data structures, adequate control flow constructs, and separate compilation. (Or so I recall... it's been a while since I've read the Report.)
First, I should say that it sort of is--I have 3 desktops up and running all the time on my large desk, and Ubuntu Linux is one of them. The other two are Windows and Mac OS X, with the Mac in the primary position, where it handles 99.9% of my desktop work. (By the way, I'm a programmer, and have even written a book on UNIX/Linux programming.)
Here are ten reasons why the Ubuntu machine isn't on center stage:
1. Clients, despite my telling them to send PDFs, keep sending Word and Excel files. Yes, I do open them with OpenOffice, even on the Mac, but I very often find that the conversion is imperfect. The main stuff is OK, but things like balloon-style comments don't appear. If it's an important document to review, it's unsafe, and unprofessional, to attempt to review it if I'm not sure I'm seeing all the text, arranged as it should be. Sometimes the Word or Excel file is on a web site (e.g., Texas Education Agency), and I can't "tell" them to send me a PDF.
2. No Photoshop. Yes, there are image-editing programs for Linux, but I need precise color management using ICC profiles, support for raw files, including Adobe DNG, Adobe's Camera Raw, and the ability to convert raw files to DNG using Adobe DNG Converter.
3. No monitor-calibration device (I use Spyder2PRO) for Linux. I can't do my work on an un-calibrated monitor.
4. Having used the Mac, I just couldn't bear to use anything less usable, powerful, agile, and attractive. It would be like trading in a BMW for a Chevy. (If I weren't a Mac user, I could easily trade the Windows GUI for Gnome.)
5. No Quicken. Sure, I could switch, but I have a check register that goes back 20 years, and it's connected to my credit card and bank accounts. Maybe I could reproduce this on Linux, but I don't have the week it would take to do the research and set it all up.
6. No iTunes. Not for buying songs--I've never bought one. But I get a download from Audible of NPR's Fresh Air every day, and it automatically shows up on my iPod. So do the CDs I rip. Yes, I've used other software, but its irritating to do all that clicking. I like it that the stuff just wends its way to my iPod without my intervention.
7. I've started to write Mac Cocoa programs in Objective-C with Interface Builder. I've used many other languages and IDEs, but this one is far-and-away the most polished and gives me the most intellectual and esthetic pleasure
8. Integration. When I first bought my Mac, I went to a web form and the Mac automatically filled it out. Strange, because I had never entered my address and phone number! Turns out that I had synched my Palm, my info made it's way into the Address Book, and Safari picked it up from there. This is only one example. All the Apple apps (and a few non-Apple apps) work together, often in amazing and unexpectedly brilliant ways.
9. AppleScript. I hate the language, but I love that most of the important Mac applications are scriptable. I like that every Mac comes with Java, too. Not yet true of Linux (wrong license type, or something like that).
10, PDFs are the basis for all document display, and the heart of the printing subsystem. Built-in, deep into the system. Having them just on the surface, as is true of Linux and Windows, doesn't work nearly as well.
OK, I'll stop at ten. The point is, why use a system that's less than the best available? And, such as the case with Java, I know that I can install the stuff myself, but why bother? I have other things to do with my time! I want a desktop system that's complete and ready to go out of the box.
It's one of those irritating web pages that's for Windows only, yet has such a Windows-centric view of the world that it doesn't even say that it's Windows only! (Not on the Home or Download pages, anyway--didn't read anything else.) The hint is when your Mac wants to know what to do with the EXE file you've just downloaded.
May be the greatest thing ever, but it definitely started off on my wrong side...
Blocks of addresses are given to organizations, so any comparison of the number of theoretical addresses to the number of addresses actually in use is misleading, because the distribution is nonuniform. In addition, while it's true that we could all work together to conserve addreses, under the present system they are assigned as they are requested, and therefore the assignining is going to find the well dry at some point. (We could all work together to conserve water, oil, or flu shots, too, but we don't seem to do that nearly as much as we ought to, right?)
By the way, for my book Advanced UNIX Programming, I calculated the number of IPv6 addresses as greater than the number of particles in the universe.
Well said! My software company's training, support, and consulting business was very profitable, but product sales were a huge loser. The sales reps used to FLY to customer sites just to give demos, often with no ensuing sale. I wish I'd been smart enough to fire the sales guys, pump up marketing, and give the product away.
Open source is an even smarter idea, especially for RedHat. You can get programmmers to work on your product without paying them. The GPL takes in one big step further: You can even get your hands on what your competitors develop.
RedHat is a public company, so their numbers are reported. I don't know about the two private companies.
This is yet another case of someone claiming a company is profitable by looking at their revenues. Meaningless.
Also, MySQL (I know less about RedHat and JBoss) has dual licensing, and I'm sure their product revenues come entirely from the non-GPL side of the business. Their services and training revenues may come from both. Where their profits come from, if they have any profits, is unclear.
To my way of thinking, a dual-license company uses the free license to gain market share, so as to reduce their customer acquisition costs. The business is really about the non-free ("as in beer") side. The risky thing, for which OSS companies should be admired, is to make even the non-free side open source, rather than closed, trusting that customers are honest. (Many companies have a watered-down free product along with their non-free product, but neither are open source. To go all open source is exceptional.)
I started a company once that reached #72 on the Inc 500, based on revenue growth over 5 years, not one of which was profitable. In fact, for a venture-capital-backed company, which we were, as MySQL is, running the business conservatively enough to make a profit may not be the best course. Also, you can't count investment money as revenue when it comes in, but it is an expense when it goes out, so this also reduces profits, and usually results in big losses.
What he actually said about "theatre" was this: "I think people like the idea of this 5,400 employee software company buying up a German Linux distributor. I think they liked the theatre of it." The paraphrasing of this in the leader is misleading.
That aside, of course, Red Hat would hope that the number of non-Red Hat distros would stay high, since that tends to increase the gap between Red Hat, the only Linux distro that most ITers know about, and the rest of the pack. In addition, the confusion, or perception of confusion, drives corporate Linux users to Red Hat.
Disparaging Novel/SuSE is also to be expected from Red Hat, since SuSE was and is the only competitor to Red Hat.
(My own opinion is that the proliferation of distros is a serious problem that wastes effort by Linux distro developers, complicates support, makes life difficult for application developers, and gives many potential users, both corporate and consumer, the impression that Linux is immature. If whatever-we-mean-by-Linux were a complete system, like FreeBSD, we wouldn't need packaging to be a separate operation, and the number of distro outfits would be very small.)
Apple and Google are indeed innovative. But, can you name any innovations in Linux? I'm not doubting its usefulness or its impact, which are both substantial, only that either is in any way due to innovation.
1. Most titles that I want to read aren't available.
2. When they are available, the e-book version costs substantially more. (Barnes & Noble and Amazon frequently sell best sellers in paper at huge discounts, which rarely apply to the e-book version.)
But, when I'm enjoying Al Franken, I sometimes wonder whether he's our "Ann Coulter." Of course, he sure seems to be telling the truth, not lies like she tells, but perhaps my bias is affecting my critical powers?
Lots of people care about what she says, obviously, judging by the reaction here and in other forums to what she says. Whether the quotes are "ridiculous" is immaterial, as long as they are accurate.
Inasmuch as many people who hear about DiDio might want to know who she is and why she upsets so many people, it's useful for the Wikipedia to provide the information. That you, and I take it, many others disagree with her is not a reason for her article to be removed or, for that matter, for her views to not be heard.
Perhaps we simply have a basic disagreement about public discourse. I want to hear views with which I disagree, and I assume that others can apply their own critical thinking. If they can't, then I question whether they should even be exposed to views with which they do agree.
Very well said, and you might have added that one reason why it's important to have articles on relatively obscure subjects is that another article may mention the subject, and then the reader can just click on a link to find out more. Much like the way a footnote works.
For example, an article on Linda Ronstadt mentioned the Stone Poneys, for which there was no article. There is now, though!;-)
I looked at the DiDio article, and it's very well done. It does indeed present her POV, but by quoting her (accurately, I assume). This is quite different from the article itself taking her POV.
I was at Bell Labs from 1970 to 1982, and I don't remember any Dept. 1127. My 1980 Bell Labs Directory shows a Dept. 1271, led by McIlroy, consisting of Cherry, Morris, Thompson, Aho, Baker, Lengaauer, Syzmanski, Weinberger, and Yannakakis. Its sibling, Dept. 1273, led by Fraser, consisted of Chesson, Kernighan, Ritchie, Stroustrup, Vollaro, Johnson, Ditzel, Elliott, and Feldman. (No Pike--I don't think he was at Bell Labs yet.)
I guess everyone thinks that Thompson and Ritchie were in the same department during the 1970s, but I do remember always knowing that they were not.
Note that by 1980 UNIX-related OS research at Bell Labs was nearly completed. Development of UNIX, which is where I worked, was very active and remained so for another 10+ years, but that's different from research. (Center 127 did research in many areas unrelated to UNIX.)
So, undoubtedly there was a recent reorg and some department went away, and maybe it was even 1127, but what that means, if anything (since Thompson, Kernighan, and others left a while ago), I have no idea.
Anyway, I think the gist of the article and most of the responses here is that it's sad that AT&T and Lucent are no longer combined and able to spend as lavishly on research as they once did. That part of this thread is true.
A few posts are from Bell Labs people who said it was a great place to work, and that's true, too.
Linux and some of the BSDs are "expected to 'run everywhere'" because their creators and supporters make that claim. OS X is not because its sole creator, Apple, makes no such claim.
It is true that UNIX-based systems have a theoretical ability to "handle hardware that all other unixen do," but for this theoretical ability to become real requires:
1. The necessaary device drivers,
2. Necessary kernel modifications outside of device drivers,
3. Lots of hardware testing,
4. Keeping up with hardware changes,
5. Support for all the combinations.
For any OS, the assumption that 1 - 5 can be done for a narrow subset of hardware will lead to many opportunities to produce a more reliable system.
Actually, with the BSD license you do get something back: Credit for the original work. This coupled with the greater likelihood that the software will be used in heavily-marketed proprietary products is sometimes exactly what authors want. For them, the BSD license is a good match.
For example, I released the source code for my book, Advanced UNIX Programming, under the BSD license, and what I wanted in return (other than more happy people in the world) was publicity for the book. I still get improvements and bug fixes emailed to me by helpful readers.
It's not true that "AJAX apps work in any browser out there." Perhaps the writer meant to say "all the major browsers have versions that support AJAX (XMLHttpRequest)?"
Most of the web references to AJAX that I've seen correctly point out the importance of checking the browser version, the necessity of testing on many different browsers and versions, and the difficulty of fallback coding if XMLHttpRequest isn't supported. For example, see the AJAX page at http://en.wikipedia.org/wiki/AJAX.
This is not to say that AJAX isn't terrific, and a major breakthrough in building more responsive web apps. But it's often important for the application to work for all users, many of whom are still running Windows 98 (with an old IE), Mac OS 8 (or earlier), Palm OS, and other systems that don't support AJAX. Sometimes these users can be ignored, and sometimes they can't. If they can't, the developer who wants to use AJAX may have to program (and document, test, maintain, etc.) a second, non-AJAX interface.
Well, many people would call me a "guru programmer," and there's no way I would do this for anything like $80 x 8 hours.
You need error-checking on printf and an error log for the error message, you need to internationalize and localize the string, you need both user and maintenance documentation, and you need a client-server architecture. Not to mention that you seem to have jumped into the project without written requirements, a specification, and usability testing.
This would take about a month at my standard rate of $200/hr.
Slashdot Mirror
If they're in a computer room, then such a scheme might work. But, if they're on user's desks, you don't really have control. They're subject to filling up, being shut off, being knocked about, crashing, etc. I don't think in this case you would really get the reliability that the diversity and independence would suggest.
--Marc
I didn't see the first article when it came out, but just read it. It had far too many factual errors to be taken seriously. Examples: OS/360 was not the first general-purpose OS; UNIX was not underfunded by Bell Labs (Thompson was a full-time employee whose job was to do self-guided computing research).
The principal error in the new article is mixing up great software with a great product/service. Craig's List is a good example: The concept and execution were so great that all it needed was workable software, and that's probably all it actually had. It wasn't great software. Google, on the other hand, was and is great software.
But, the articles are certainly enjoyable. Too bad they aren't better researched.
--Marc Rochkind
ImageIngester.com
Yes, I took the default... expanding disks. I didn't try any alternatives to see if performance would be different. However, my tests were run on a fresh system, so I doubt that fragmentation would have played much of a role. My test (running ImageIngester to copy several hundred ~5MB raw image files) involved mostly sequential I/O, so my guess is that Windows (or any OS) running on real hardware would employ various techniques (large disk blocks, read ahead, disk optimization etc.) that might not be working as effectively, if at all, on an emulated disk. Maybe there's a difference there as well between fixed-size and expanding.
--Marc
The answer is "No" for the Mac version, but there's no way to know if the problem is with Parallels.
A virtualized Parallels machine is a vanilla PC, not an iMac, and OS X won't run on a vanilla PC. Since Mac Parallels is an OS X app, there's no way to run Mac Parallels inside of Parallels.
However, Parallels is available for other OSes, too (e.g., Linux), so someone out there may have a different answer for the non-Mac versions. Also, one could run a non-OS X Parallels inside the Mac OS X Parallels, and that might yield a different answer also.
Whatever is done, I'm sure the virtualized machine doesn't support hardware virtualization, so that feature would be gone. (Parallels still works without it.)
--Marc
Some other points: I found the mouse a bit jerky, and no machine in Parallels can have close to the amount of physical RAM (e.g., on my 1.5GB iMac, I could have at most 1GB for a Parallels machine).
Nonetheless, Parallels works really, really well, and it's trememdously convenient. But the article goes overboard in its assessment of its performance.
--Marc
1. The Report on the language used a formal syntax specification, one of the first, if not the first, to do so. Semantics were specfied with prose, however.
2. There was a distinction between the publication language and the implementation language (those probably aren't the right terms). Among other things, it got around differences such as whether to use decimal points or commas in numeric constants.
3. Designed by a committee, rather than a private company or government agency.
4. Archetype of the so-called "Algol-like languages," examples of which are (were?) Pascal, PL./I, Algol68, Ada, C, and Java. (The term Algol-like languages is hardly used any more, since we have few examples of contemporary non-Algol-like languages.)
However, as someone who actually programmed in it (on a Univac 1108 in 1972 or 1973), I can say that Algol60 was extremely difficult to use for anything real, since it lacked string processing, data structures, adequate control flow constructs, and separate compilation. (Or so I recall... it's been a while since I've read the Report.)
First, I should say that it sort of is--I have 3 desktops up and running all the time on my large desk, and Ubuntu Linux is one of them. The other two are Windows and Mac OS X, with the Mac in the primary position, where it handles 99.9% of my desktop work. (By the way, I'm a programmer, and have even written a book on UNIX/Linux programming.)
Here are ten reasons why the Ubuntu machine isn't on center stage:
1. Clients, despite my telling them to send PDFs, keep sending Word and Excel files. Yes, I do open them with OpenOffice, even on the Mac, but I very often find that the conversion is imperfect. The main stuff is OK, but things like balloon-style comments don't appear. If it's an important document to review, it's unsafe, and unprofessional, to attempt to review it if I'm not sure I'm seeing all the text, arranged as it should be. Sometimes the Word or Excel file is on a web site (e.g., Texas Education Agency), and I can't "tell" them to send me a PDF.
2. No Photoshop. Yes, there are image-editing programs for Linux, but I need precise color management using ICC profiles, support for raw files, including Adobe DNG, Adobe's Camera Raw, and the ability to convert raw files to DNG using Adobe DNG Converter.
3. No monitor-calibration device (I use Spyder2PRO) for Linux. I can't do my work on an un-calibrated monitor.
4. Having used the Mac, I just couldn't bear to use anything less usable, powerful, agile, and attractive. It would be like trading in a BMW for a Chevy. (If I weren't a Mac user, I could easily trade the Windows GUI for Gnome.)
5. No Quicken. Sure, I could switch, but I have a check register that goes back 20 years, and it's connected to my credit card and bank accounts. Maybe I could reproduce this on Linux, but I don't have the week it would take to do the research and set it all up.
6. No iTunes. Not for buying songs--I've never bought one. But I get a download from Audible of NPR's Fresh Air every day, and it automatically shows up on my iPod. So do the CDs I rip. Yes, I've used other software, but its irritating to do all that clicking. I like it that the stuff just wends its way to my iPod without my intervention.
7. I've started to write Mac Cocoa programs in Objective-C with Interface Builder. I've used many other languages and IDEs, but this one is far-and-away the most polished and gives me the most intellectual and esthetic pleasure
8. Integration. When I first bought my Mac, I went to a web form and the Mac automatically filled it out. Strange, because I had never entered my address and phone number! Turns out that I had synched my Palm, my info made it's way into the Address Book, and Safari picked it up from there. This is only one example. All the Apple apps (and a few non-Apple apps) work together, often in amazing and unexpectedly brilliant ways.
9. AppleScript. I hate the language, but I love that most of the important Mac applications are scriptable. I like that every Mac comes with Java, too. Not yet true of Linux (wrong license type, or something like that).
10, PDFs are the basis for all document display, and the heart of the printing subsystem. Built-in, deep into the system. Having them just on the surface, as is true of Linux and Windows, doesn't work nearly as well.
OK, I'll stop at ten. The point is, why use a system that's less than the best available? And, such as the case with Java, I know that I can install the stuff myself, but why bother? I have other things to do with my time! I want a desktop system that's complete and ready to go out of the box.
--Marc Rochkind
It's one of those irritating web pages that's for Windows only, yet has such a Windows-centric view of the world that it doesn't even say that it's Windows only! (Not on the Home or Download pages, anyway--didn't read anything else.) The hint is when your Mac wants to know what to do with the EXE file you've just downloaded.
May be the greatest thing ever, but it definitely started off on my wrong side...
Blocks of addresses are given to organizations, so any comparison of the number of theoretical addresses to the number of addresses actually in use is misleading, because the distribution is nonuniform. In addition, while it's true that we could all work together to conserve addreses, under the present system they are assigned as they are requested, and therefore the assignining is going to find the well dry at some point. (We could all work together to conserve water, oil, or flu shots, too, but we don't seem to do that nearly as much as we ought to, right?)
By the way, for my book Advanced UNIX Programming, I calculated the number of IPv6 addresses as greater than the number of particles in the universe.
Open source is an even smarter idea, especially for RedHat. You can get programmmers to work on your product without paying them. The GPL takes in one big step further: You can even get your hands on what your competitors develop.
I am so dumb...
This is yet another case of someone claiming a company is profitable by looking at their revenues. Meaningless.
Also, MySQL (I know less about RedHat and JBoss) has dual licensing, and I'm sure their product revenues come entirely from the non-GPL side of the business. Their services and training revenues may come from both. Where their profits come from, if they have any profits, is unclear.
To my way of thinking, a dual-license company uses the free license to gain market share, so as to reduce their customer acquisition costs. The business is really about the non-free ("as in beer") side. The risky thing, for which OSS companies should be admired, is to make even the non-free side open source, rather than closed, trusting that customers are honest. (Many companies have a watered-down free product along with their non-free product, but neither are open source. To go all open source is exceptional.)
I started a company once that reached #72 on the Inc 500, based on revenue growth over 5 years, not one of which was profitable. In fact, for a venture-capital-backed company, which we were, as MySQL is, running the business conservatively enough to make a profit may not be the best course. Also, you can't count investment money as revenue when it comes in, but it is an expense when it goes out, so this also reduces profits, and usually results in big losses.
That aside, of course, Red Hat would hope that the number of non-Red Hat distros would stay high, since that tends to increase the gap between Red Hat, the only Linux distro that most ITers know about, and the rest of the pack. In addition, the confusion, or perception of confusion, drives corporate Linux users to Red Hat.
Disparaging Novel/SuSE is also to be expected from Red Hat, since SuSE was and is the only competitor to Red Hat.
(My own opinion is that the proliferation of distros is a serious problem that wastes effort by Linux distro developers, complicates support, makes life difficult for application developers, and gives many potential users, both corporate and consumer, the impression that Linux is immature. If whatever-we-mean-by-Linux were a complete system, like FreeBSD, we wouldn't need packaging to be a separate operation, and the number of distro outfits would be very small.)
Apple and Google are indeed innovative. But, can you name any innovations in Linux? I'm not doubting its usefulness or its impact, which are both substantial, only that either is in any way due to innovation.
1. Most titles that I want to read aren't available. 2. When they are available, the e-book version costs substantially more. (Barnes & Noble and Amazon frequently sell best sellers in paper at huge discounts, which rarely apply to the e-book version.)
Anyway, thanks for the info.
But, when I'm enjoying Al Franken, I sometimes wonder whether he's our "Ann Coulter." Of course, he sure seems to be telling the truth, not lies like she tells, but perhaps my bias is affecting my critical powers?
Naaaaaaaahhhh.....
Inasmuch as many people who hear about DiDio might want to know who she is and why she upsets so many people, it's useful for the Wikipedia to provide the information. That you, and I take it, many others disagree with her is not a reason for her article to be removed or, for that matter, for her views to not be heard.
Perhaps we simply have a basic disagreement about public discourse. I want to hear views with which I disagree, and I assume that others can apply their own critical thinking. If they can't, then I question whether they should even be exposed to views with which they do agree.
For example, an article on Linda Ronstadt mentioned the Stone Poneys, for which there was no article. There is now, though! ;-)
I looked at the DiDio article, and it's very well done. It does indeed present her POV, but by quoting her (accurately, I assume). This is quite different from the article itself taking her POV.
That makes sense... can anyone confirm?
I guess everyone thinks that Thompson and Ritchie were in the same department during the 1970s, but I do remember always knowing that they were not.
Note that by 1980 UNIX-related OS research at Bell Labs was nearly completed. Development of UNIX, which is where I worked, was very active and remained so for another 10+ years, but that's different from research. (Center 127 did research in many areas unrelated to UNIX.)
So, undoubtedly there was a recent reorg and some department went away, and maybe it was even 1127, but what that means, if anything (since Thompson, Kernighan, and others left a while ago), I have no idea.
Anyway, I think the gist of the article and most of the responses here is that it's sad that AT&T and Lucent are no longer combined and able to spend as lavishly on research as they once did. That part of this thread is true.
A few posts are from Bell Labs people who said it was a great place to work, and that's true, too.
It is true that UNIX-based systems have a theoretical ability to "handle hardware that all other unixen do," but for this theoretical ability to become real requires: 1. The necessaary device drivers, 2. Necessary kernel modifications outside of device drivers, 3. Lots of hardware testing, 4. Keeping up with hardware changes, 5. Support for all the combinations.
For any OS, the assumption that 1 - 5 can be done for a narrow subset of hardware will lead to many opportunities to produce a more reliable system.
For example, I released the source code for my book, Advanced UNIX Programming, under the BSD license, and what I wanted in return (other than more happy people in the world) was publicity for the book. I still get improvements and bug fixes emailed to me by helpful readers.
You're right, of course. I was being imprecise.
Most of the web references to AJAX that I've seen correctly point out the importance of checking the browser version, the necessity of testing on many different browsers and versions, and the difficulty of fallback coding if XMLHttpRequest isn't supported. For example, see the AJAX page at http://en.wikipedia.org/wiki/AJAX.
This is not to say that AJAX isn't terrific, and a major breakthrough in building more responsive web apps. But it's often important for the application to work for all users, many of whom are still running Windows 98 (with an old IE), Mac OS 8 (or earlier), Palm OS, and other systems that don't support AJAX. Sometimes these users can be ignored, and sometimes they can't. If they can't, the developer who wants to use AJAX may have to program (and document, test, maintain, etc.) a second, non-AJAX interface.
You need error-checking on printf and an error log for the error message, you need to internationalize and localize the string, you need both user and maintenance documentation, and you need a client-server architecture. Not to mention that you seem to have jumped into the project without written requirements, a specification, and usability testing.
This would take about a month at my standard rate of $200/hr.