Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!


Forgot your password?
The Courts Government News

Microsoft Video Blunder 130

An anonymous reader wrote in with this gem: "Microsoft was trying to show that a PC loaded with Windows 3.1 took longer to access the Internet than a PC loaded with Windows 98 using Internet Explorer. It turns out that the Windows 98 PC was connected to a faster modem than the Windows 3.1 PC. " I'll refrain from (further) comment on this one. Plenty of meat here. Update: 02/10 09:25 by J : Oops. More like "from the egg on my face dept." - it seems I made an honest yet huge mistake: the article says "configuring". Sorry I couldn't update when I realized this. There were some DB troubles, which seem to now have been fixed. Sincerest apologies to all.
This discussion has been archived. No new comments can be posted.

Microsoft Video Blunder

Comments Filter:
  • Did I miss something in the trial...why do we care that Windows98 is faster than Win3.1? Shouldn't they have tested Win98 and win95?
  • And screw all you whiners who can't handle me, or others, "bashing" them. They've been incredibly stupid these past few weeks (moreso than usual, even), and deserve all the derision that's being heaped onto them.

    - A.P.

    "One World, One Web, One Program" - Microsoft Promotional Ad

  • OR hit "Flat Mode", that usually works for me.
  • It seems to me that Microsoft is trying to decieve the DOJ in court (under oath?). Certainly these are grounds for a contempt of court charge, yet all I hear is:
    "One bright eyed DOJ employee noticed the Microsoft Word and Excel icons appeared on the taskbar for the second half of the demo. The DOJ asked Microsoft to remake the video." (Presumeably so they don't forget to hide the taskbar this time).

    Plus this I can connect faster to the internet with 98...and a faster modem stuff.

    Clinton could have been impeached for stuff like this (well, maybe, but we certainly would have heard about it about a million times).
  • It depends. If you're dialing into a 33.6k modem with a 33.6k modem, both will try to sync up at 33.6 first. If you have a 28.8 modem dialing into a 33.6, the 33.6 will try to sync at 33.6, then 31, then keep going down the line until it hits 28.8, at which point they sync up.
  • Posted by Alonzo The Great:

    'doze 9.11 uh windoze 3.11 Yeah, I know I'm lame. If you got a better OS and want to tell me how great it is e-mail Alan Grimes at the link above... :)
  • ... and I'll say it again.

    Microsoft Internet Explorer is an Application.

    A TCP/IP Stack is not...

    Microsoft has the source code to Windows.

    Microsoft can REMOVE Internet Explorer by rewriting Windows.

    End of story.

    Someone needs to inform the DOJ of these simple facts.

  • You could easily reverse the results by using a 9600bps modem for Windows 98 for which only Win3.1 drivers exist, and a 14.4 with drivers for both for the Windows 3.1 box.

    The problem is not the slight difference in speed. It's that the modems are different, so the result is not truly objective.
  • Yes, the first version of Windows 95 did have PPP dialup features. (It was the first Windows release to ship with it - unless NT did. Shrug.)
  • Actually people want to play games on machines like the PlayStation rather than Windows. Otherwise why are there so much interest in emualating these machines on the PC. BTW I haven't bought a game for the PC since I bought my Playstation a couple of years ago. The system requirements for PC games these days are becoming just too absurd, and the games on the PlayStation are far more interesting, especially if you are an Anime/Manga fan.
  • Someone please show Mr. Barkto the door. . .
  • A similar test could be performed in the opposite direction--pit a machine with decent buffered UARTs and Windows 3.1 with an external 56k modem against a Win98 box with a WinModem. Note how slowly the machine runs while "downloading the Internet" or whatever these morons call it. Make it illegal to sell WinModems.
  • They COULDN'T be that silly.. These tests must be conducted by utter morons.. Shouldn't these tests be done by neutral, third party vendors? I mean, who's the say they didn't just speed up the friggen tape to make it appear faster?

    "Joe, move the mouse REAL slowly.. We need to speed up this tape, and don't want it to look phony.."
  • Yes, there is a big difference when one is judging speed.. As a matter of fact, unless they both had dedicated access to the SAME T1 line, the results can be SIMPLY screwed by someone simply ftping a large file across the LAN..
  • Lemme get this strait.. They mislead the court on 3 different occasions:

    1) Using survey results that where 'ordered' to turn out the way they did..

    2) Using systems that are NOT CONFIGURED as they stated to demo something..

    3) They demo something on different speed modems, and suprisingly, the slower modem lost..

    Does this mean that I can 'forget' to mention 30,000 of income when doing my taxes and use the 'gimme a break' as a reason?
  • They weren't comparing thouroughput, they were comparing the time it took to configure the system. Which doesn't have *anything* to do with modem speed. (well, I guess configuring a 1980s 1200 baud modem might be difficult these days, but...)

  • by diakka ( 2281 )
    you know, that 28.8 "purely academic" difference could be alot difference if the modem was the old style VFC protocol, the pre-V.34 standard protocol. I had one of these modems, and it was completely unreliable. It never connected at 28.8. On the average it connected at 19.2 and it was not rare to get 14.4. Not to mention, that when these things got errors, the modem would lock up for a few seconds, until it could renegotiate the connection with the other end. Basically this "academic" difference, could have resulted in the difference shown by the video. In fact, even if there was nothing fishy about the modems themselves, they could have used telephone lines that differed significantly in quality. While the modems may have been nearly the same, the actual connection speeds could be far different, thus accounting for the big difference.
  • Hey genius... In case you didn't notice, this article's topic is on the MS vs DOJ case, and the video screw up. Not Linux, not FreeBSD, not BeOS. Microsoft. Natrually people are gonna express their feelings on MS. It's not as if they're offering unsolicited or off-topic comments.
    Leave MS alone, if you hate them so much, ignore them.
    If you hate the bashing comments on MS stories on slashdot, ignore them.
  • THANK you... I am not normally one to defend Microsoft (at best I keep my mouth shut), however in this case, while you may dispute the point of comparing Win98 to Win3.1 in terms of Internet access, they were comparing the time it takes to CONNECT to the Internet, including, presumably, the configuration of TCP stacks on each. This may actually (!!) be a valid point as difficulties in this process are what plague many users (do you remember the fucked up TCP stack in 3.1?), and making the setup faster and easier is certainly a plus.

    What I dont get is the comparison: why compare technically distinct OSes when they had a perfect comparison between Win95 and Win98 (essentially the same product except for the features that are currently in question). Thats akin to an engine maker claiming its new model is faster and better in every way than its previous model, but then opting to compare it in lab tests to its 1927 model that doesnt even share a similar design.

    Also, on a side note, even IF the Win98 machine was using a faster modem, and all other things were equal, it sure seems to me that with each successive generation of modem standards the handshakes take LONGER, therefore if Billy-boy had a 2400 in one box and a 56k in another, my bet is on the 2400 connecting faster.

    With that said, I still hope someday thge Justice Dept. will pull their heads out of their collective bureaucratic ass and develop an actual strategy in this case... otherwise issues like this will continue to plague the trial until its dismissal.

    Please dont post such obviously slanted articles (or respond to them) without thinking first.. being rabid Microsoft-haters will discredit us all in the long run... why nitpick when we have MORE than enough ammo already?
  • And can someone explain why MS would want to compare Win3.1 to Win98 rather than Win95 to Win98? Someone at the DOJ should ask Microsoft on stand why this comparison is so relevant.

    Yep. I was thinking the same. Why are they comparing Win98 against Win3.1 instead of Win95? It does not make any sense, unless Win98 and Win95 perform the same with and without IE. And that is obviously the reason they used Win3.1.

  • I recall from the article that "most of the time was spent" loading the browser application from the disc. What a stupid comparison!

  • Slashdot itself has done nothing to make itself look bad here. Slashdot is a news site reporting the news. Moreover, it's not bashing MS's product, it's insulting their perceptions of our intelligence. If some comment-posters feel obliged to make their usual childish remarks, it's their own problem. Slashdot has done nothing here to stain its journalistic integrity (unless the summary is incorrect).
  • I think there are always grounds for appeal, but mistakes made in your own defense are not one of them. Also, is not an appeal actually a review of the trial proceeding, its fairness and the legal logic used for the various rulings?

    Perhaps a lawyer can comment on what recourse defendants have to deal with their own inept presentation of their case.
  • explain to me how microsoft is allowed to do these tests. Think about this, if you were just arrested for murder, would they allow you to go to the crime scene to collect the evidence? i think not! Why are they allowing microsoft to create and perform the tests? is the DOJ so computer retarded that they can't hire a impartial third party to perform the test? much less do the tests themselves? For some reason i have a feeling that no matter what MS do, they will win this case. The DOJ has done nothing to prove their case, while MS is doing everything (and screwing it up big time) to win.
  • Yet another "Anonymous Coward" wrote:

    According to the article, one modem was a 33.6 and the other was a 28.8.

    Wow. Big difference there. Are Microsoft haters now reduced to this level of nitpicky-ness?

    Let's see. 33.6 is about 17% faster than 28.8. How about we take 17% of your income as additional tax? You wouldn't be nit-picky about it or anything would you?

    17% is a big difference.

  • ohmygod... the wacky graph from the (every month repeating itself) microsoft ad in wired with webserver "benchmarks" comes into mind. crafty folks they are,- in _that_ business, i mean.
  • Modem Configurations for Internet Speed Comparison

    • Modem Type
      • Windows-98: External
      • Windows 3.1: Internal WinModem
    • Speed
      • Windows-98: 56K V.90
      • Windows 3.1: 2400 bps
    • Error Correction
      • Windows-98: On
      • Windows 3.1: Off
    • Connection Type
      • Windows-98: RJ-11-style jack
      • Windows 3.1: Acoustic coupler

    Yeah, the differences are academic.

    :-) :-) :-),

  • The win3.1 machine was probably at 1200 baud and the 98 machine with a pocket rocket DSL modem... Way to fool em Bill!!
  • One of Microsoft's suits said that the difference in modem speeds is a red herring. Well, this whole demo was one big red herring. MS proved what, exactly? That it is faster to set up a "trial account" with an ISP or online provider if you already have a browser installed. Big whoop.

    There was no need for Microsoft to "integrate" MSIE into 32-bit Windows to provide this benefit. All they'd have to do for the same results is to preinstall a web browser. Or better yet, encourage OEMs to preinstall a choice of web browsers.

    I'd like to see a fair test instead of this sham. Same OS (or at least both Win9x), same processor, same browser preinstalled, same modem on both machines. Then set up a dialup account. I'm sure the time required would not be vastly different for the two computers then.

    I still have Windows 3.11, and I can tell you that setting up Netscape or (ugh) AOL is dead easy. MSIE is a pain in the neck to configure, though, because they think they know better than you, and try to hide everything.

  • Why must everyone post this tired, tired, old quote whenever a reference to Windows 95 comes up? It was mildly entertaining the first time. Barely so the next few times. It has no entertainment value now. Let it die.
  • There seems to be some misconception about why the DoJ would raised the issue of different modem speeds. If one has been following Boiles cross-examination of the MS witnesses, one realizes that the DoJ is systematically trying to destroy their credibility. Furthermore, the DoJ is also trying to show that MS has exhibited a pattern of evasiveness and deceit in their testimonies (written and oral). The flawed tapes and the hidden modem differences all move towards this goal. Remember, this is all cummulative in it's effect.

    In contrast, MS's cross-examination of DoJ witnesses seemed to be primarily concentrated on disputing (or correcting) individual little points of contention. Furthermore, while the DoJ witnesses were from a wide-range of organizations, MS's witnesses are almost exclusively MS employees. If one destroys the credibility of the MS witnesses, you destroy the credibility of MS. This will then make life tough for MS if this gets to an Appeals Court.

    The DoJ can still call two more witnesses. If too many bogus technical claims are made, I would suggest that the DoJ do their own technical demonstrations (video presentation of course). And yes, sorry MS, you are not allowed to be there when the DoJ sets up or films their demo.
  • I read somewhere that the only truly intuitive interface was the nipple. Everything else had to be learned. :-)
  • I have to agree with this thread, why can a company like this be so blatently misleading and then just say 'Oops...I didn't really mean to say that.' If any one of us tried this, we would be forking over a bunch of money for a contempt charge, or spending an evening in the care of the state.
  • On a related note, visit today's stories at http://www.mercurycenter.com/business/microsoft/tr ial/ where it was reported that M$ VP Myhrvold admitted under corss-examination that the reasons behind Microsoft's restrictive agreements with ISPs to push IE was the fear that Netscape would win in side-by-side comparisons. This is in *direct* contradiction of his written testimony, where he states that IE's share increase was due to "superior product".

    I hesitate to say perjury... nah, no I don't.
  • Yep. You would be jailed. But that is because you are not a multi-million dollar monopoly with tons of lawyers at your disposal.
  • by Dast ( 10275 )

    This is a linux page, don't point out MS flaws

    What? Where does it say "Slashdot: Linux news for nerds?"

    We are not pointing out the "flaws" of MS, but their lies. If you don't like it, ignore it. If you can't stand it, leave.

  • Ha, the laugh of the day from our dear friend Bill. :) It's kind of sad the level of inteligence that M$ thinks we have.
  • As others have pointed out, this was a usability test to find out how long it took to set up a pre-IE versus a post-IE system, not a test of the relative speeds of the two OSes.

    If my memory serves, Trumpet Winsock was a horrible, horrible product that took ages to get working properly. I also recall that Windows 3.1 also had no support for modem speeds above 19.2k, and that was pretty darn shaky as it was.

    But I don't have a clue what this has to do with Internet Explorer integration. 99% of the work done to set up Windows 3.1 networking was spent getting Trumpet to work; the browser always worked as soon as you could ping a remote host using a domain name.

    When Windows95 was introduced, there was an option called the "Plus Pack" which included IE. If you want to see the real benefit in user convenience of having IE "integrated", you would have to compare installing W98 with installing W95 and the Plus Pack (or Netscape on CD ROM). I would argue this difference is negligible.

    As far as I know, nobody even attempts to dispute that it was a Good Thing for the consumer to have TCP/IP integrated into Windows 95, even though it surely damaged the competitive positions of a number of companies marketing Windows 3.x TCP/IP, including Trumpet.

  • I've used KDE, and I don't see how anyone could describe it as harder to use then Windows - it's a virtual Windows clone!

    Now, if you were to say that the applications aren't there, I'd be forced to agree with you - but that doesn't seem to be your argument.

    Incidentally, I now use Enlightenment because I can't resist the cool look. One day, one day, I promise myself, I'll do a Theme of my very own, or at least figure out how to get one of the non-default ones to work.

    But I would cheerfully recommend KDE to anyone who wants a Windows-like experience.

    I guess you could say I'm one of the few agnostics in the KDE vs Gnome wars. :-)

  • It's pretty much obvious to anyone who's actually trying to follow along that Microsoft will lose. So, I'm not surprised.
  • EJECTED? I find that a little hard to believe, most seminars I would suspect (even mickeysoft) are fairly friendly. They want you to buy their product and continue to use it, and those seminars usually ain't free.

    I guess personally it would be bigger news to me that Microsoft is "ejecting" people from its conferences than Microsoft making uneven comparisons.

    (Correct me if I'm wrong Sandy, but if I kill all the golfers won't they lock me up and throw away the key?)
  • This is nothing more than a case of people loving to bash MS. Hey give them a break.
  • i've been having the same problem. Just keep hitting reload, sometimes it doesn't work.
  • Isn't that roughly like saying linux 2.2.x is faster than 1.2.x or MacOS 8.5 is faster than System 7. Or that is faster than . I would hope it's faster, win3.1 is about 5 or 6 years old, isn't it? Since the issue is integrating ie4 wouldn't a more comperable test be win95 vs win95 w/ ie4?
  • They put a winmodem in the 3.1 machine? First of all, I worked Tech support for a large ISP and winmodems, IDEPENDENT of the OS, took much longer to get the freeggin POS's to work in the first place. Also, I was under the impression that Win 3.1, if it doesn't have some sort of error correction or flow control, first time you try to tax the harddrive or CPU justa little bit, disconnection is the result. Winmodems are simple really crappy DA converters with no hardware flow control whatsoever, which is why if you run win (of anything) and you playing an mp3, you disconnect. Basically what I'm trying to say is what kind of a moron would use a winmodem in a test in the first place much less on a win3.1 box. Oh yeah, I forgot, he was told to make 3.1 look like it sux... sorry..
  • I'll give M$ a break when M$ gives me a break. I am sick and tired of the "This is what you really want to do" garbage. I'll tell my computer what I want to do, not vice versa.
  • I think you miss the whole point here. This is being submitted as evidence in a trial. It's not like it was part of some silly ad campaign.

    Microsoft's major contention in this trial is that Windows98/MSIE are inseperable, and part of their defense is that this linkage of browser to operating system provides a superior experience than the "old way." Microsoft has, time and again, fudged the numbers in their favor, even if only slightly. Every little bit counts.

    Give Microsoft a break? When Microsoft starts comparing apples to apples, we'll give them a break. Even if it was a difference between a 33.6k and a 28.8kbps modem, that's significant enough to make it an unfair (and inaccurate) comparison. Apples to apples means: same components, same conditions, only the operating system version should be changed.

    How do we know that Microsoft didn't fudge things further? They could have the Win3.1 box connecting to a UNIX machine, and firing off a chat script to start PPP manually; while the Win98 box connects to a WinNT RAS server. What were the conditions of the test? Without knowing that, it's just a case of more Microsoft FUD.

    If it was you being charged with a crime, and you fudged the evidence this way in your favor, you'd be sitting in jail on a contempt of court charge. Why isn't Microsoft?

    Give me a break.

  • actually, ppp is easier to setup when using ANY gui....
    whether windows or kde or afterstep!
    having said that, i noted the difference between stup on afterstep (approx 2 mins less than a win setup)
    and kde (5-10 mins less)
    (and none of this *(^&*^&P rebooting every step of the process, either..i didnt include the reboots in the time thing, btw!)
    now, i have set up so many windows dial ups for friends who just cant do it...but they all got the kde setup i showed em..
    people dont want windows, they want an easy to use system, with a great gui...
    sounds like linux to me..the o/s with a CHOICE (windows != choice) of guis....
  • Yes, but with Win3.1 (same modem, different machine) I had a lot better luck at connecting consistently, because the interface and scripting was better. On the average it takes me longer to dial in with Win98 than it did with 3.1, and requires more mousing and user input (rather than one click and it dials).

    I'd rather spend 3 more minutes setting it up the first time than 1 minute extra PER DAY to connect!

    Also, were the machines comparable? Win3.1 doesn't recognize memory over 64megs, which, especially if we're talking about IE, could be a factor.

    I agree, however, that the modem difference is small, but probably one of a plurality of factors that dragged the Win3.1 performance down.
  • Er...where does it say that this is a "Linux page?"

    Still, I do agree that "Your operating system sucks" is a poor way to sell Linux.
  • How about their collective, flaccid pud...

    What can their legal team be thinking at this point? "How can I get out of this with a tattered remnant of a career?"
  • According to the report from the Seattle Times, the Win98 system was using a 56k modem while the Win3.1 system had a 33.6k one. The Win98 system could be 25% slower than the Win3.1 system and still appear faster.

    http://www.seat tletimes.com/news/technology/html98/micr_021099.ht ml [seattletimes.com]
  • I'll tell you why MS wants to compare Win3.1 to Win98, rather than Win95 to Win98. Because comparing Win3.1 to Win98 is completely bogus, and has nothing to do with the advantages of "integrating" IE with Windows, while comparing Win95 w/out IE to Win98 is more valid.

    Even that would be stupid. The Connection Wizard is a completely separate function from the browser. This test is yet another attempt by MS to deliberately use smoke and mirrors to hide the real issues. I'm not sure why the DOJ didn't address this point directly. This "demo" is completely irrelevant to the issue of browser integration, because they are demonstrating a function that is separate from the browser.
  • They were originally showing BOTH actual connection speed AND how long it took to actually get connected. However, once the discrepancy in the modems came up, they changed their tune and said the demo was really more about how much easier it was to get setup only. ZDNet has more detail on this.

    In other words, once the flaw was found, they changed their mind on what it was about. They did the same thing with the video from last week: once the government found major problems with the video, and they couldn't reproduce the original results two days later, they changed the label from "Demo of actual problems" to "Hypothetical simulation of theorized problems".
  • I read the article first. And I read it twice: first, when the original blunder happened. At that point, MS was trying to prove BOTH that the setup time was faster, AND that the actual connect speed was faster with Win98.

    Then the blunder happened. It was a big deal in court; the MS witness mumbled that he wasn't sure that they were the same speed, though that was half of what they were trying to show. At that point, I kid you not, this is what one of their lawyers said immediately afterwards: "The difference in speed between the two modems is academic." (ZDNet article here. [zdnet.com]).

    After all that, MS found out the difference in speed, but it didn't matter; they changed their tune to say that the demo was about the setup time, not the actual connection speed. Just like their video blunder last week, they changed what their demo was about after mistakes were found. That's why if you read articles on it now, you'll hear MS say that the demo was really about setup time; but that's only because you are reading the article well after the blunder. If you had seen articles on it immediately afterward, you would have seen that MS meant to show problems with actual connect speed too.

  • It ended up being a minor difference. But as someone already mentioned, the point is that MS was trying to demonstrate difference in connect speed due to SOFTWARE, not due to HARDWARE. The fact that they didn't show similar hardware puts into question the whole demo. I mean, if they couldn't be trusted to make the modems the same, how do we know the rest of the hardware was the same? Say CPU, memory, hard disk, etc.? It is simply an extremely deceptive thing to do, even if the end result is a minor difference. Take this in conjunction with last week's complete screwup (read: falsification of evidence), it's a very serious thing indeed. The entire credibility of the demo is in question.

    Of course, how MS wrote it off BEFORE THEY FOUND OUT THE ACTUAL DIFFERENCE IN SPEED is like this: "The difference in speed between the two modems is academic." Again, this statement is BEFORE they knew the actual speed difference; when the lawyer made that point, MS didn't know what the difference was.
  • The modems do seem trivial, but they also use Win 3.1 instead of Win95 sans browser. I bet even Win 95 is faster than the bloated Win 98, which would just ruin the demonstration.

    It is also interesting that they didn't demonstrate connecting to the Internet on a machine on which the OEM had chosen to pre-install Netscape.
  • Thank god, someone who actually is prepared to say it. I think all the people who slate Microsoft are just a bunch of fashion victims. Most of you are just jealous of Microsoft's success.

    HAHAHAHAHAHAHAHAHAHAHAHAHA, glad you have a sense of humour.

    Oh, you weren't joking. Okay. Then you're an idiot.

    "Lick a rock follow fashion, drop of a conka tree"


    Kick em, while they're down. Who's down again?

    Can't any of you be constructive and do something good rather than critisize people all the time.........

    Well, in theory I can speak only for myself, but I believe that many, if not most, of /.'ers, are hackers - 'real' hackers. Now we have a tendency to be rather constructive people - as ESR said it, we'd rather build than destroy.

    Some of the countless reasons why we hate Micro$oft are: * it practices evil marketing techniques; * it's not Open Source; * their code is ridiculously losing; * their employees are known to be fornicating marketroids; * their software is impossible to properly interface; * they've managed to make the overwhelming majority of computer users work on their ripped-off system, even though it's Not Good, it crashes often and it usually becomes the expert's job to fix the frustrated lusers' computers - until it crashes again.

    If this sounds harsh it's because I'm pissed and England have just lost to France.....

    Well, I'm too tired to make my point completely, but nonetheless I think you can have an idea. Peace, --

  • Boise has had a good trial, but he missed the bus here. The difference in modem speeds had nothing to do with how long it takes to connect to the internet. He should have pressed the witness on what programs were being used to configure the connection. I wouldn't be surprised if the PPP-configuration software would still work after the browser-remover program thing had been run.
  • Get a f**king clue! Unless web pages are used
    for the config screens, then bundling a browser
    makes no difference.
    internet != web

  • Why not?
    They're big, they're powerful, they control the welfare of puny clueless ignorant users..
    Why shouldn't they be as much a target in our society as our political system, our entertainment complex, or our governmental body?

    They really haven't done anything for me to be appreciative of, so I don't praise them.

    They tried to comandeer html and internet standards, they've tried to control the desktop graphics routine(OpenGL is still very much alive and well, thank you!), they try to sell shoddy software, bloated office packages...

    Why should we give them a break? They can afford the heckling!

  • People don't want servers. People wants to play the latest and coolest computer games. Conclusion: People wants Windows!

    People don't want to work. People want to feel good. Conclusion: People want to smoke crack!
  • I always noticed that when someone wants to make there 'product' look good, they just insult the other. This is a linux page, don't point out MS flaws, instead point out Linux's high points. Saying one thing is crappier than your supported angle doesn't make anything better. Leave MS alone, if you hate them so much, ignore them.

"The following is not for the weak of heart or Fundamentalists." -- Dave Barry