Also, I wasn't certain of it, but didn't the pre-Y2K investments companies made provide a lot of oomph to the economy?
I never got the whole Y2K thing since I start at January 1, 1960 when working with dates.
Moby Hick comes from a universe in which Unix was developed ten years earlier.
I take it there's no UnfoggedCon tonight?
I spent 1999 working for a Y2K auditing firm. There were something like 50,000 date-related bugs, mostly Y2K or related pivot-year bugs, in the reports I put together from our analysts that year. My feeling at the end of it, however, was that most of them were minor or would only affect a small part of something, and we already live with vast piles of bugs in our software, so Y2K bugs will just become part of the noise background.
I did develop a habit of always writing dates as "30 Dec 2009" instead of "12/30/09" that year, as well as a preference among alternate date reckoning systems (a 64-bit counter of 100-nanosecond ticks since midnight, 17 Nov 1858).
Sorry - no UnfoggeDCon tonight. Too hard to put something together with traveling across the country. But Capps, Cath, and I have been talking about the possibility of another Con before the Flophouse lease is up.
Allow me to suggest mid-february. I have already ascertained that there are cheap flights then.
develop a habit of always writing dates as "30 Dec 2009" instead of "12/30/09"
Me too! But mine came as a result of dealing with data from international trials where 05/06/09 could mean May 6th or June 5th, depending on the country of the person recording it.
I have already ascertained that there are cheap flights then.
a 64-bit counter of 100-nanosecond ticks since midnight, 17 Nov 1858
But … why?
The most ridiculous thing was the Y2K partners made at the big six accounting firms. Expire much?
On the Greenspun boards, where the first message board I inhabited then flourished, there were two Y2K boards. There had been just one, but about a year after Y2K there had been a tremendous schism, with flamewars and rage, largely over whether Y2K was actually in full effect, we just didn't know it yet. By c.2004, the last time I peeked, most posters had at last slunk away, but a handful still battled it out on one of the boards, over whether the apocalypse had in fact already begun, and we were too dim to realise.
It was all very bracing.
A test in Australia today showed global warming could be fixed cheaply by setting the calendar back 180 days.
11: 17 Nov 1858 is already an important reference point for timekeeping - it's the start of the Modified Julian Date epoch (The Julian Date minus 2400000.5, originally so the MJD could fit in an 18-bit counter). It's also conveniently before the birthday of anyone alive, so it's pretty good for medical records. 100-ns units is fairly arbitrary, but it's nice not to need a separate counter for sub-second precision, and it's what OpenVMS uses, so it's somewhat established. 64 bits is a natural pick once you've chosen an interval that small.
13: You poor bastard. That explains so much.
I too never got the Y2K thing. A two character year (e.g. "00") would presumably occupy 16 bits, but a 16-bit integer can easily represent every year since the dawn of history. Something weird about COBOL, I guess.
A two character year (e.g. "00") would presumably occupy 16 bits,
Someone will undoubtedly correct me, but I vaguely remember the problem as involving binary coded decimal, or BCD (or the IBM version, EBCDIC, or something like that). With BCD you could encode a single decimal digit in 4 binary bits, or get two binary digits into an 8 bit byte. If I checked wikipedia it might all come back to me, but I'd rather not remember more clearly. Some things are best forgotten. I remember playing with BCD to 7-segment decoders in 74XX TTL logic chips to build a digital clock. This was back when digital clocks were really nifty. I can't remember how many chips it took to decode the four 7 segment digits necessary for a clock.
Yeah, I was a little bitter about the whole "Y2K turned out to be nothing!" conventional wisdom, since I spent at least some of 1999 fixing incredibly boring Y2K bugs. Dammit, it wasn't a problem because we geeks are so awesome at our boring, tedious jobs.
To the original point (of the OP):
We've done it once on a smaller scale and we can do it again
We hope so. I take the overall point, of course, but we wouldn't want to think that global warming/climate change is (just) Y2K times 10 or even times 100, either.
It's a quibble, I admit it, but this isn't (just) Y2K 2. It wouldn't be a bad idea if someone, I've no idea who, penned a longish column about just what was done to fend off potential Y2K disasters, and just what that accomplished, in order to demonstrate the effectiveness of, you know, measures. Being taken.
I'm actually curious now which TV shows and blog posts have been mocking the Y2K measures as the work of pearl-clutching nancy-boys.
I believe the post is basically wrong in that Y2K was vastly overhyped and nothing very bad would have happened if much less had been spent on preventing problems.
And I agree with 6.
... we already live with vast piles of bugs in our software, so Y2K bugs will just become part of the noise background.
18 19
As to where 2 character year fields came from, my case is punch cards where you only had 80 columns to play with.
James B. Shearer is really excellent at trolling!
17: It's not that bad. I'm mostly a data analyst, so I don't do the heavy database work. Plus, knowing SAS is a kind of salary floor for me.
Becks gets it exactly right.
Now, who wants to talk about ceilings?
Ceilings are ok, but I'd rather have a roof, you know?
If you are forced to choose, that's the way to go.
Honestly, you'd think even the possibility of a 'Con would merit front page mention.
If forced to choose, I'd take a floor.
No takers for countertops? Lawns? The blue moon?
I cried because I did not have a pretty patterned ceiling. Until I met a man with no floor.
He did have a roof though, which was weird.
If forced to choose between a ceiling and a roof, none of floors, lawns, blue moons, or countertops are options.
Then I met a man with a floor, walls, a roof, patterned ceilings, granite countertops and a two-car garage. I sold him an interest-only, sub-prime, adjustable mortgage with a balloon-payment at year three.
WELL JESUS IF THERE ARE BARELY ANY CHOICES I CHOOSE A PAIR OF SANDALS.
Allow me to suggest mid-february. I have already ascertained that there are cheap flights then.
And if the stars align correctly, you could find yourselves having it on baby snarkfox day!
Surely an occasion worth celebrating.
I too never got the Y2K thing. A two character year (e.g. "00") would presumably occupy 16 bits, but a 16-bit integer can easily represent every year since the dawn of history. Something weird about COBOL, I guess.
COBOL is weird in ways that seem designed precisely to produce the Y2K problem. Data records are generally specified in terms of digits/letters, not bits, so your choice is "two digits or four?" rather than "eight bits or sixteen?". Worse, there's no central repository of data record definitions; each file that uses them has to redefine them, which means that changing the layout of a record (such as going from two-digit to four-digit year representation) has a high risk that you'll miss one somewhere, causing very weird errors. This explains the popularity of the "pivot year" approach, where you decide that you'll keep two-digit years, but instead of assuming that they represent 1900-1999, you decide instead that they'll represent, say, 1950-2049. This works fine (gives you a y2049 problem, but that's yet more diffuse to have bugs spread over 50 years than all at once) except when different programmers in different parts of the program pick different pivots, and one part operates on 1950-2049, when another works on 1930-2029. This kind of thing was probably more common than straight-up year-wraparound errors.
C only has one major native source of Y2K errors, which is that there's a standard date structure which includes a 'years since 1900' field. That can go to more than 100, but lots of formatting routines were printing "19%02d", and would go from 1999 to 19100, instead of printing the value + 1900.
MUMPS was probably my favorite obscure language (sort of COBOL-flavored Perl from the 1960s, used for medical databases) that managed to totally dodge the problem; it used pairs of integers for time representation, where one was the modified Julian day number, and the other was the number of seconds that had passed in the day. Pretty innovative for the era.
Y2K is totally mockable, because the fears lagged way out of line with the responsible steps taken. Most people were completely unaware that there was any sort of issue until they heard the message "Responsible people have been diligently fixing and reviewing miles and miles of computer code, so don't worry, but here are the basics of their concerns," and otherwise rational people responded "YOU MEAN PLANES ARE GOING TO FALL FROM THE SKIES???"
Y2K is totally mockable, because the fears lagged way out of line with the responsible steps taken.
Everything is mockable. Come on. Orange formica is mockable. Granite countertops are mockable. Popcorn ceilings are mockable. Some guy wearing a John Lennon t-shirt while wearing John Lennon glasses (at the farmer's market) is mockable. We might want to get over the mocking thing every once in a while. It's simple-minded.
If some people's fears about Y2K caused them to lay in stores of bottled water and refuse to get on an airplane, well, that may have been silly in retrospect.
It's also mockable because of all the gibberish spouted by the people who understand all that stuff. It is even real? Only a geek can tell if the other geeks are faking.
Nothing happened, therefore they probably were.
It was mockable all December 1999. I mocked it. Outloud.
I can't imagine mocking a guy in a John Lennon shirt.
I can't imagine mocking a guy in a John Lennon shirt.
In fact, the events of December 31, 1999 did nothing to increase or decrease the mockability in my eyes. If anything, it became less fun because everyone now agreed how it would go down.
40 sounds exactly right to me. (Though surely the media tended to exaggerate the scope and significance of the Y2K survivalist movement, members of which were apparently stocking up on canned goods and ammunition because they didn't trust the computer programmers to fix the glitches in the code?)
I mocked it. Outloud.
You mocked it aloud. Bono dreamt out loud. Haste, haste to bring him laud.
I can't imagine mocking someone who would double-post.
I was in Edinburgh in December 1999. What I most remember about Y2K was the millenial Hogmanay.
Everything is mockable. Come on. Orange formica is mockable. Granite countertops are mockable. Popcorn ceilings are mockable. Some guy wearing a John Lennon t-shirt while wearing John Lennon glasses (at the farmer's market) is mockable. We might want to get over the mocking thing every once in a while. It's simple-minded.
I find a lot of humor in the world, sourpuss.
44 and 45: LB got annoyed about some guy in a Lennon t-shirt once, because he looked kind of like Lennon as he was wearing it. I don't look up the post. In the field of mockability, as that works, it was mockable, sure, but so is pretty much everything.
Chickens say bawk. Interrupting cows
I think Standpipe's looking for a cock joke.
I'd have no problem mocking someone for mocking someone in a Lennon shirt.
We went to Chico Hot Springs, one of the best places on earth. And, coincidentally, we're here again now. I just realized that tomorrow we're going dogsledding on the last day of the decade having done so on the first day. (By which I mean 1/1/2000 -- don't mock me for calling that the first day!)
39: Revelation, a spin-off of the Pick OS avoided the Y2K problem but I did have to check it out for a couple of old clients back in Alabama.
Given some of the hyper-alarmist claims it's no wonder some people were stocking up on food, water, and ammo (which one needs also in case there's an earthquake).
We went to Chico Hot Springs, one of the best places on earth.
In general, I agree, but when Magpie and I were there in '04 there was (what I surmise to be) a rape or assault just down the hall from our room. (The room was taped off, I saw a cop comforting a young woman, and a guy got marched out through the lobby in handcuffs.) It kind of cast a pall over what had otherwise been a very pleasant stay.
Lamb necks, at least around the vertebrae, make for kind of inconvenient eating.
Mares eat oats, and goats eat oats, and ivy eats little lamb necks.
Pretty delicious, if I do say so myself. Braised with white wine and quince. Brussels sprouts and bread with muhammara alongside.
Neb eats lambs and Nosflow eats lambs and little lambs are braised with white wine and quince. A kiddleat little lambs braised with white wine and quince, too, wouldn't you?
What awful luck. Some people really suck.
I just learned that Chico has a wing dedicated to Warren Oates.
So that would be the Hall of Oates?
I'm going to post, then go to bed. Enjoy! Write clever things for me to mock in the morning!
Only a geek can tell if the other geeks are faking.
This is true, and it's a problem in all kinds of areas. Climate change (as Becks mentioned) and economics and law and medicine and innumerable other fields. It's somewhere between difficult and impossible for a non-specialist to differentiate between bullshit and current accepted wisdom. That leads to things like climate denialists and evolution denial and crazy legal beliefs and quack medicines and a host of other nonsense.
I wish I knew what could be done to fix the problem so that more people could do a better job of figuring out what to believe.
I'm going to post, then go to bed.
I most royally shall now to bed,
To sleep off all the nonsense I've just said.
Me too.
71
This is true, and it's a problem in all kinds of areas. Climate change (as Becks mentioned) and economics and law and medicine and innumerable other fields. It's somewhere between difficult and impossible for a non-specialist to differentiate between bullshit and current accepted wisdom. That leads to things like climate denialists and evolution denial and crazy legal beliefs and quack medicines and a host of other nonsense.
This isn't really true either or at least doesn't identify the real problem which is people have trouble being objective about things where they have an emotional stake. It isn't all that hard for an intelligent layman to get a general picture of technical issues. But you have to be able to put tribal or partisan or emotional or religious allegiances aside and it is this that some people find difficult to impossible.
My question is: are we ready for Y10K?
I compromised between panic-based stockpiling and assuming nothing was going to go bad: in addition to my normal weekend shop, I bought three tins of carrots.
I also attended the worst party I ever went to IN MY LIFE on Millennium Eve. Actual worst bit: I had left the party by the time the host's GF arrived, not knowing he was even throwing a party, to find him doing coke and another woman in the bathroom. The 2000s went downhill from there on, for me.
As one of the hacks involved in correcting reams of Y2K code, I can safely say that there was a non-negligible risk, not of planes falling out of the air, but of companies losing a lot of money through non-compliance on performance indicators and other such unromantic stuff.
It is also true that plenty of companies lost money by paying squillions to fly-by-night consultancies, but this was an instance of bad management, just like most instances of companies being taken in by sharks who are cleverer than their own managers.
The outfit I worked for, like the vast majority, costed the risk in good time, and allocated sufficient resources to do the drudgery necessary to see them all right. It was profoundly dull, but contra Shearer et al. it addressed real issues and would have cost a lot of money if it hadn't been done.
On "Millennium eve", I spent my time arguing that the millennium really started in 2001 and there was nothing special about the year 2000.
The Millennium actually really starts tomorrow.
but contra Shearer et al.
I don 't think anyone is claiming that there was no technical problem associated with Y2K. But there was a lot of panic-mongering in respectable circles, and panic was inappropriate. Also:
It isn't all that hard for an intelligent layman to get a general picture of technical issues.
James gets it right. I knew to a reasonable certainty before Y2K that no serious problems would ensue.
It is, in fact, global warming denialists who would find the Y2K experience rhetorically useful. After all, it was a situation where the experts, while agreeing there was a problem, disagreed on the size and tractability of the problem. And in the end, nobody was much inconvenienced by the solution. "No doubt the panic-mongers are exaggerating about global warming, just like Y2K. We'll work out a solution when the need arises, just like Y2K."
Stephen Jay Gould had the best answer on the issue of the beginning of the millennium. He asked an autistic savant, who responded that the millennium begins on Jan. 1, 2000.
The savant explained: "The first decade only had nine years."
75.last: So the worst party of your life was the best part of the '00s for you? Geez.
76
The outfit I worked for, like the vast majority, costed the risk in good time, and allocated sufficient resources to do the drudgery necessary to see them all right. It was profoundly dull, but contra Shearer et al. it addressed real issues and would have cost a lot of money if it hadn't been done.
I am not saying it was unnecessary to do anything. Just that there was no good reason for the alarmism (much of it obviously driven by millennialist religious hysteria). The conversion to decimal stock pricing also required some care but was accomplished without all the drama.