Cory Hamasaki's DC Y2K Weather Report V2, # 11 "March 11, 1998 - 660 days to go." WRP67 FINAL (c) 1997, 1998 Cory Hamasaki - I grant permission to distribute and reproduce this article as long as this entire document is reproduced in its entirety including this notice. I do not grant permission to a commercial publisher to reprint this in print media. As seen in USENET:comp.software.year-2000 http://www.elmbronze.demon.co.uk/year2000/ http://www.kiyoinc.com/HHResCo.html Please fax or email copies of this to your geek pals, especially those idiots who keep sending you lightbulb, blonde, or Bill Gates jokes, and urban legends like the Arizona rocket car story. If you have a Y2K webpage, feel free to host the Weather Reports. Don't forget- April 2, 3 1998, Geek Out. Project Dumbass needs you. In this issue: 1. RATES! 2. Yuppies 3. FAA, source info 4. errata - hams 5. ALC 11 6. CCCC ------------ Rates ------------------ Solving the Year 2000 problem. 660 days now. Divide the number in half to account for weekends, holidays, sick leave, vacations, meetings on the pension plan, the new parking policy, the Lan failures, the Superbowl, etc. Still 10-15 years of work to do, but now, less than two years to do it in. I was worried, well, actually optimistic that this would have caused a programmer hiring panic in 1997. It didn't happen. Denial was too great. Yes, rates did go up a little. Journeyman COBOL programmers who went begging for $35/hour in 1996, closed out 1997 at an easy $50-60 or slightly higher. Expert mainframers with specialties, available in 1996 at $50-60, closed out 1997 at about $100. These are approximate 1099 numbers and vary widely by geographic region. Companies that read the early WRP's and took action have benefitted by having a stable workforce, a full year's worth of remediation and some are entering the test phase. Those companies have used the WRP's to monitor the pulse of the remediation industry and to address their HR issues proactively. However, however, we are entering the most dangerous phase of Y2K remediation for companies... ...and I am not going to give away the ending in the WRPs. That story will be in a for-fee, Corporate Edition DC Y2K Weather Report. It will carry the same irreverant editorial content, cursing and grumbling from the Geek slave pens (you need to know what they're thinking and scheming.) You want some rate stories? Here's a good one. A mid-low end geek was getting itchy. He was earning 54K, told his boss to get him more money. At the same time, he went on ONE interview, planned to ask for 75K and was prepared to settle for 70K... well maybe 69K. Before he could shoot off his yap, the interviewer, a VP at the company, wrote a number on a piece of paper and said, "Will this do?" The number? $110K. True story, happened two weeks ago, his boss is now down 2 geeks and doesn't have the budget to replace them. The boss, a team lead earning mid 60's and has a masters and 15 years in IT. ...that's 110K, W2, full benefits. Here's more to the story, the boss's boss is considering bailing out too, he also has a masters, earns $75K and has an offer for $125K. There are discontinuities in the workplace. There are still programmers available at under $70K but not for much longer. These are dangerous times for companies. There are things you as a company can do but again, no free rides for corporations. ...and please, corporations, don't try to think this one out yourself, you'll have to think way, way outside THE BOX. I expect the crisis to extend until 2003. Take the mid-geek in the story above. He can't afford to stay, none of them can and they're well paid, 54K, mid-60's, and 75K. You want more rates? I know a supercranker earning 105K; call came in from a pal, we need you, how about $160K? It's getting weird out there. ----------- Yuppies ---------------- ...I am very disheartened by the lack of remediation. At this point, we, the Information Technology business should be deciding which 20% of the systems are absolutely necessary. We should be dismantling the power generation 'grid' and making each city go it alone. It's not happening and the contingencies are not being prepared. When I say I'm optimistic about Y2K, that doesn't mean I'm minimizing the problem. My car has 4 wheel disk brakes with redundant systems, it also has a mechanical third braking system actuated by a cable. Although it has dual airbags, I always wear my seatbelt, a three point inertial reel harness. I don't drink, I don't tailgate, and for me, speeding is 5 over the limit. I have medical insurance and to back it up, I have life insurance. I count 6-7 layers of systems to address the problem... hey, there's a large object in my way, the problem of 30,000-40,000 deaths in cars in this country each year. When I look at Y2K, I can clearly see the systems failing. I know the remediation is not being done, I know the poor quality of work, the small amount of work done to date; it doesn't look good. What I can't see is whether Y2K will kill 50 people or 50 million in the U.S. If the number is under 500,000 deaths due to riots, starvation, powerplant explosions, I won't experience it personally unless I am very unlucky. Even 500,000 deaths, while a large and exciting number is about one person in five hundred. In any year, cancer, heart disease, influenza will kill more than that. To achieve the 500,000 person level, we'd have to see places like NY, East Saint Louis, South Chicago, burn for weeks. Or some kind of civil war would have to break out. The problem is that Y2K has never happened before and is one of those discontinuities that happens in .... a thousand years. Y2K will have an impact like the computerization of the last 35 years... except it will happen fast, almost overnight and it is not a positive change, it is an undoing of the efficiencies of automation. Last night, I saw a Sprint commercial... or was it Folgers. In the commercial, a Yuppie woman is shown logged on to a network, drinking a cup of coffee, the scene is a high-tech, stylish, very modern office, she's looking at a spreadsheet and obviously thinking grand Yuppie thoughts, a smirky, butthead smile on her face. ...a spreadsheet? Spreadsheets are scary. It is hard to know what's going on in them; easy to switch a +AA123 with a -AA132 but we use them because they are easy for the dullest to click 'n drag. How can society afford to house, cloth, feed the near brain-dead Yuppies and Gen-X'ers, pay them to sip coffee and smirk at spreadsheets? Spreadsheets which are wrong, guarenteed incorrect (unless independently verified by a professional). How? The back room systems, the S/390's, COBOL, 450 page per minute Enterprise laser printers, VSAM databases, are so efficient, provide such a high return on investment that it doesn't matter that cities are full of drones, drones who click on spreadsheets, attend meetings about the Holiday party... in September. Have a birthday party for someone every week. Welfare for the well dressed and connected. The dole for the useless and lazy. The back room systems were hand crafted by programmers who earned a big $4.50, $6.25/hour in the 1960's and 1970's, who were right-sized in the 1980's and 1990's because their software was too good, too efficient, too reliable, because it didn't have to be made new every year. That rule is over. It's time to pay for 15 years of deferred maintenance and the only way to do it is to get rid of the drones, fire them, fire them now and crank up the programmers who understand Enterprise systems. Here's the choice, you can buy them from the current market, or you can try to hire them Post Y2K when any amount you name will be topped by someone else. I've seen one company take action. They fired all their meeting attenders and kept the real programmers. If you talk to their IT director (who I know personally), he will tell you that he has geeks who will code til they drop; yes, he got that expression from me and uses it to express his policy, code 'til they drop, code well and live. Is this harsh talk? Ravings or a plan for corporate survival? Some corporations have started to stockpile diesel and other essential materials. Hints for the clueless. The programmer hiring boom is yet to begin. If you think Citicorp and the others are desperate now, wait until January 2001 when the tricked up data hits year-end processing for the first time, when everyone knows that the offshore factory jobs have failed. ---------- Gooood FAA Info --------- click up http://www.gao.gov Yes! the United States General Accounting Office, GA-Oh, GA-Oh Day light come and I want to go home. <There's DD, drinking rum from a coconut; Frank's playing the steel drums; Sing it Greg, sing it!> Order File Number: GAO/T-AIMD-98-63 Testimony Year 2000 Computing Crisis FAA Must Act Quickly to Prevent Systems Failures Statement of Joel C. Williamssen Director, Civil Agencies Information Systems Accounting and Information Management Division It's eight pages of yum-yum, good reading. Here are a few excerps: 'With only 696 days remaining....' 'Hundreds of critical FAA computer systems make its operations possible; without these specialized systems, FAA could not effectively control air traffic, target airlines for inspection, or provide up-to-date weather conditions to pilots...' '...too slow... ...current pace, will not make it in time... ...severely behind schedule...' '...quickly running out of time, making contingency planning even more critical.' '...cause for serious concern...' '...domino effect...' It's an incredible, butt clinching, wake up screaming piece and IT'S NOT PAUL MILNE, it's not a DC Y2K Weather Report written by the king of cluelessness; it's congressional testimony and available free from the GAO. The GAO will send you a printed copy for free, send a fax to (202) 512-6061. It's very professional, has the GAO seal on the cover, beautifully formatted, has lots of hot words, risk, emergency measures, critical... it's very scary. You can download the .pdf from their website too. Get your copy and use it to swat the denial-heads. Wack-em on the nose. The nerve of some people, how dare they say, I'm a commercial pilot/air traffic controller and I know there's no danger of planes falling from the sky because of a Y2K computer bug. Don't you denial-heads get it, you're meat-ware, the bio part of the system. Sure you have expertise and an important job but you are just one component of a system. Just because you're a pilot or ATC, that doesn't mean your expertise extends into complex software systems, it doesn't, get over it, keep your keyboard buttoned. --------- Errata ----------------- DC WRP mentioned EEB as a source of Ham Radio transceivers. EEB closed its doors a month ago, an alternate vendor is: http://www.aesham.com 1-800-558-0411 AES will send you a 160 page catalog on just ham equipment, more knobs, dials, buttons, LCD displays than you've ever seen before. Also check out: http://www.icomamerica.com http://www.kenwood.net http://www.yaesu.com http://www.ameritron.com http://www.mfjenterprises.com http://www.vectronics.com http://www.mirageamp.com These are the home pages of the manufacturers, lots of spec sheets and technical info. -------- ALC 11 -------------- The dump. Read from the top, <my comments, oops yyddd alert.> JOB ASM STEP GO TIME 050316 DATE 98047 ID = 000 CPUID = 010001580158 PAGE 0001 COMPLETION CODE SYSTEM = 0C1 PSW AT ENTRY TO ABEND 078D1000 000A5F98 ILC 2 INTC 0001 <Yes, this is a 24 bit pre-XA MVS system. The address counter portion of the PSW, the Program Status Word, contains the address A5F98.> <formatted ASCB, TCB deleted> ACTIVE RBS < RBs deleted > LOAD LIST NE 00000000 RSP-CDE 00FF9D80 CNT 00010001 CDE 9ECAA0 NCDE 00000000 RBP 009EC980 NM GO EPA 000A5F58 XL/MJ 009ECEC8 USE 00010000 ATTR 0B20000 FF9D80 NCDE 00FFA128 RBP 00000000 NM IGG019DK EPA 00F99000 XL/MJ 00FF9DA0 USE 00010000 ATTR B122000 <We loaded a module named GO at A5F58. A5F98-A5F58 = 40 hex. Look for that in your assembly listing.> XL <deleted> TIOT 9C3020 JOB ASM STEP GO OFFSET LN-STA DDNAME TTR-ST STB-UC + 0018 14010100 PGM=*.DD 9CF4C000 800024F8 + 002C 14010102 SYSUDUMP 9CEE8000 80000000 <The TIOT, Task Input Output Table is related to the JCL and DD names.> VSM <6 lines deleted> ***TCB SUMMARY*** ----------- <about 120 lines deleted> REGS AT ENTRY TO ABEND FLTR 0-6 0000000000000000 0000000000000000 0000000000000000 0000000000000000 JOB ASM STEP GO TIME 050316 DATE 98047 ID = 000 . PAGE 0007 REGS 0-7 009ECED8 000A4FF8 000197B0 00000000 0000197B 000A5FFC 009C3018 FD000000 REGS 8-15 009ECA28 809EC6B0 00000000 009EE080 400A5F5E 000A4FB0 0001BBB0 000A5F58 <Reg 4 has 000197B in it.> ACTIVE LOAD MODULES <The hex on the left is the absolute machine address. The hex in the middle is the program and date in storage. The EBCDIC is on the right. Some people put eyecatchers in their programs and look for messages like, "This is my data". In this simple example, you should be able to match every byte in the assembly listing to the hex, including the data areas. The key is knowing where to find the instruction address (PSW), the registers, where the module was loaded (the CDE) and understanding the base displacement addressing. Referring back to the assembly listing: 000010 5840 C08E 00094 23 L R4,COUNT means R4 = count; /* put the 4 bytes of count in R4 */ C 08E means take value in Register C, add 8E to it. A5F5E + 8E = A5FEC get 4 bytes from machine address A5FEC, COUNT, and put them in R4. <linkpack area or jobpack area module.> LPA/JPA MODULE GO 0A5F40 90ECD00C 05C050D0 * ........* 0A5F60 C04A41D0 C0461744 5840C08E 88400004 41300003 4150C092 41404001 18248920 *......... ... ........... .....* 0A5F80 00045020 5000960F 50034150 50044630 C01A58D0 C04A0000 98ECD00C 41F00000 *.............................0..* 0A5FA0 07FE0000 00000000 000A4FB0 00000000 00000000 00000000 00000000 00000000 *................................* 0A5FC0 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 *................................* 0A5FE0 00000000 00000000 00000000 0001978C 0001979F 000197AF 000197BF 00000000 *................................* <A5F98 is the 98 EC D 00C, the load multiple. It's preceeded by 0000, our forced 0C1. S/390 updates the PSW before executing the instruction. Notice the years starting at A5FEC. The entire program and data area is contained in the few hex lines above.> <hundreds of lines deleted.> If you took the original assembly language program and ran it on your MVS or OS/390 system, you should be well on your way to puzzling out object code and sysudumps. It's important that you took the time to memorize the few dozen instructions and instruction formats. Also the Base+displacement and Base+Index+displacement addressing of S/390 is the key to reading the dumps. I glossed over a lot of stuff, these lessons don't make you an assembly language programmer but it should give a COBOL'er or PL/I'er enough info that they can tear apart the dumps that we will be getting. We WILL be getting. I can't work with you directly but if two or three COBOL heads will team up on these lessons and help each other, you shouldn't have a problem. Pee Cee Wee Nee's who know the Intel architecture will realize how much simpler and cleaner the S/390 is. Although VAX-heads might complain that the instruction set is not as orthogonal as the VAX. Any Intel assembler of VAX macro assembler programmer should easily understand S/390 assembler. Denial-heads and the clueless who don't think that programmers deserve every cent are invited to critique and clarify any part of this lesson. Yes, I am mouthing off to the butt-head who made the crack about programmer's tiny hands last year. What an idiot. I don't care why you think you're special but if you can't fix this stuff, you're useless. You're worse than useless, you're a cause of this problem, butt-heads who dare to shoot their mouths off, on the next Mauri. The program that we've been discussing for weeks is a simple, contrived example, a couple dozen lines of code. Production systems are hundreds of thousands of lines, millions of lines in some cases. -------- CCCC ---------------- So here it is, 660 days left, denial is over, the panic is raging in enclaves. I want to talk to you geeks again... find a seat, the floor is fine... We've got a big problem, some heavy choices to make. Do we run, abandon the cities, or stay at the helm, do what we can to fix the systems. This is something that each geek needs to decide. There are c.s.y2k'ers who have headed for the hills. I can't, won't tell you to stay or go, you decide for yourself. I can't tell you that remediation will make a difference or not. This is not within my experience. Yes, I know about enterprise scale systems, understand the clueless who infest the corporate world, can analyze complex situations in new, breakthrough ways. What I haven't been able to do is see past the barrier at December 31, 1999. My guesses and what I will do. Guesses - We're looking at a complete collapse of the government's systems and partial collapse (50%) of private industry's computer systems. Analogous to the dissociation of the former Soviet Union. 10-20% of the military will resign when they aren't paid for months. Rioting, looting, and burning in the usual places. Avoid Miami, New York, Annacostia, East St. Louis, L.A., etc. A nice fearsome recession, DJI down 5000 points in 6 months, hyper inflation for a couple years, the usual brushfire wars in unstable 3rd world countries. Crop production off 20%... some deaths but less than 50,000 directly attributable to Y2K in the U.S. Lots of personal financial misery. At 50,000, we're looking at one death in 5,000. You'd have to be very unlucky to be that one person. This is just a wild-assed guess (WAG) and could be off by an order of magnitude. What am I doing - keeping my options open. My pal's farm is one of my options. I will invest some time and money in it this year. I see three possible futures and responses, 1 month of excitement, a rough ride for 6 months, and forever. 1 month, the minimal state, real confused computer systems but civil order is maintained, food supplies and other goods and services remain intact. Deaths? a few hundred, mostly the elderly, infirm, weak, or extremely unlucky, stock market gyrates and ends up higher as companies rebuild. 6 months of power interruptions; when the foodstamps don't come, localized riots, looted grocery stores; the smart thing to do would be to escape to the farm until the dust settles. This is a rough ride but the deaths are less than a hundred thousand, maybe only a few thousand; the major hit is the financial loss, stock market falls 5000 points. Total collapse, forever on the farm, a scratch out an existance life style in a Ted Kaszinsky-style shack. Giving it all up and rebuilding from zero. No one wins if Y2K takes us down to total collapse. In this scenario, civil war breaks out and there are millions of deaths in the U.S. Money is worthless. I own a fine assault rifle, a Colt AR-15 model SP1, in this scenario, I'd need it. Suppose I covered my bases this way; this spring, help my pal plant a half acre in beans, squash, pumpkins, etc. Kick in a hundred bucks for diesel. Build a 12x12 shed and store a wood stove in it. Cache beans and rice in the shed. Get a .22LR adapter kit for the AR-15. Put a 3 amp solar panel on the shed. Not great but survivable. 3 amp solar and a car battery will run an 8 watt florescent tube and my ham gear on receive every night. 8 watt florescent is like a 40 watt incandescent bulb. If things turn out well, my pal gets a nice storage shed and a hundred gallons of diesel. Monitor the situation and be prepared to bug out. If society collapses, I'm not especially interested in a palace in the woods. Cover your bases, monitor the situation, take care of helpless and aged, make your preparations, crank code 'til you drop. We control our future. cory hamasaki 660 days now.