Tuesday, December 31, 2013

My father's "formation of proper habits" and the workplace

On this New Years Eve, as I “review my life” and walk back (like Benjamin Button) a few decades, I’m quite struck by how long it took me to learn the proper maturity for the IT workplace, especially as it was in the largely batch mainframe environment in the 70s and early 80s, continuing all the way to Y2K in the culture of elevations and nightly (and end-of-month and end-of-year) cycles.
Or, say it’s about “formation of proper habits” as my own father used to lecture.
Maturity in the workplace had a lot to do with understanding the damage you could do before it was discovered.  So the recent flap at Target (in my favored city Minneapolis) has plenty of precedence in the old mainframe world.
One habit had to do with being extremely careful with production files.  It wasn’t until the late 1980s when mainframe installations started implementing security software like “Top Secret” (or simply RACF) that limited ordinary programmer update access to production files.  Chilton (in Dallas) did this in 1987.  At NBC, from 1974-1977 on a Univac 1110 system, we had old TTY terminals at our desk.  You could keep paper trail of your work to prove you did not access production files.  It was important not to lose the paper rolls.
Likewise, production closings at end-of-month were critical.  You learned not to schedule vacation across them.

Another risk was loss of source-load module integrity.  It wasn’t until around 1990 that installations started using source management software like Panvalet, CA-librarian, Endeavor and ChangeMan rigorously.  These packages would require that a source element by locked or “processed” before it could be promoted.  Some companies had these packages for a while before they woke up to the fact that they needed to be used correctly for security.  It was important for employees or associates to pay attention to the capabilities of these packages and follow the rules even if they weren’t always enforced at first.  

Friday, December 27, 2013

Misadventures with a Toshiba Satellite synpatics touchpad under Windows 8

Well, I’m having misadventures with the Synaptics touchpad on my Toshiba Satelitte P875-S7102.  It’s jerky and stalls, and sometimes the Windows 8 key becomes inoperable if I use it.

I bought a Logitech keyboard and mouse some time back and that always works. Well, with a couple of caveats.  If the system goes to sleep and the Windows Action Center starts to run, and I shake the mouse to wake the computer up, I sometimes get a “green screen of death”.  But the machine comes back OK if I press the power button on the main laptop once, then it unlocks everything. So apparently the driver code on the Logitech is not accessible when Action Center runs.

The other thing is that I can get the touchpad to wake up by touching it with two fingers.  I’ve read that there are three pads underneath on the Toshiba instead of just two.  But then, when I start using the Logitech, if it goes to sleep and wakes up, the pointer keeps jumping, until I touch the touchpad with two fingers a second time to “close” the original Toshiba driver.  Make sense? 

Today, I bought a Logitech mouse pointer at a Best Buy, because I would need a smaller “pad” to travel with. 

Before I install it, I’ll try to get to the bottom of the problems with the original Toshiba touchpad driver.  It seems to be software, not hardware, based on the evidence.

Maybe a Windows 8.1 install (which requires Toshiba pre-installs) would fix it?
The best site I can find to start looking into this is eHow, here.  Oh, yes, the first time I went to it, Windows stalled on a “cache” issue, and worked the second time after I backed out of the site.  Google Chrome does this, but displays a “waiting for cache”.  Firefox can do it too. 

Tuesday, December 17, 2013

A visit to software alley near Raleigh NC brings back career memories (like SAS programming)

Whenever I drive through North Carolina, around the major cities, the state looks more modern than a lot of Virginia.  Monday I passed through Cary and then Research Triangle Park.  In the later, I saw two entrances (along US 70) to IBM, both gated near the street so there no good place for a picture.  In Cary, the SAS Institute (wiki ) is right off I-40, on the south side, and appears to have a huge and mostly gated campus.
SAS stands for Statistical Analysis System, that comes with a “4th generation language” for retrieving and reporting data that became very popular in the late 1980’s on mainframe systems. A typical SAS “program” would comprise a series of statements (for mainframe SYSIN input) that have the components of a typical batch program, but easy to code.  There would be a DATA step that would describe the format of the input dataset (like a COBOL FD layout) and then build it as a “free form” SAS dataset where the elements could be referred to in subsequent processing steps as if they were names of elements in a table.  SAS Procedures could output reports or other files, and could easily sort and merge files by various keys,  Database formats (like IMS, etc) were available to DATA.  The language became so popular that some installations hired SAS gurus that did nothing else.
At Chilton, in Dallas in the 1980s, the monthly billing job had a SAS merge step before printing the final member statements from an old ALC program. 
At Lewin-ICF, where I worked in 1988-1989, SAS was used heavily to prepare reports for lobbying clients on health care (most of all Medicare operating margins for hospitals).  I remember one report for a podiatry association required data being grouped into “bundles” and that was done easily in SAS. Lewin also used SAS on the PC, because in the computing environment then (an IBM 4341 and 4381) it had to pay for mainframe use.

ING-ReliaStar-USLICO did not use SAS, but instead preferred DYL280 (whith followed the old DYL260 which looked like RPG).  Another popular 4GL was Easytrieve.  

Near the SAS Institute there appears to be a private high school with a most interesting softball field (and foul lines that measure exactly 200 feet), and a half mile away is a wooded state park, whose pine-tree lined trails reminds me of the bivouac at Fort Jackson, SC (160 miles farther south) in 1968, in Basic Combat Training, or perhaps Ft. Bragg (60 miles away).   

Friday, December 13, 2013

Startups fire unfocused new hires very quickly -- they have to

Startups are quick to fire new hires who don’t have the right stuff.  And many hires find that old-fashioned marketing skills, including cold-calling, are far more important than they had expected in this era of “self-sufficiency” and privacy on the Web.  The tension that this produces is odd, and may help account for some of the grossly inappropriate marketing pitches (not just spam) that we all see these days.
That’s the gist of a “Marketplace” story in the Wall Street Journal Friday, At starups, pink slips come early and often; workers who don’t measure up can be gone within weeks; strategy changes devalue some skills”, in a story by Stephanie Gleason and Rachel Feintzeig, link here
One problem for job seekers or hires is that many startups have very narrow focuses on customer needs.  Grandiose results like those of Facebook, Twitter and Google are rare.  The startup “Plated”, belonging to Nick Taranto, for example, sells ingredients for home-cooked meals.  It’s hard to sell a narrow service or item.  Startups that develop new mobile apps (like some auto-parking-finder apps described today on ABC’s World News Tonight) might make a better fit for more technie-like job seekers.  Yet, selling practical household items (some of them make up those 800-number sites that advertise on CNN) may make up a significant portion of the market.  I can think of other startup ideas (like laser depilatory products, often on cable television) that would obviously seem problematic for many people. 
The WSJ article noted that about 6% if people in “established companies” get fired or laid off a year (Facebook is too young to be “established”), whereas about 25% get the ax in startups.

Does anyone know what working at Facebook is like?  

Tuesday, December 10, 2013

H-1B program IT workers has sometimes destroyed jobs for US citizens (so they say); should severance have strings attached?

There has been a lot of attention recently to the H-1B Visa Program, and Mark Zuckerberg’s plan to increase visas for foreign workers with specialized technical skills.
But the book “Who Stole the American Dream?” by Hedrick Smith (2012) argues that H-1B workers are indeed taking away IT jobs from Americans, even though legally they are not supposed to be paid less.

I found it common throughout the 1980s and 1990s to work with mainframe programmers from India and Pakistan.  It was an everyday thing, and some workers were very skilled.  No one ever mentioned religious issues pre-9/11.  
On p. 293, Smith describes an incident in 1994 at AIG (the insurance company that tanked in the 2008 financial crisis due to insuring so many credit default swaps) where associates were called in, told they would be laid off in 60 days, and they would have to train H-1B replacements to earn severance.
In the later part of the 1990s, Y2K work was often outsourced to India, or given to H-1B hires, but there was so much work in 1998-1999 that generally people did not believe that regular mainframe jobs could be affected.

The AIG thing tocuhed an odd coincidence. Last night, I had dinner at Kammerbooks on Dupont Circle in Washington.  I used to eat there a lot when I worked for "The Consolidated Consulting Group" which would later merge with Lewin, to become one of the top health care policy consulting firms (by running simulations and reports of policy choices) in the area. Dinner there brought back memories. I happened to sit next do a young man who, at the briefest glance, appeared to be coding java on a laptop (no, I didn't hack); when a girl friend showed up, she started talking about her job offer or internship or something at AIG -- just before I found the story about AIG in the book. What a coincidence, or providence.  The young couple talked loud enough for me to hear everything, and they never mentioned the 2008 crisis.  It was if they were oblivious to recent history.  
To return to the year of Y2K: I was working at ReliaStar in Minneapolis in 1999, we got several hires or consultants (including one project manager) who had been let go from Prudential in Plymouth, MN on Feb. 4, 1999.  People described coming to work that morning and finding themselves locked out of their logons, and they weren’t told what was happening until late morning.
During the wave of layoff at ING-Reliastar, not related to visas, at the end of 2001, some associates were kept until Feb. 15 and told they had to meet certain goals to get severance, but others (including myself) were suddenly cut loose, although with full severance (and health insurance during the severance period, which was pretty generous) suddenly in mid December.  My own logon failed while I was on the phone talking to an internal customer about a problem. 
There is another question here, does severance come with strings?  Usually the employee signs an agreement giving up the right to sue (and often paperwork is shuffled to prove illegal discrimination, especially by age, did not occur, by breaking employees into small groups).  In a few gases employees have to sign non-compete agreements, not to work for a competitor, but I did not encounter that.  If it happens, it is troublesome (and sometimes it is enforced).  I would wonder if non-disparagement clauses (regarding internet and social media postings) could happen in the future.  Again, I haven't seen that.  Of course, confidentiality agreements have to be honored, but usually employees have to surrender all materials and have no further access.  

Tuesday, December 03, 2013

Back-ground mainframe processing, and communications among players, for Obamacare has many problems because of lack of mature personnel

It’s becoming ever more apparent that the treatment of older mainframe professionals in the period following Y2K, encouraging early retirement and buyouts and offering at best a questionable mechanism of temporary W2 assignments in various cities (in specific areas like MMIS) has damaged the workforce available to get the processing behind Obamacare right.  The mature talent, used to the full systems development life cycle and the pains of implementation, is no longer around.
On Monday, the New York Times, in an article by Robert Pear and Reed Allison, reported “Insurers claim health website is still flawed; White House praises software upgrade”, link here. The article has a link to a mult-media Visio-style systems flowchart showing how the processing among exchanges, insurance companies, the federal government, and individuals and employers is supposed to work.  The chart reminds me of the overview of the “Combined A&B Medicare System”, or “CABCO”, that seven Blue Cross and Blue Shield plans tried to develop in Dallas, TX in the early 1980s but failed because of turf bickering.  I was employed almost three years in this effort.  The processing was broken into “subsystems” in a dogmatic way with development software from “Pride-Logik”.   Curiously, the management then, although of retirement age now, could have been of help in developing the new system, due to experience.  Did political turf battles hinder this system?
The Washington Post reports Tuesday morning, in a story by Amy Goldstein and Juliet Eilperin, “HealthCare.gov makes frequent enrollment errors; up to one-third possibly affected; Some people may not have the coverage they expected”, link here.

The entire workflow is supposed include a daily “reconciliation” which sounds like a mainframe batch job that would send xml files for reports to all the various components (including the insurance companies) linked in the system for daily corrective action.