Article

2013-11-07 Home Front: Politix

Sebilius says healthcare website has hundreds of errors

Archived material is restricted to Rantburg regulars and members. If you need access email fred.pruitt=at=gmail.com with your nick to be added to the members list. There is no charge to join Rantburg as a member.

Posted by Frozen Al 2013-11-07 00:00|| || Front Page|| [11127 views ] Top

#1 Yup. Every Congress hag or Congress dupe, senator, Justice, or President that allowed it.

Posted by newc 2013-11-07 00:34|| 2013-11-07 00:34|| Front Page Top

#2 CGI Experience the Commitment.

33.98
+0.21 (0.62%)
Nov 6 - Close
NYSE real-time data - Disclaimer
Currency in USD

Range 33.66 - 34.39
52 week 22.51 - 37.41
Open 33.85
Vol / Avg. 0.00/206,572.00
Mkt cap 10.57B
P/E 90.53
Div/yield -
EPS 0.38
Shares 311.07M
Beta 0.83
Inst. own 54%

Posted by Besoeker 2013-11-07 07:31|| 2013-11-07 07:31|| Front Page Top

#3 Docs show website could only handle 1,100 users the day before launch.

Posted by Besoeker 2013-11-07 07:38|| 2013-11-07 07:38|| Front Page Top

#4 I doubt the website could have even handled 100 users. I believe very little this government has to say about itself.

Posted by Anguper Hupomosing9418 2013-11-07 09:03|| 2013-11-07 09:03|| Front Page Top

#5 Normally, during Quality Assurance tests during alpha, well before beta, a site like this will not have hundreds of errors. It will have thousands. Depending on how large the QA test team is.

After a few hundred bugs are fixed, those fixes open up more functionality on down the line of code that has not been tested yet until the first bugs were fixed, which adds hundreds more.

If the development team consists of more than one person, which usually a project of this size has many developers, copies of code must be controlled through source control programs such as Visual Source Safe, or a developer working with an old set of code for several days that does not include bug fixes rolled out since he got his copy, can overwrite the new bug fixes when he submits his old stuff.

Democrats don't know this. But common practice in the Evil private sector enterprise.

A Rantburg Exclusive Analysis and Commentary.

Posted by Guillibaldo McCoy1948 2013-11-07 10:24|| 2013-11-07 10:24|| Front Page Top

#6 My guess is that the specs were never nailed down until just before the go-live. Designed by multiple committees and rampart feature creep as things get added, and added, and added, oh and this well-connected interest group wants this little feature so you have to re-write whole sections of code.

So the development window shifts right radically toward the go-live date and QA and Release management gets squeezed tight. What was originally a few months of QA gets squeezed to a couple of weeks. QA and Source control gets very sloppy with little or no controls and soon they are simply throwing changes into production practically directly from development with little or no QA (we can catch up on that later... besides, it compiles and the prima-donna developers tested it right? We *must* make this date!)

Recipe for disaster.

Posted by CrazyFool 2013-11-07 10:44|| 2013-11-07 10:44|| Front Page Top

#7 this well-connected interest group wants this little feature so you have to re-write whole sections of code

Plus the issue of the White House wanting complete secrecy.

Then there's the reluctance to inform the Oval Office that things weren't going well.

Shades of the media's explanation for Saddam's chem/bio warfare program, ironically.

Posted by Pappy 2013-11-07 10:56|| 2013-11-07 10:56|| Front Page Top

#8 I'm sure it has a lot of runs and drips too.

Posted by Thing From Snowy Mountain 2013-11-07 11:30|| 2013-11-07 11:30|| Front Page Top

#9 We *must* make this date!

Schedules are the natural enemy of software projects.

Posted by SteveS 2013-11-07 12:09|| 2013-11-07 12:09|| Front Page Top

#10 My guess is that the specs were never nailed down until just before the go-live.

Uhhhhhh, yeahhhhhh. Like I said before, imagine the likes of Hilary Clinton, Kathleen Sebelius and Nancy Pelosi putting a software specification together. And just for kicks, throw in Harry Reid and Moochelle. C'mon, it'll be fun!

Posted by Ebbang Uluque6305 2013-11-07 12:15|| 2013-11-07 12:15|| Front Page Top

#11 Administration officials say they are committed to having the website fixed by the end of November, but the timeline leaves little room for error given the list of problems.

"It's a pretty aggressive schedule," Mrs. Sebelius told the Senate Finance Committee at a hearing Wednesday.

"Aggressive schedule" is management speak for "the programmers are on a death march". Let's put it this way, they won't be seeing much of their families for the next few weeks.

Posted by Ebbang Uluque6305 2013-11-07 12:20|| 2013-11-07 12:20|| Front Page Top

#12 ...it's getting better (link): Pro-Obamacare team trains reporters on covering Obamacare website problems

Posted by Uncle Phester 2013-11-07 12:46|| 2013-11-07 12:46|| Front Page Top

#13 But remember, "It's the Law of the Land".

So was Prohibition.

Posted by Bobby 2013-11-07 12:47|| 2013-11-07 12:47|| Front Page Top

#14 Pro-Obamacare team trains reporters on covering Obamacare website problems

The vital quote from that article:

"What is surprising is that an organization claiming to represent professional journalists would endorse 'training' delivered by advocates for the program they are covering, which would violate SABEW's code of ethics. That code encourages journalists to 'avoid any practice that might compromise or appear to compromise objectivity or fairness.'"

So we'll see "hundreds of errors" become "dozens of minor problems," to then become "a few feature enhancement issues?"

Posted by Pappy 2013-11-07 13:27|| 2013-11-07 13:27|| Front Page Top

#15 Any sting those journalists feel will be their pride...if they have any. Maybe the trainers can buy them some drinks afterwards. It might help.

Posted by Ebbang Uluque6305 2013-11-07 13:48|| 2013-11-07 13:48|| Front Page Top

#16 *sigh*

The "hundreds of serious problems" was a hint to give up.

Posted by Ptah 2013-11-07 14:03|| 2013-11-07 14:03|| Front Page Top

#17 #15 Any sting those journalists feel will be their pride...if they have any. Maybe the trainers can buy them some drinks afterwards. It might help.

Bill Clinton: "put some ice on that"

Posted by Frank G 2013-11-07 14:33|| 2013-11-07 14:33|| Front Page Top

#18 You can do all the BS propaganda actions, but the thing that you cannot cover up the fact that people are getting notices that their insurance is cancelled.

Some people will submit and buy health insurance that does not pay anything and is more of a tax. Some people will not submit.

The bottom line is mean olde Mr. Arithmetic. There is not enough money to pay for those that don't contribute by those that do.

The system is dead. It just does not know it yet.

Posted by Alaska Paul 2013-11-07 14:58|| 2013-11-07 14:58|| Front Page Top

#19 Sebilius says healthcare website has hundreds of errors.

Same could be said for the stones in Hadrian's Wall, from a supine view.

Posted by Besoeker 2013-11-07 16:24|| 2013-11-07 16:24|| Front Page Top

#20 Yow... Frank G. has a long memory.

Posted by Shipman 2013-11-07 18:15|| 2013-11-07 18:15|| Front Page Top

#21 In other words AP: 2 plus 2 does not and never will equal 5. Or 3!

Posted by CrazyFool 2013-11-07 18:40|| 2013-11-07 18:40|| Front Page Top

#22 The thing is, in software there are many kinds of errors. There can be simple formatting errors - you expected dates to be printed out as month/day/year, and they came out as day/month/year. These sorts of errors should show up in initial testing, and should be fairly easy to fix.

Then there are simple coding errors, where the programmer meant to add two numbers together, but instead multiplied them. These errors might not be easy to find, but they are usually easy to fix - just correct the code.

Then there are design errors, where you tell the programmer to do something, but s/he was supposed to do something different. These are much harder to find, and can be a real bitch to fix.

Next you have the connection errors, where you need to get information from another system, say the IRS. You expect the IRS to send you the data in XML, but they send it in their own private binary code. These should show up when you do your connection tests - if you do them. Fixing these errors could take a long time - somebody needs to decree a common format, and then everybody has to code to it.

Add to these the numerous other places where things go wrong - you call a subroutine, and send the parameters in a certain order, but the subroutine expects them in a different order. Again, these sorts of issues should be settled in a design conference - if you have them. Sometimes you need the system integrator to establish the protocols between systems. If no one with experience is in charge, you get chaos. If you don't test, you find out only when the system goes live.

So there are hundreds of errors - but that doesn't address what kind of errors they are, where they are, how critical they are, how hard they are going to be to fix. That is the real challenge.

Posted by Rambler in Virginia 2013-11-07 22:02|| 2013-11-07 22:02|| Front Page Top

23:29 Herb Gloluger9960
23:24 Herb Gloluger9960
23:16 Herb Gloluger9960
23:07 SteveS
23:06 Herb Gloluger9960
22:51 Herb Gloluger9960
22:40 Iblis
22:02 Rambler in Virginia
21:48 abu do you love
21:37 USN, Ret.
21:37 Alaska Paul
21:16 tipover
21:06 Barbara
21:02 Barbara
21:00 gorb
20:59 Barbara
20:53 Glenmore
20:33 gorb
20:00 Thing From Snowy Mountain
19:49 rammer
18:40 CrazyFool
18:34 Besoeker
18:27 Chantry
18:15 Shipman