If you can read this, either the style sheet didn't load or you have an older browser that doesn't support style sheets. Try clearing your browser cache and refreshing the page.

(Politico)   Healthcare.gov site should have used Wordpress to make their blog suck   (politico.com) divider line 29
    More: Fail, WordPress, White House, 14th state  
•       •       •

1230 clicks; posted to Geek » on 25 Oct 2013 at 1:40 PM (1 year ago)   |  Favorite    |   share:  Share on Twitter share via Email Share on Facebook   more»



29 Comments   (+0 »)
   
View Voting Results: Smartest and Funniest
 
2013-10-25 01:05:37 PM  
On a press call today, the White House said the site will be functional by the end of November. I'd like to think they wouldn't say that unless they are EXTREMELY confident of that.
 
2013-10-25 01:44:52 PM  
They probably didn't use WordPress or any other of the tools listed because security vulnerabilities are found and exploited in them all of the time.
 
2013-10-25 01:46:11 PM  
Is the fail for HealthCare.gov or for the article? If the latter than subby is an idiot, as the article is exactly right. Using a tested and proven back-end would have made the whole process cheaper and less error prone. If you think WordPress is just a blog platform, you're a moron.
 
2013-10-25 01:52:52 PM  

DamnYankees: On a press call today, the White House said the site will be functional by the end of November. I'd like to think they wouldn't say that unless they are EXTREMELY confident of that.


Given that the end of November is still a month away, and how much of a train wreck the main site is, I'd remain pretty skeptical.  Unless they fired everyone and brought in actual programmers.

MindChild: They probably didn't use WordPress or any other of the tools listed because security vulnerabilities are found and exploited in them all of the time.


IIRC, they are pretty steady with security updates.  For the amount of money spent on the website, you could have a full time wordpress dev just sit there and wait for patches.

entropic_existence: Is the fail for HealthCare.gov or for the article? If the latter than subby is an idiot, as the article is exactly right. Using a tested and proven back-end would have made the whole process cheaper and less error prone. If you think WordPress is just a blog platform, you're a moron.


I don't think it would have made any difference.  You can give a shiatty driver a Lamborghini or a Pinto.  Both will end up upside down in a ditch.
 
2013-10-25 01:56:31 PM  

fang06554: Given that the end of November is still a month away, and how much of a train wreck the main site is, I'd remain pretty skeptical.  Unless they fired everyone and brought in actual programmers.


There's 2 things about this. First is that the main site has gotten much better. People already forget, but in the first week of October is was basically impossible to even create an account. Now that goes through smoothly. In their call today they said the number went from 30% at the beginning to 90% now.

Secondly, I heard an interesting theory that the front-end problems are actually sort of artificial. The idea is that the real issue is on the back-end, in communicating enrollee information to insurers. The HHS people know this needs to be fixed first. At the moment, with a slow trickle of enrollees, any errors in processing that data can be fixed manually by the insurers. But once the flood begins, there would be no way to fix transmission errors. So the theory is that in reality they know how to fix the front end, and it would be relatively simple, but they aren't doing it yet because they know that doing so would open the floodgates before the back-end was fixed. So its possible that they have the fix for the front-end done and totally ready to go, they just can't pull the trigger yet.
 
2013-10-25 02:02:33 PM  

entropic_existence: Is the fail for HealthCare.gov or for the article? If the latter than subby is an idiot, as the article is exactly right. Using a tested and proven back-end would have made the whole process cheaper and less error prone. If you think WordPress is just a blog platform, you're a moron.


Because a trendy CMS would've solved problems with dependencies on a ton of disparate backend services? Or because a framework tied to PHP / MySQL scales crazy well?

/not subby
 
2013-10-25 02:05:11 PM  

DamnYankees: fang06554: Given that the end of November is still a month away, and how much of a train wreck the main site is, I'd remain pretty skeptical.  Unless they fired everyone and brought in actual programmers.

There's 2 things about this. First is that the main site has gotten much better. People already forget, but in the first week of October is was basically impossible to even create an account. Now that goes through smoothly. In their call today they said the number went from 30% at the beginning to 90% now.

Secondly, I heard an interesting theory that the front-end problems are actually sort of artificial. The idea is that the real issue is on the back-end, in communicating enrollee information to insurers. The HHS people know this needs to be fixed first. At the moment, with a slow trickle of enrollees, any errors in processing that data can be fixed manually by the insurers. But once the flood begins, there would be no way to fix transmission errors. So the theory is that in reality they know how to fix the front end, and it would be relatively simple, but they aren't doing it yet because they know that doing so would open the floodgates before the back-end was fixed. So its possible that they have the fix for the front-end done and totally ready to go, they just can't pull the trigger yet.


While there were definitely a  lot of coding bugs, there were also very fundamental design problems.  As an example, if you are filling out an application, and have to stop for some reason, you can save it, and continue later.  Except, you don't actually continue later, you start from the very beginning and go through the entire application all over again.  In the short term, not really  that big of a deal.  It is pretty indicative of the design structure as a whole though.

Odds are, that once the flood gates open, and suddenly people are able to submit applications, then another whole new set of bugs will appear.  What we pretty much have right now is a full scale alpha release, getting ready to go into beta.  I guess time will tell though, maybe I'm just a little bit too jaded with the whole evolution.

/Has never been able to submit an application
 
2013-10-25 02:07:12 PM  

fang06554: While there were definitely a  lot of coding bugs, there were also very fundamental design problems.  As an example, if you are filling out an application, and have to stop for some reason, you can save it, and continue later.  Except, you don't actually continue later, you start from the very beginning and go through the entire application all over again.  In the short term, not really  that big of a deal.  It is pretty indicative of the design structure as a whole though.


Yeah that's pretty infuriating. I had to do that a few times - what the fark is the point of all the "save and continue" buttons if NOTHING IS SAVED?

Its very annoying, but its not crippling.
 
2013-10-25 02:15:10 PM  

MindChild: They probably didn't use WordPress or any other of the tools listed because security vulnerabilities are found and exploited in them all of the time.


Drupal already runs whitehouse.gov

/just sayin'
 
2013-10-25 02:29:16 PM  
Honest question does anyone know if there is any sort of contingency plan in the event they don't get enough people enrolled? I've heard Republicans say it will collapse in on itself and Democrats insist it wont happen but have yet to hear any serious discussion if that did happen what would be done.
 
2013-10-25 02:45:57 PM  
this whole thing is a horrible, horrible mess.  And it will only get worse ...
 
2013-10-25 02:49:47 PM  

fang06554: I don't think it would have made any difference.  You can give a shiatty driver a Lamborghini or a Pinto.  Both will end up upside down in a ditch.


Oh that is definitely true. It just seemed that subby was suggesting the fail for even the thought of usign WordPress, or something similar. Given many other state-level exchanges use it for their backends it seems like labelling that suggestion a fail is, well, a fail.

zyrian: Because a trendy CMS would've solved problems with dependencies on a ton of disparate backend services? Or because a framework tied to PHP / MySQL scales crazy well?


Didn't say anything was perfect. I said labelling it as a "fail" suggestion is moronic. I also wouldn't call WordPress or many other CMS "trendy" they've been around for awhile and many of them get quite heavy day-to-day use in very high level production environments. And that speaks to scalability as well. Yes, having to coordinate disparate backend services is a nightmare but it isn't one decent developers can't address or one that is terribly unique. People have to deal with issues like that all of the time. Using an out of the box, maintained, and robust solution to at least get you part way there tends to work much much better than building your own unique backend from scratch. Where you just contribute to the problem of disparate services needing to talk to each other.
 
2013-10-25 03:06:27 PM  
This is just an advertisement for the crappy builder WordPress.   WordPress was designed for bloggers no gigantic databases, even if you cram a database into WordPress, if the database doesn't work, the website wont work.  The problems with the website seem to be with the database and interacting with it.  WordPress would have done NOTHING to change that.
 
2013-10-25 03:06:37 PM  

entropic_existence: fang06554: I don't think it would have made any difference.  You can give a shiatty driver a Lamborghini or a Pinto.  Both will end up upside down in a ditch.

Oh that is definitely true. It just seemed that subby was suggesting the fail for even the thought of usign WordPress, or something similar. Given many other state-level exchanges use it for their backends it seems like labelling that suggestion a fail is, well, a fail.

zyrian: Because a trendy CMS would've solved problems with dependencies on a ton of disparate backend services? Or because a framework tied to PHP / MySQL scales crazy well?

Didn't say anything was perfect. I said labelling it as a "fail" suggestion is moronic. I also wouldn't call WordPress or many other CMS "trendy" they've been around for awhile and many of them get quite heavy day-to-day use in very high level production environments. And that speaks to scalability as well. Yes, having to coordinate disparate backend services is a nightmare but it isn't one decent developers can't address or one that is terribly unique. People have to deal with issues like that all of the time. Using an out of the box, maintained, and robust solution to at least get you part way there tends to work much much better than building your own unique backend from scratch. Where you just contribute to the problem of disparate services needing to talk to each other.


I don't think using WordPress, or any of the other CMS mentioned, would be a bad choice, but I hardly think it would address the underlying problem. WordPress, without significant modifications, would only address the front end. I suspect the real performance issues arise from back end problems. AFAIK, the only "backend" that WordPress has is sending web form data to MySQL. This is basic web application programming, I don't think that was the problem with healthcare.gov.

Getting a web application to talk to a disparate bunch of systems, some likely arcane, in an efficient manner sounds like the real challenge here. I don't see WordPress solving that issue.
 
2013-10-25 03:18:02 PM  
WordPress is a content management system, not a goddamn panacea.
 
2013-10-25 03:44:22 PM  

DamnYankees: On a press call today, the White House said the site will be functional by the end of November. I'd like to think they wouldn't say that unless they are EXTREMELY confident of that.


How hard could it possibly be to find a website designer who can create a site that charges money?
 
2013-10-25 03:45:07 PM  
DamnYankees:  I'd like to think they wouldn't say that unless they are EXTREMELY confident of that.

Have you not been paying attention for the last few years?
 
2013-10-25 03:48:44 PM  
Hell, they could have used Squarespace and it would have been an improvement.
 
2013-10-25 04:33:09 PM  
What are they using to power whitehouse.com?
 
2013-10-25 04:40:15 PM  
Anyone else getting tired of those with no software or web development knowledge asking what they think are revealing questions? You're wasting everyone's time.
 
2013-10-25 04:43:10 PM  

TigerzDad: What are they using to power whitehouse.com?


The blood of white Christian babies.
 
2013-10-25 06:41:34 PM  
It's government.
The development contract was either awarded to the lowest bidder, or whomever was highest in the power chain with a relative who "knows computers".

The entire site is rife with code that would make even a first year programming student embarrassed to be associated with the profession.  "Design-by-committee" has NEVER been a good idea because of consistent issues like this, and this project absolutely stinks of it.

I wonder how much Amazon invested in-total on their site to get it to the level it's at now.  Or eBay.  Or Google.  Somehow I doubt the total would be anywhere near 640 million dollars.  Somewhere out there, there are some dudes who dropped out of college after a year with powerful relatives contemplating the purchase of their second Lamborghini.  :(
 
2013-10-25 07:07:52 PM  

MindChild: They probably didn't use WordPress or any other of the tools listed because security vulnerabilities are found and exploited in them all of the time.


Most security problems come from the plugins and misconfiguration rather than Wordpress itself.
 
2013-10-25 07:45:37 PM  

ItachiNai: I wonder how much Amazon invested in-total on their site to get it to the level it's at now.  Or eBay.  Or Google.  Somehow I doubt the total would be anywhere near 640 million dollars.


Oh, well if you doubt it then I guess

(Actually, Amazon spent over 4.5 billion dollars on "technology and content" in 2012 alone.)
 
2013-10-25 08:11:39 PM  

1000 Ways to Dye: MindChild: They probably didn't use WordPress or any other of the tools listed because security vulnerabilities are found and exploited in them all of the time.

Most security problems come from the plugins and misconfiguration rather than Wordpress itself.


Yep.  From my experience setting up a lot of wordpress servers, the problems are plugins written by people who know just enough coding to be dangerous.  Their sql queries are unfiltered so it's ripe for simple exploit.  They also do queries on unindexed columns so they are slow as hell when the database grows.

A vanilla wordpress install is relatively secure.  It's all the shiat people add that get them pwned.


imgs.xkcd.com
 
2013-10-26 08:22:04 AM  

manbart: I don't think using WordPress, or any of the other CMS mentioned, would be a bad choice, but I hardly think it would address the underlying problem. WordPress, without significant modifications, would only address the front end. I suspect the real performance issues arise from back end problems. AFAIK, the only "backend" that WordPress has is sending web form data to MySQL. This is basic web application programming, I don't think that was the problem with healthcare.gov.

Getting a web application to talk to a disparate bunch of systems, some likely arcane, in an efficient manner sounds like the real challenge here. I don't see WordPress solving that issue.


True. I guess I'm also just thinking that it looks like the developers for both the front and back ends decided to do everything from scratch instead of using off the shelf components where appropriate. That's never a good sign for competence as a web development company IMO.
 
2013-10-26 10:37:41 AM  

entropic_existence: True. I guess I'm also just thinking that it looks like the developers for both the front and back ends decided to do everything from scratch instead of using off the shelf components where appropriate. That's never a good sign for competence as a web development company IMO.


The trouble is, there aren't off-the-shelf components for "record quotes and connect a healthcare exchange with XYZ Insurance". You want an ecommerce system, a blog, somewhere to put all your user guides? Great - I'll install something off-the-shelf.

I build business systems that integrate with other systems, so let me explain: the front end is the smallest, easiest bit of the job. On the systems I've built, we've used a load of open source stuff like jQuery and Knockout.js to do the job. On the back end, we use a bunch of tools to do things like WSDL discovery, or to generate DBs from entity models, and I'll occassionally pinch classes that do stuff for me, or use some classes provided to interface to something like an SMS gateway, but most of it has to be built from scratch. That's what most programmers are doing - building stuff that there isn't something off-the-shelf for, because otherwise the people in charge would just get something off-the-shelf.

And frankly, anyone suggesting WP is a moron. There's almost nothing in there that fits what a healthcare system needs. You'd have to rewrite so much of it, you would be better starting with a clean sheet of paper.
 
2013-10-26 11:55:09 AM  

OgreMagi: ep. From my experience setting up a lot of wordpress servers, the problems are plugins written by people who know just enough coding to be dangerous. Their sql queries are unfiltered so it's ripe for simple exploit. They also do queries on unindexed columns so they are slow as hell when the database grows.

A vanilla wordpress install is relatively secure. It's all the shiat people add that get them pwned.


Any particular plugins you'd care to point out? I'm genuinely curious, as I run a few blogs (which, no doubt, suck) for myself and others.

I generally stay away from public-facing plugins, with only a few exceptions: Yubikey/Google Authenticator for securing the login system (seriously, GA support should be built-in), a caching plugin for busier sites, something to make XML sitemaps to make things easy for search engines, and one to add Piwik (essentially a self-hosted Google Analytics system) code to public pages in an easily-maintainable way.

I'm quite excited that the 3.7 WordPress release will automatically check for and install security updates without user intervention (it can be disabled, but it defaults to on). Hopefully that'll help keep a lot of sites from getting compromised.
 
2013-10-26 12:11:54 PM  

MindChild: They probably didn't use WordPress or any other of the tools listed because security vulnerabilities are found and exploited in them all of the time.


www.reactimg.com
 
Displayed 29 of 29 comments

View Voting Results: Smartest and Funniest


This thread is closed to new comments.

Continue Farking
Submit a Link »






Report