Skip

To Apply For FEMA Aid Online, Katrina Survivors Will Need IE 6
September 26, 2005 6:11 AM   Subscribe


 
Aha! Natural selection has not caused us Mac users to go extinct, so now they are implementing intelligent design to finish us off...
posted by fairmettle at 6:17 AM on September 26, 2005


If I could be bothered to write a list of 'Things I Couldn't Give a Fuck About' I promise I would put this story right at the top.
Which is the opposite of where it would be if I was writing a list of priorities related to the Katrina clean-up.
posted by biffa at 6:18 AM on September 26, 2005


So a search of Metafilter using the "Katrina + 'Internet Explorer'" yields this.
But, you know, bring it up again if you want.
posted by oddman at 6:22 AM on September 26, 2005


Food for thought: Someone had to work to make the FEMA form IE-only.

If they hadn't done that work, the form would have worked just fine in any browser that supports POST. Which is, you know, any browser.

Since I'm not interested in filing a bunch of fradulent claims, I'm going to leave my conjecture at the reasons at guessing that the developers couldn't figure out how to prevent multiple submissions without doing something funky and related to idiosyncracies in IE6's http headers.

On preview: WHY I GIVE A FUCK

This is an example of incompetence. Someone was either too lazy to figure out how to do something right, instead of as a half-assed kludge, or someone was engaging in a dogmatic crusade.
posted by lodurr at 6:26 AM on September 26, 2005


How many people that are applying for this relief are applying by computer? I was under the impression that these were people applying for disaster relief, meaning they just lost their house. I doubt that very many of these folks have electricity much less an internet connection and functional computer.
posted by Pollomacho at 6:33 AM on September 26, 2005


I doubt that very many of these folks have electricity much less an internet connection and functional computer.

It's my understanding that a lot of them are going to makeshift application centers that are set up at public libraries and other areas specifically for this purpose. There was a big call for computers from the State Library a few weeks ago for this purpose.
posted by jessamyn at 6:36 AM on September 26, 2005


That makes sense Jessamyn, but couldn't they just set up the folks that need to apply at pc's?
posted by Pollomacho at 6:46 AM on September 26, 2005


Damn. I work in marketing, and my mission is to make sure everything we do on the web is viewable by as many audiences as possible -- to move past 80/20 and instead say "let's get as close to 100 as possible."

Shouldn't someone have that job at governmental agencies too?
posted by VulcanMike at 6:47 AM on September 26, 2005


Yeah, really, why should you give a fuck? It's only your tax dollars hard at work, giving corporate welfare to Microsoft.

And if you think this is a PC/Mac thing, just wait until you can't use your precious Firefox to ask for help from your government to recover from an emergency, let alone download tax forms and the like.
posted by Rothko at 8:17 AM on September 26, 2005


FEMA hates open source.
posted by dial-tone at 8:20 AM on September 26, 2005


I applied online from a friend's house. Didn't work. Then from my mother's house. Didn't work. The FEMA Web page has not worked, in fact, on any computer I have attempted.

You can get through if you call. It takes between an hour and a half to three hours, if you can get through at all.

Most people are applying for an emergency relief of $2,000. I applied three weeks ago and have yet to see dime one. Apparently, "emergency relief" for FEMA means "whenever the hell we get around to it."
posted by maxsparber at 8:24 AM on September 26, 2005


It is another sign of the cluelessness that exists in the government agencies that are allegedly there to help you. This would be a (very late) double-post newsfilter item, except this is not news to anybody.
posted by spock at 8:33 AM on September 26, 2005


I have had terrible luck with federal websites, most of which are IE only. If they can't get those right for one browser, how the hell can you expect the feds to get 'em right for others?
posted by mischief at 8:35 AM on September 26, 2005


FEMA hates doesn't care about open source.

Dare I point out that a lot of government agencies worked much better in January 2001 than they did in 1993? Or, for that matter, than they do now...
posted by lodurr at 8:36 AM on September 26, 2005


My guess is simply that they never tested on any other browsers, or maybe use ActiveX.

Still, weak.
posted by delmoi at 8:39 AM on September 26, 2005




So that's what they mean by standards compliance?
posted by Joeforking at 9:10 AM on September 26, 2005


The Web Standards Project notes that this restriction doesn't square with FEMA's accessibility statement.

Of FEMA and IE: Why are bloggers obsessed with the trivial?
posted by kirkaracha at 9:27 AM on September 26, 2005


Yeah, really, why should you give a fuck? It's only your tax dollars hard at work, giving corporate welfare to Microsoft.

And if you think this is a PC/Mac thing, just wait until you can't use your precious Firefox to ask for help from your government to recover from an emergency, let alone download tax forms and the like.


Oh for fuck's sake, that's nonsense. This isn't corporate welfare in even the most basic way. How exactly is Microsoft profiting from this in any way? Are people running out to buy PC's with Windows on them because they can't apply for FEMA aid on their Mac? That's such a stretch.

IE accounts for something like 80% of the browser market, and Windows accounts for around 90% of the desktop market. The Government, in an attempt to save you tax money, is following the current "de-facto" standard for your average person who uses the internet. You guys are the exception, not the rule.

I can't believe anyone has the energy to get righteously upset over piddling shit like this.
posted by SweetJesus at 9:35 AM on September 26, 2005


"I can't believe anyone has the energy to get righteously upset over piddling shit like this."

Sounds suspiciously like a tagline to me.
posted by mischief at 9:40 AM on September 26, 2005


Metafilter: I can't believe anyone has the energy to get righteously upset over piddling shit like this
posted by Who_Am_I at 9:50 AM on September 26, 2005


The fact that it's such a trivial matter is exactly why it's so sad that FEMA can somehow manage to fuck it up. It's like if someone goes to change a lightbulb, and they accidentally replace the old burnt-out bulb with itself instead of using a new one. No real harm done, but it demonstrates a truly awesome level of stupidity.
posted by sfenders at 9:50 AM on September 26, 2005


IE accounts for something like 80% of the browser market, and Windows accounts for around 90% of the desktop market. The Government, in an attempt to save you tax money, is following the current "de-facto" standard for your average person who uses the internet. You guys are the exception, not the rule.
posted by SweetJesus at 12:35 PM EST on September 26 [!]


Designing for one company's web browser costs more than to design to W3C and federal accessibility standards. You pretty much have to buy Microsoft's Visual Studio package at hundreds of dollars per license in order to develop websites that are broken on any other browser but IE. Complete waste of taxpayer's money.

And, yes, it is indeed corporate welfare to Microsoft if you're pushing people to buy Windows machines in order to access taxpayer-funded government resources.
posted by Rothko at 9:50 AM on September 26, 2005


Designing for one company's web browser costs more than to design to W3C and federal accessibility standards. You pretty much have to buy Microsoft's Visual Studio package at hundreds of dollars per license in order to develop websites that are broken on any other browser but IE. Complete waste of taxpayer's money.

That's just stupid. I worked in web design for a few years, and even working within software frameworks that are deigned to be cross-browser compatible, it's a pain in the ass. It involves a lot of hand-tweeking. These guys are using FrontPage to produce web sites in the quickest amount of time possible. They're not going to hand-code things so it's works in Safari on OS 9.3, or whatever other pony you want.

I won't respond to the whole corporate welfare thing, because it's just so utterly worthless. Under your definition any standardization amounts to corporate welfare. We all know you hate Windows and Microsoft, Alex, but this is silly.
posted by SweetJesus at 10:09 AM on September 26, 2005


Under your definition any standardization amounts to corporate welfare.

You don't need to let corporations dictate your feature set when there are publicly-accepted standards already in place.

We all know you hate Windows and Microsoft, Alex, but this is silly.

We all know FEMA is incompetent. What's silly and stupid is giving Microsoft a free ride at the taxpayer's expense. The money that could be saved could, say, help out people in an emergency, to give one example.
posted by Rothko at 10:24 AM on September 26, 2005


You don't need to let corporations dictate your feature set when there are publicly-accepted standards already in place.

Hahaha... I'm sure Microsoft is secretly behind this.

We all know FEMA is incompetent. What's silly and stupid is giving Microsoft a free ride at the taxpayer's expense. The money that could be saved could, say, help out people in an emergency, to give one example.

Explain to me how spending extra time and money making sure Government web sites work with every browser saves any time or money.
posted by SweetJesus at 10:30 AM on September 26, 2005


Apparently the web interface was built as an internal application, and then later opened to the public. I believe this. The "help prompts" are formatted as a script, which a FEMA employee might use to assist a phone caller.

Here is more information on that: http://arstechnica.com/news.ars/post/20050906-5278.html

Please remember this example the next time some moron says your intranet application can be IE-only.

By the way, although the blogosphere has been drooling over this incident, the worse incident is the quality of the application. I was helping residents in the Austin shelter file their FEMA app and open an on-line account, and my success rate was probably 20%. I never saw a single piece of software with so many distinct failure modes.
posted by chipr at 10:32 AM on September 26, 2005


You would think the government would be a little more thoughtful on making access/viewing available to everyone. Isn't that its job...making sure everyone is taken care of? Todd Lokken.
posted by toddlok at 10:40 AM on September 26, 2005


Explain to me how spending extra time and money making sure Government web sites work with every browser saves any time or money.
posted by SweetJesus at 1:30 PM EST on September 26 [!]


Explain to me why you need to spend extra time and money coding to community standards that apply to all standards-compliant browsers, instead of Microsoft's "standard" that applies to one non-compliant browser.
posted by Rothko at 10:41 AM on September 26, 2005


Which is to say, pretty much the main reason testing is done is to make sites designed for IE from the start work on other browsers. Sites designed from the start to community standards work on all browsers with much less extra testing required.
posted by Rothko at 10:43 AM on September 26, 2005


Explain to me why you need to spend extra time and money coding to community standards that apply to all standards-compliant browsers, instead of Microsoft's "standard" that applies to one non-compliant browser.

Simple. It's more cost-effective to have some IT G2 rapidly prototype a simple web application using FrontPage than it is to hire someone to write standards compliant code, by hand. It also helps that 90% of the market uses IE for web browsing.

It's logic, not conspiracy. But then again, you still haven't answered my question.
posted by SweetJesus at 10:45 AM on September 26, 2005


say the following: sofa kingdom

now say this: "this is sofa kingdom"

can we move on now?
posted by chibikeandy at 10:47 AM on September 26, 2005


Explain to me why you need to spend extra time and money coding to community standards that apply to all standards-compliant browsers, instead of Microsoft's "standard" that applies to one non-compliant browser.

I understand your argument, Alex, but this wasn't some well thought-out, long-term project; this was a short job to fix a need. Perhaps their in-house project designers and programmers are old-school, pre-w3c developers...as such, developping outside of their specialty would've taken more time and more money.

The reality is, as much as you and I have heard the preaching of standards compliance, can you really blame people who are making the decisions who they hire as developers who have not heard it because they just don't get it? If there's one criticism of the standards compliance movement that I have, it doesn't do a very good job of explaining to the common user why its message is important. It's missing a lot of people by the design of its marketing.

In any discipline I can think of, the most cutting edge and talented people aren't working for the government. There are levels of management making decisions on things they don't know about, and unfortunately the majority of programmers make lousy managers.

You don't need to let corporations dictate your feature set when there are publicly-accepted standards already in place.

Publicly-accepted usually means a majority; can you honestly say that a majority of americans know and/or support these standards?
posted by dflemingdotorg at 11:01 AM on September 26, 2005


It's logic, not conspiracy. But then again, you still haven't answered my question.
posted by SweetJesus at 1:45 PM EST on September 26 [!]


Your question was politely answered by indicating your assumptions are demonstrably incorrect. It does not cost more to code to larger standards bodies, and in fact costs less when done properly. Further, FEMA's application is preventing a non-insignificant number of people from getting assistance, which is a significant cost in itself. You don't like these answers, however.

As an aside, static pages are written (poorly) with FrontPage, and not so much web applications, which are usually written and served up with the aid of other software packages. Depending on the audience you're writing for (just IE, or everyone) you pick different software packages. I advise you to learn about W3C and pre-existing (!) federal accessibility standards.
posted by Rothko at 11:02 AM on September 26, 2005


I should revise:

Publicly-accepted usually means a majority; can you honestly say that a majority of americans know and/or support these standards?

Should read:

Publicly-accepted usually means a majority; can you honestly say that a majority of american programmers know and/or support these standards? If so, I'd love to see a study/article that says so.
posted by dflemingdotorg at 11:05 AM on September 26, 2005


I advise you to learn about W3C and pre-existing (!) federal accessibility standards.

Thanks for the advice, Capt. Condescension.

Fuck it, this is silly. You're impossible to have a discussion with about anything. The only evidence you offer is what you pull out of your ass and declare as Gospel. I have better things to do.
posted by SweetJesus at 11:08 AM on September 26, 2005


This really is silly. Your profile says you're an engineer. Let's say you're a civil engineer. Do you build your bridge from the same ISO 9001 vendor-manufactured parts as your colleagues would use, or do you invent your own parts out of thin air, spend a lot of money testing them, and then cross your fingers that your bridge won't collapse and kill someone?
posted by Rothko at 11:15 AM on September 26, 2005


This really is silly. Your profile says you're an engineer. Let's say you're a civil engineer. Do you build your bridge from the same ISO 9001 vendor-manufactured parts as your colleagues would use, or do you invent your own parts out of thin air, spend a lot of money testing them, and then cross your fingers that your bridge won't collapse and kill someone?

You know what's silly? That analogy. They know their program works with IE, they're not crossing their fingers and there's no chance of collapse within the confines of what they've designed it for.
posted by dflemingdotorg at 11:22 AM on September 26, 2005


They know their program works with IE, they're not crossing their fingers and there's no chance of collapse within the confines of what they've designed it for.

Well, apparently they've decided to airlift some folks over their hypothetically shoddy bridge in lieu of better ideas.
posted by Rothko at 11:26 AM on September 26, 2005


They know their program works with IE, they're not crossing their fingers and there's no chance of collapse within the confines of what they've designed it for.

Um: "I was helping residents in the Austin shelter file their FEMA app and open an on-line account, and my success rate was probably 20%. I never saw a single piece of software with so many distinct failure modes."
posted by Optimus Chyme at 11:27 AM on September 26, 2005


Well, apparently they've decided to airlift some folks over their hypothetically shoddy bridge in lieu of better ideas.

No, it's more like shuttling walkers over a bridge designed only for cars, should you really be looking for a proper analogy. It works for the majority and they provide an, perhaps inconvenient, alternative to those who fit outside of the majority.
posted by dflemingdotorg at 11:29 AM on September 26, 2005


I wonder how much it costs to set up, maintain and man a toll-free hotline, as well as hire web developers to shoehorn their shoddy IE code into something workable for everyone? Imagine what the costs would have been with better planning — imagine how much better the service to all taxpayers in need. It isn't like FEMA: The Organization was invented post-Katrina and this particular application needed to be a fly-by-night cook-up job.
posted by Rothko at 11:31 AM on September 26, 2005


I'm a simulation engineer, and I work with standards all the time for simulation interoperability.

W3C holds no power, and has no teeth. Until, if ever, it does, the market will dictate what is cost-effective to support. Most people use IE on a Windows PC, so most web pages will be built with that in mind. Most people don't care that you have some Quixotic problem with Microsoft, and wont care that you don't like to use IE.
posted by SweetJesus at 11:33 AM on September 26, 2005


"I was helping residents in the Austin shelter file their FEMA app and open an on-line account, and my success rate was probably 20%. I never saw a single piece of software with so many distinct failure modes."

This might have to do a lot with this;

Due to the heavy load on our online registration application some of you are receiving an error message when you try to apply. We apologize for this and ask you to call the FEMA Technical Helpdesk at 1-800-745-0243 for further assistance.

I don't know for sure but I think that dealing with heavy load in programming would be the type of thing one would do if they had more time; this, obviously, was a rush job to fix a need they had immediately.
posted by dflemingdotorg at 11:36 AM on September 26, 2005


Is government policy now dictated by market share? Let's scrap the ADA while we're putting shareholder value ahead of public service. Those wheelchair ramps cost dollars.
posted by Rothko at 11:37 AM on September 26, 2005


Is government policy now dictated by market share? Let's scrap the ADA while we're putting shareholder value ahead of public service. Those wheelchair ramps cost dollars.

If you can't tell the difference between an inconvenience and a necessity, we're screwed.

Lets bring Hitler into this right now, and get it over with.
posted by SweetJesus at 11:40 AM on September 26, 2005


If you can't tell the difference between an inconvenience and a necessity, we're screwed.

I'd love to see you explain to taxpayers trying to apply for aid from FEMA — so that they have a place to stay for the night and clothes on their back — your nuanced version of the difference between inconvenience and necessity. Thanks for your time.
posted by Rothko at 11:43 AM on September 26, 2005


Whatever.

Your logic chain is more twisted than a Gordian Knot. I'm upset at myself for getting pulled back into silly circle-jerk.
posted by SweetJesus at 11:50 AM on September 26, 2005


Thanks for your time.

I'd love to see you explain anything to these people with your condescending, holier than thou attitude. Do you ever wonder why your message doesn't get through to people, that it could entirely have to do with the way you present it?

Honestly, without the snark, it would be a lot easier to actually take you seriously. Have you ever thought about that?
posted by dflemingdotorg at 11:51 AM on September 26, 2005


sfenders: and they accidentally replace the old burnt-out bulb with itself instead of using a new one. No real harm done, but it demonstrates a truly awesome level of stupidity.

It's worse than that: They had to work to find a "burnt out" bulb. They had to make an effort to make the site broken.

SweetJesus: Simple. It's more cost-effective to have some IT G2 rapidly prototype a simple web application using FrontPage than it is to hire someone to write standards compliant code, by hand. It also helps that 90% of the market uses IE for web browsing.

It's logic, not conspiracy. But then again, you still haven't answered my question.


In case you're not satisfied with Rothko's answer (which has the merit of being concise), I'll put it another way: You're wrong. Because your premise is wrong. It would not be more cost-effective to have a grunt rapid-prototype a high-volume application in browser-specific FrontPage code. It would be more cost effective to do it using simple, standard POST forms that are standard to every web browser out there -- including Microsoft's.

How do I know that? The same way that Rothko does: It's one of the things I get paid to know.

This is just so not about Microsoft bashing. It's about inexplicability. As in, there's really just no good explanation for why this kind of thing happens. I'll say this again, not just because you don't seem to get it, but because it amazes me: It takes extra work to make a simple form that fails for non-MS browsers. It takes less work to make the form work. This is not rocket surgery -- there's not even any DHTML or JavaScript invovled. It's just HTML and HTTP. How hard coudl it be? If you're an IT G2 and you don't know how to build a simple form, you are a sad piece of shit and you should not be employed with tax payer dollars.

Jesus....
posted by lodurr at 11:53 AM on September 26, 2005


I wasn't snarky with SweetJesus until he somehow Godwin'ed the conversation. In fact, up until that point, I was pretty damn polite and respectful, despite repeated baiting and insults. Enough is enough.
posted by Rothko at 11:54 AM on September 26, 2005


... oh, and after lookng at OptimusChyme's efficient response farther up thread: Rapid prototyping an application in FrontPage is one of the surest ways to make sure you get a profoundly non-scalable web app. That's also something even a G2 should know. So the "they knew it would work, why fix it" argument just doesn't fly.
posted by lodurr at 11:58 AM on September 26, 2005


If you're an IT G2 and you don't know how to build a simple form, you are a sad piece of shit and you should not be employed with tax payer dollars.

I see you've never worked for the Government before.
posted by SweetJesus at 11:59 AM on September 26, 2005


Cute response, but what does it mean? Are you saying that we ought to just suck up and deal? That we ought to say "It could be worse, we should be grateful they got off their asses to do anything."

Maybe I'm not beng clear, here: This is really web 101. I know secretaries who can code POST forms. They had to work to make this break in non-MS browsers. Then they had to work some more to make it report errors when it broke in non-MS browsers.
posted by lodurr at 12:05 PM on September 26, 2005


I wasn't snarky with SweetJesus until he somehow Godwin'ed the conversation. In fact, up until that point, I was pretty damn polite and respectful, despite repeated baiting and insults.

Bullshit.
posted by SweetJesus at 12:14 PM on September 26, 2005


SweetJesus: producing a cross-browser compatible website is only a pain in the ass if you were stupid enough to use browser-specific extensions in the first place. The whole 'standards-compliant development is expensive' line is a fraud perpetrated by incompetents who need the FrontPage crutch.

Under your definition any standardization amounts to corporate welfare.

Ugh. I think it's pretty clear that the problem isn't with 'any standardization,' but with a choice that benefits one company to the exclusion of all others. I don't believe there's any kind of conspiracy here, and I don't expect that anyone will choose microsoft products solely on the basis of their compatibility with the FEMA website, but it's foolish to think this has no effect on purchasing decisions.

It's also unreasonable to discount this problem because it only affects 10% (or whatever) of the population. The actual flooding in NO certainly affects much less than 10% of the US population; maybe we should just tell them to upgrade their real estate if they have a problem with all the water? The government could save lots of my taxpayer dollars that way!
posted by sriracha at 12:16 PM on September 26, 2005


Bullshit.
posted by SweetJesus at 3:14 PM EST on September 26 [!]


Your comments are all up-thread for fair-minded people to read. Are you saying you didn't Godwin this thread?
posted by Rothko at 12:22 PM on September 26, 2005


SweetJesus: producing a cross-browser compatible website is only a pain in the ass if you were stupid enough to use browser-specific extensions in the first place. The whole 'standards-compliant development is expensive' line is a fraud perpetrated by incompetents who need the FrontPage crutch.

My whole point is that these web pages are written (ok, not written, designed) by people who's main job isn't doing web development. If you talked about POST their eyes would gloss over, never mind talking about standards compliance.

Find me a WYSIWYG web editor with good technical support that produces standards compliant code, and is so easy a monkey could use it.

Then you'll get your wish.
posted by SweetJesus at 12:26 PM on September 26, 2005


My whole point is that these web pages are written (ok, not written, designed) by people who's main job isn't doing web development. If you talked about POST their eyes would gloss over, never mind talking about standards compliance.

... And if that were true (which I really, really doubt) that would be a problem. Worse than a problem: It would be negligence. It would mean that you didn't even bother to consider scalability (oh, yeah, eyes would glaze over then, too).

In other words, you are making our point for us: This is incompetence at least bordering on negligence, and the people responsible for it should be at least reprimanded. "That's gommint woik for ya!" just don't cut it.

It's got nothing to do with how good the WYSIWYG editors are; if your scenario is true, it's got to do with making responsible management decisions. Regardless of whether we have any reason to expect Bushite appointees to do that, we have a right -- and, I would argue, a duty -- to expect it.
posted by lodurr at 12:32 PM on September 26, 2005


Lodurr, you're wasting your time on him.
posted by Rothko at 12:37 PM on September 26, 2005


I quite simply don't agree with you, is all. I just don't think it's a priority in the way you do, and I don't think it would save any money despite your insistence. I don't think you can convince me otherwise.

... And if that were true (which I really, really doubt) that would be a problem.

Again, I can tell you've never worked for the government.
posted by SweetJesus at 12:38 PM on September 26, 2005


It's got nothing to do with how good the WYSIWYG editors are; if your scenario is true, it's got to do with making responsible management decisions.

So, how is the w3c preaching standards to managers? I'm curious as to how this is being done today, because it seems as though the only ones who are getting this information are programmers.
posted by dflemingdotorg at 12:40 PM on September 26, 2005


So let me put it to you this way, SweetJesus and dflemingotoro -- let me ask you this again, this time explicitly:

Should we just shut up and be glad that someone "working for the government" actually decided to do anything, and not criticize its quality?

'Cuz that seems to be where you've gotten to: We shouldn't be criticizing the quality of this website because it's "government work."

Which is a pretty civically irresponsible position, to say the least -- but, oh well.
posted by lodurr at 12:45 PM on September 26, 2005


My whole point is that these web pages are written (ok, not written, designed) by people who's main job isn't doing web development. If you talked about POST their eyes would gloss over, never mind talking about standards compliance.

It's the year 2005. The federal government needs web pages. This is not news. If they're trusting that job to some dimwit using motherfucking FrontPage, then we should just burn the whole goddamn system to the ground.

Wait a minute. SweetJesus, you don't happen to have a government job, do you? Maybe in some technical field?
posted by Optimus Chyme at 12:45 PM on September 26, 2005


'Cuz that seems to be where you've gotten to: We shouldn't be criticizing the quality of this website because it's "government work."

Don't criticize FEMA, citizen; they might let one of five die in a hurricane from managerial negligence, but that's just how market share works.
posted by Rothko at 12:49 PM on September 26, 2005


Lodurr, I'm not saying that at all. What I'm saying is that if the message of standards compliance isn't getting through, perhaps there's a problem in the way we're presenting it? If eyes are being glossed over, is it not the fault of the presenter for not targetting his audience?

To a manager, 80% of users probably sounds really, really good. Granted, you and I both know that there's a better way, but if what we're saying is insular and missing the point, is it not the responsibility of the community to make sure the message gets through?

Managers, usually, don't know enough to make this type of decision by themselves. They manage things and people, they don't see the front-line.
posted by dflemingdotorg at 12:50 PM on September 26, 2005


Should we just shut up and be glad that someone "working for the government" actually decided to do anything, and not criticize its quality?

You can feel free to criticize all you want, but I just see it as nitpicking. In the totality of everything the government does, this is next to last on my "importance" list. I just feel you guys, and by extension most of Metafilter, cares about this a whole hell of a lot more than most people.

I see it as a tempest in a tea cup.
posted by SweetJesus at 12:53 PM on September 26, 2005


Also, I should mention that I wouldn't write Windows software if you paid me. I use Linux and Firefox, but I understand that I'm in the minority.
posted by SweetJesus at 12:55 PM on September 26, 2005


I see it as a tempest in a tea cup.

SweetJesus, I see you as having taken several different positions to prolong this discussion. You could have simply said up top that you thought it was a waste of time to care, and then not said anything more. It would have been a lot less work and a lot more true to your stated beliefs.


dflemingdotorg: In this context, I actually don't care about standards compliance. It's damn near irrelevant. The application fails for a significant portion of its users, because it was badly designed. Furthermore, making it work the way it does took extra work -- that is to say, they had to make a decision to make it work such that it is guaranteed to fail for something around 20% of users. (That's not counting the poor back-end design that, by Rothko's account, makes it then fail for 80% of the remainder.)
posted by lodurr at 1:06 PM on September 26, 2005


What I'm saying is that if the message of standards compliance isn't getting through, perhaps there's a problem in the way we're presenting it? ... To a manager, 80% of users probably sounds really, really good.

Standards compliance for government work should be dictated by law. If that makes the manager's eyes glaze over, they should find themselves in court. Managers in the corporate world have the luxury of making the choice to ignore 10% of their market; managers in the government should not.

For a government site that serves the general public, 80% should sound like an utter failure (which it is).
posted by sriracha at 1:12 PM on September 26, 2005



SweetJesus, I see you as having taken several different positions to prolong this discussion. You could have simply said up top that you thought it was a waste of time to care, and then not said anything more. It would have been a lot less work and a lot more true to your stated beliefs.


What? My position hasn't changed one iota. I just don't think it's that big of a deal, and I don't see how it would be less expensive to hire someone to hand-code standards complaint HTML. I think it's a waste of time to care, and that energy could be put to better use elsewhere. I don't think you understand the ways in which the government operates in these situations, and are taking a very simplistic view.

But please don't tell me what I think.
posted by SweetJesus at 1:24 PM on September 26, 2005


Pollomacho writes "I doubt that very many of these folks have electricity much less an internet connection and functional computer."


Libraries, friends, relatives, internet cafes, schools, shelters, airports.

SweetJesus writes "The Government, in an attempt to save you tax money, is following the current 'de-facto' standard for your average person who uses the internet."

This money saving scheme is in violation of the ADA, in spirit if not fact.
posted by Mitheral at 1:41 PM on September 26, 2005


SweetJesus writes "It's more cost-effective to have some IT G2 rapidly prototype a simple web application using FrontPage than it is to hire someone to write standards compliant code, by hand."

Are you under the impression that FrontPage is the only WYSIWYG style web page editor? If not why all this talk about hand coding?

SweetJesus writes "If you can't tell the difference between an inconvenience and a necessity, we're screwed. "

Ever ran a screen reader SJ?

SweetJesus writes "My whole point is that these web pages are written (ok, not written, designed) by people who's main job isn't doing web development"

So what your saying is this is a problem of incompent HR and Managers hiring the wrong kind of developer? How hard is it to find a decent web designer? It's not like this is the dot com era or something.

dflemingdotorg writes "To a manager, 80% of users probably sounds really, really good."

Not to any goverment manager I've ever worked for. When you dealing with 100s of thousands of people your generally wanting to get well above 99% accessability. At 99% and 100,000 clients your talking 1000 people emailing their MPs, Ministers and Deputy Ministers. This lack of compliance is the kind of thing that gives goverment middle managers ulcers.
posted by Mitheral at 2:00 PM on September 26, 2005


This money saving scheme is in violation of the ADA, in spirit if not fact.
posted by Mitheral at 4:41 PM EST on September 26 [!]


Don't bring up the ADA, SweetJesus will bring up Adolf Hitler.
posted by Rothko at 2:20 PM on September 26, 2005


So what your saying is this is a problem of incompent HR and Managers hiring the wrong kind of developer? How hard is it to find a decent web designer? It's not like this is the dot com era or something.

It's not hard to find a good web designer, it's hard to find a good web designer who will work for 35K a year in a job that eats your soul.

Are you under the impression that FrontPage is the only WYSIWYG style web page editor? If not why all this talk about hand coding?

It's a generic example.

Ever ran a screen reader SJ?

Nope, because I'm not blind. But I know what they are, and I understand the concept behind them. I still think Alex's example is specious reasoning taken to the extreme.
posted by SweetJesus at 2:23 PM on September 26, 2005


You can feel free to criticize all you want, but I just see it as nitpicking. In the totality of everything the government does, this is next to last on my "importance" list.

Well, as someone who is quickly slipping into poverty and homelessness because FEMA cannot be trusted to do something as simple as create a Web page that works, I say, fuck you. If it desn't matetr to you, don't participate in the conversation. It's dreadfully important to me, and to a half million others who are displaced and running out of options as a result of the destruction of our city.

At this moment, I have not even receved the packet FEMA promised me two weeks ago that would explain the process.
posted by maxsparber at 2:26 PM on September 26, 2005


Well, as someone who is quickly slipping into poverty and homelessness because FEMA cannot be trusted to do something as simple as create a Web page that works, I say, fuck you. If it doesn't matter to you, don't participate in the conversation. It's dreadfully important to me, and to a half million others who are displaced and running out of options as a result of the destruction of our city.

Well, obviously that's because FEMA didn't comply to browser standards for Firefox.

I feel for your situation, but FEMA's incompetence does not stem from browser standards. Apparently they can't even mail shit properly, which is a much lager problem. As I said before, I think FEMA's got a lot more to worry about than getting their web pages to work in linux browsers.
posted by SweetJesus at 2:33 PM on September 26, 2005


It's not hard to find a good web designer, it's hard to find a good web designer who will work for 35K a year in a job that eats your soul.

Working eight-hour days with great benefits and the assurance I will never ever ever be fired, all for taking five minutes to write a couple of pages in Notepad and reading MeFi the rest of the day? Sign me up.
posted by Optimus Chyme at 2:34 PM on September 26, 2005


I still think Alex's example is specious reasoning taken to the extreme.
posted by SweetJesus at 5:23 PM EST on September 26 [!]


How about chipr and maxsparber's examples?

On preview: SJ, you are simply too much! Go down with that flaming ship.
posted by Rothko at 2:35 PM on September 26, 2005


"If they're trusting that job to some dimwit using motherfucking FrontPage, then we should just burn the whole goddamn system to the ground."

That's what I've been trying to tell everyone all along.

You can expect quality help when your employer's main perk is how many paid time off you get each year. Not just holidays, vacation time and sick leave either, but arriving late in the morning, leaving early in the afternoon, flex time (what a joke!), extended lunches, and "oh what the fuck I think I'll wander down the street to Dunkin' Donuts" mid-afternoon breaks.
posted by mischief at 3:41 PM on September 26, 2005


You can't ... ^^
posted by mischief at 3:42 PM on September 26, 2005


I don't know who this mythical government IT employee is that SweetJesus et al keep holding up as some archetype. I did write HTML for a federal entity. We did not use FrontPage and we did code to W3C and 508c standards using a freeware text editor geared toward an HTML environment (syntax highlighting + autocompletion). Mind you, this was well before the technorati were pulling the standards bandwagon. We kept all schedules, brought the project in under budget, and managed to not turn away any member of the browsing public. The inefficient federal employee sure is fun to ridicule, but it was my experience that they are fairly difficult to find. In fact, the reason I was assigned to this team was because my own department was temporarily over-staffed--the work was "seasonal". Futher, I found most ineffciencies occur when trying to comply with some over-sight fuckwittery brought about because of this predisposition to view government employees as incompetent and lazy. I work in the private sector now and I cannot see any difference in work ethic.

Back to the topic: I imagine the rumor that this was an internal application that FEMA released into the wild by necessity is true.

Diverging yet again from the topic: Then again, it serves as yet another example to bring to one's managers' attention when they justify cutting corners and using FrontPage/.Net/Visual Studio as a package because everything in-house is Microsoft/IE. Write to a published standard—with a little care in how it degrades—and your app/site/whatever will work better and far longer than any application-specific code. And, ultimately, it does cost less. Investment up-front yields better dividends as my uncle always said.
posted by Suck Poppet at 4:41 PM on September 26, 2005


SweetJesus: ... and I don't see how it would be less expensive to hire someone to hand-code standards complaint HTML.

That's because you haven't got a freaking clue what you're talking about.

Really, you're making a fool out of yourself by saying stuff like that on a site that's thick with web professionals -- designers, coders, admins -- and interested amateurs. People, in other words, who understand that coding to standards is actually easier than not coding to standards. You clearly lack the experience to make that judgement; I don't.
posted by lodurr at 5:56 PM on September 26, 2005


"The inefficient federal employee sure is fun to ridicule, but it was my experience that they are fairly difficult to find."

There's a reason for that. Just ask New Orleans. ;-P
posted by mischief at 6:13 PM on September 26, 2005


Not sure he was talking about cabinet positions, mischief. ;-P
posted by Rothko at 6:16 PM on September 26, 2005


Good thing I wasn't talking about cabinet positions either.
posted by mischief at 6:29 PM on September 26, 2005


My cabinet position is "crouched".
posted by lodurr at 7:06 PM on September 26, 2005


There's a reason for that. Just ask New Orleans. ;-P
posted by mischief at 9:13 PM EST on September 26 [!]


1. Notice the qualifier "fairly difficult" in my statement. Notice the mention about oversight fuckwittery. In your specific example, consider the joys of large-scale reorganization under the neglectful watch of the current administration might as possibly having an impact as well. How much of the train-wreck that is FEMA's response to Katrina due to the laziness of the front-line FEMA staff verses inefficiency introduced by meddling from above?

2. Inefficiency as a result of poor leadership is at least as prevalent in the private sector in my experience. Enjoy your confirmation bias though.

3. In other words: nyaah!

gee this is fun!
posted by Suck Poppet at 7:09 PM on September 26, 2005


Consider that the rank and file will survive this administration relatively unscathed. As ever. And still slurping down lattes on the public dime.

Unlike the federal bureaucracy, inefficiency in the private sector has a limiting factor, bankruptcy.

Your faith is misguided.
posted by mischief at 7:16 PM on September 26, 2005


The ten minutes I spent reading this post are ten minutes that I will, regretablly, never get back.
posted by trey at 7:24 PM on September 26, 2005


trey: that's cuz the discussion is focusing on a symptom, not the disease.
posted by mischief at 7:27 PM on September 26, 2005


your mother and father were misguided, norquist!

which is about as pertinent to your comment as yours was to mine

weeeeeeeeeeeeeeeeee!
posted by Suck Poppet at 7:28 PM on September 26, 2005


... and the "disease" would be.....?

coyness is unbecoming in the righteous, mischief. If you really want to say something, say it, and don't leave us to fill in your blanks. Like, say, with the observation that you seem to be harboring some resentment over the perceived excesses of free time enjoyed by the crowds of "latte-slurping" gummint teat-sippers....
posted by lodurr at 7:46 PM on September 26, 2005


Teat-sippers. Good one. I'll have to note that.
posted by mischief at 8:22 PM on September 26, 2005


"Where do you want to go today?"

Into a HOUSE, for the LOVE of GOD! Please, PLEASE, just get me away from this WATER!
posted by EtJabberwock at 8:41 PM on September 26, 2005



That's because you haven't got a freaking clue what you're talking about.

Really, you're making a fool out of yourself by saying stuff like that on a site that's thick with web professionals -- designers, coders, admins -- and interested amateurs. People, in other words, who understand that coding to standards is actually easier than not coding to standards. You clearly lack the experience to make that judgement; I don't.


If you say so, since you're such an expert. I've been dabbling with web development for 8 years, and spent three 3 years doing it both professionally for money and altruistically for college credit. Professionally, I've gotten out of the web world a bit over the past two years, but I've been playing with the new AJAX framework for fun in my spare time, and I don't think you're the expert you think you are. You simply keep telling me that it is the way you say it is, and don't offer anything other than that - not even an anecdote.

Anyway, I work with, not for, the government helping to design distributed simulation software. I deal with FrontPage constructed websites and Visual Basic network applications each and every day. In my experience, they're not written by people with training in Computer Science, and they're usually poorly constructed and prone to error. That, I agree with you, is a problem. But it's also a different conversation.

But really, this discussion has spun off course. My initial comment was directed at Alex's post that choosing designing for a browser that 90% of the world uses is somehow corporate welfare, which I still think is profoundly stupid. The overall topic of the post is that the government is forcing Katrina survivors to use a website that only works in IE 6. This is annoying, yes, and in an ideal world there would be enough time to write standard's complaint code for everyone. But taking in the totality of the Katrina disaster, I can't muster one gram of outrage over the disaster website not working in Firefox. FEMA employees are working quickly to develop something that can be used by the majority of people in the quickest amount of time. If you want to have a discussion of why this work wasn't done pre-disaster, that's a different talk.

The biggest problem I have with the web standards movement (espically comments in this thread realting to it) is the zealotry that is involved with it. I have similar feelings about Cory Doctorow and the anti-DRM movement - "I agree with you in principal, but when you're foaming at the mouth like that it's hard to take you seriously". I think, in some situations, you're probably right. I'd like to most government web site front-ends work well in all browsers, but in this situation I really find it hard to give a shit - at least the amount of shit you guys give.

Pun intended.

The ten minutes I spent reading this post are ten minutes that I will, regretablly, never get back.

Just imagine if you wrote part of it...
posted by SweetJesus at 10:40 PM on September 26, 2005


My initial comment was directed at Alex's post that choosing designing for a browser that 90% of the world uses is somehow corporate welfare, which I still think is profoundly stupid.

You're welcome to think that, but you're wrong.

First, FEMA is not writing a website for 90% of the world. They are writing a website for 100% of the taxpaying citizens of the U.S. of A. Secondly, your client numbers are wrong. Thirdly, (poorly) writing an application for one company's product to the deliberate exclusion of others is corporate welfare by any definition.

You really don't know much about web design or application development. You've had a handful of knowledgeable people demonstrate to you that you don't know what you're talking about. Take some advice from someone who's been there and close your mouth already.
posted by Rothko at 12:10 AM on September 27, 2005


[Data point: The form pages in the FEMA application wizard were definitely not "rapid-prototyped in FrontPage." They could have been done (rapidly or otherwise) in Interdev, but if so, they were done by an at least moderately-experienced coder -- someone not using the default Interdev page templates. (Alex, you didn't check? Tsk!) Meaning that the "clueless G2" theory can be pretty much tossed: This thing is intentional; it's just badly done. Also, I think I pretty much know why they used IE-only, now, in case anyone is curious. Suffice to say that they appear to be using modal dialogs for action-confirmation (another conversation), and may also be using IE-only DHTML menu code. That's as much as I'm going to work at it, at least for now.]

sj: I've been dabbling with web development for 8 years, and spent three 3 years doing it both professionally for money and altruistically for college credit.

Well, so it's down to comparing resumes, is it....

I got my first job as a web developer in February of 1997. I spent about six years working in HTML and ASP, have worked mostly in PHP since 2002. I did most of the presentation layer coding for a couple of high-volume online training apps, designed two different presentation layer frameworks on my own (one PHP, one VBScript), started hand-coding cross-browser DHTML menu systems back when that meant making them work with NN4, have done numerous project estimations and requirement docs for apps ranging from trivial to ones expected to have up to thousands of simultaneous users. These days I'm mostly out of the hard-core technical game, though I do a lot of mostly-CSS formatting for a design firm (just to keep fresh) and I still do a fair amount of specification and project planning. So that, theoretically, makes us peers. Which I find interesting, because a lot of the stuff you're saying -- it just doesn't make sense.

You don't seem to really have an understanding of what parts of the DOM vary between browsers, for example. Cross browser problems are largely issues of rendering, and occasionally of DOM reference, but not of form handling -- at least, not in my experience. In fact, I can't think of any currently relevant differences in form handling between IE and any other DOM-compliant browser. There are differences, but they're not relevant to this type of application.

Put another way: There really isn't any reason for this app to be browser-specific.

I'll be blunt, SweetJesus: I don't know what you could have been doing for those eight years. Because you honestly, truly do not seem to know what you are talking about, with regard to standard behaviors, proprietary extensions, development environments, and project costing for web application development, to name what i can think of off the top of my head.

[Fun experiment for someone who needs FEMA assistance: Try spoofing the browser detection code. I looked at the form with Opera identifying itself as IE6, so I know that their browser-exclusion routine is in JavaScript and only uses the browser identification string. I'd love to know if there's actually not any browser-specific functionality in the app at all, aside from the modal dialogs -- that would be pretty amusing at this point.]

sj: My initial comment was directed at Alex's post that choosing designing for a browser that 90% of the world uses is somehow corporate welfare, which I still think is profoundly stupid.

(And from that initial point, you progressed to several other points, including: it's more cost-effective to limit the scope of your clients; it's more cost-effective to rapid-prototype, so we shouldn't complain when rapid-protoyped apps fail; and, my personal favorite, Government workers are lazy and stupid and we shouldn't expect better. But I digress...)

You know what? I agree: The corporate welfare angle was overdone. But I disagree, strongly, that it's "stupid" to want to service that lost "10%". (Alex is right, BTW -- your numbers are either old or are skewed somehow.) Or that it excuses FEMA in any way to point out that the app was "released into the wild" after internal use. If the online forms had been "released" just to service Katrina evacuees, you might have a case -- but i see no indication that's true. The browser exclusion, combined with the embarrassing array of failure modes noted up-thread, indicate that people planning for this kind of thing at FEMA planned poorly, to put it mildly: They didn't anticipate the load the application would come under, or they didn't anticipate delivering it to the field. That's embarrassing incompetence, and it's incompetence that's not limited to the browser. It's indicative of wrong attitude somewhere in an organization that's supposed to be beholden to us. We have a duty to expect better.

It's something worth caring about. Yes, this discussion is about one "symptom" (to borrow mischief's metaphor) inasmuch as we're discussing one web application -- but it's about the "disease" in that what we're talking about is a profoundly cavalier attitude toward customer service and professionalism. Where we're the potential customers, and where the service, when needed, is critical.

Now, as for your disingenuous protestations that you don't really care about this -- well, hey, you're still here, dude. You can leave any time you want to.

posted by lodurr at 5:29 AM on September 27, 2005


Changing tack for a moment: I followed links to a story about the Copyright Office, and its plans to "restrict" users to IE6 and NN4 for form submissions.

That set off my logic-bells. Because if a form will work in both NN4 and IE, it will work in any DOM browser. Unless Siebel (who actually wrote the app) is doing something really obselete and weird, like delivering LAYER tagsets based on the browser ID string....

If you read between the lines, what you've got there is an extreme case of CYA: Whoever is responsible for establishing the browser standards for the Patent Office's site would rather play hard-line rules-man ("you MUST use browser x") than run the small risk of displeasing anyone -- or the geeks at Siebel can't be bothered to spend five minutes on the phone with them to tell them "dude, it'll work in Mozilla."

I used to spend a lot of time looking at PeopleclickVMS screens. At that time, they wouldn't certify that the app worked in Mozilla browsers, but I used it that way all the time. (Made it easier to get multiple logins -- one login in IE, one in Mozilla, a third, if necessary, in Opera...) The developers had clearly built cross-browser support in, and done a reasonably good job of it, but it's as though nobody higher than a team lead knew about it. I'm sure it was in their design docs -- the JavaScript was nicely structured, they were clearly using consistent libraries -- but the mid-level people were too scared of the details to let customers know that it would work.

There seems to be an attitude in many corporations that if it's not MS, it's somehoe not "serious". I would get the feeling from MSP reps that they didn't want to admit that they had even considered the possibility of using their apps in anything but IE -- that doing so would mark them as Bolsheviks or something. They'd get clam-mouthed and wide-eyed when I'd bring it up. This was true even when they were ex-geeks, who ought not to have been "challenged" by a little technical adventurism.

I was looking to the future, and thinking about potential browser-sensitive clients (like AOL or Sun or some college or university). Part of my job was to do bullshit detection for our own sales teams. So I'm testing this stuff whenever I get the chance. Plus, it gave me a chance to see how serious they were about their error-trapping -- if our customers lost data due to a problem with an MSP's app, they screamed at us, not the MSP. So I'll complain, for example, about a lot of things to do with PeopleclickVMS, but I never saw a Moz-related error that caused data loss. Even though non-IE browsers were a guilty, dirty-little secret, they at least dealt with it.

That told me that, all other things aside, they were serious about making something that had quality. It gave me a reason to trust them, at least a little.
posted by lodurr at 6:00 AM on September 27, 2005


I'll be blunt, SweetJesus: I don't know what you could have been doing for those eight years. Because you honestly, truly do not seem to know what you are talking about, with regard to standard behaviors, proprietary extensions, development environments, and project costing for web application development, to name what i can think of off the top of my head.

Well, so it's down to comparing resumes, is it....

Well, when you tell me I don't know what the fuck I'm talking about, yeah I guess it comes down to that.

(Alex is right, BTW -- your numbers are either old or are skewed somehow.)

The 90% number is wrong? Sorry, I was off by 1.24%. Way to pick a nit.

And from that initial point, you progressed to several other points, including: it's more cost-effective to limit the scope of your clients; it's more cost-effective to rapid-prototype, so we shouldn't complain when rapid-protoyped apps fail; and, my personal favorite, Government workers are lazy and stupid and we shouldn't expect better.

Ugh, you read way too much in to what I say. When did I say government workers are lazy? I said it's hard to find good people willing to do the work for the amount of money they pay. I said most of the time the people doing the web work are not computer scientists, but know how to write a word document, so those skills are transferred over. Anything else you infer is your own business. I don't remember calling anyone stupid or lazy.

I also never said you shouldn't complain when rapidly prototyped applications fail, I said it's often less expensive to rapidly prototype an application. Ever heard of RAD methodology? How about agile software development? You may not agree with me, but don't misstate what I said. There are plenty of people who believe these methodologies save time and money.

Now, as for your disingenuous protestations that you don't really care about this -- well, hey, you're still here, dude. You can leave any time you want to.

If you keep telling me I don't know what I'm talking about, I'm going to keep coming back. I'm smart enough not to get into a pissing match with Rothko, because those never go any where. But I'll continue to discuss this with you.

I still don't care very much about the overall topic of the post, which is the FEMA app not working in Firefox. What I care about is you coming back, over and over again, to say I have no idea what I'm talking about. I have a very good idea what I'm talking about, I just disagree with you. We can have a fucking professional disagreement over these sorts of things without resorting to fucking name calling.
posted by SweetJesus at 8:03 AM on September 27, 2005


Very well, then: What are you talking about?
  • Do you agree that 90% is a poor quality target?
  • Do you agree that it was poor planning to release a web application which could be guaranteed to fail in at least 10% of cases?
  • Do you think, based on your experience, that there is a reason for using IE that justifies sacrificing 10% of the audience? (If your justification is cost savings, please provide supporting examples and arguments. Tokens will do.)
  • Do you agree that it was poor planning to release a web app that was so poorly tested, it failed for 80% of cases under heavy use?
  • Do you agree that we ought to expect the release of mission-critical applications, whether created by the government or not, to be well-planned? Should the fact that an app was produced by the gov't have any bearing on our expectations?
  • I'm thinking we agree that selecting IE only wasn't a conscious subsidy, but do you agree that specifying IE only would constitute a de facto endorsement of MS technology?
posted by lodurr at 8:13 AM on September 27, 2005


We can have a fucking professional disagreement over these sorts of things without resorting to fucking name calling.
posted by SweetJesus at 11:03 AM EST on September 27 [!]


Assuming all parties to the table are professionals. It's generally not professional to bring Adolf Hitler into a professonal conversation, but that's just this one professional's opinion.
posted by Rothko at 9:16 AM on September 27, 2005


Assuming all parties to the table are professionals. It's generally not professional to bring Adolf Hitler into a professional conversation, but that's just this one professional's opinion.

Lodurr, I'll get back to you on those points later tonight, I don't have time now.

Rothko, I've resisted your childish Hitler baiting for the whole thread. Your capacity for willful ignorance astounds, because if you couldn't tell I was being facetious when I said "Lets bring Hitler into this right now, and get it over with" then you're a fucking moron. It's more likely that you're just trolling around looking for another dramatic, pig-ignorant fight about nothing.

Fuck off.
posted by SweetJesus at 10:44 AM on September 27, 2005


I'm still not getting why it would have had to have been browser-specific, though. Considering how much most web development firms charge for their services, it's not like they couldn't have been a little charitable and put in a little extra effort to test the thing in other browsers to ensure a broader range of customer satisfaction. Especially when "customer satisfaction" in this case means "helping people who lost everything in a natural disaster start over again".

And it's not like the government wouldn't be able to toss on a couple extra grand for that additional development and testing, either, if the company demanded it. Of course, it's not like the government really understands any of that, either. For instance, Rhode Island's health department seems stuck on using MS Access for everything, which has to be one of the most useless pieces of database software imaginable. But they won't even consider using online databases, for some reason, because Access is a Microsoft product, and so it MUST be the best they can possibly do.

...or the geeks at Siebel can't be bothered to spend five minutes on the phone with them to tell them "dude, it'll work in Mozilla." --lodurr

Geeks... using phones? You might as well go back in time and ask them to go to their senior proms.


On a related note, I made this comic about a month ago.
posted by EtJabberwock at 12:30 PM on September 27, 2005


Ever heard of RAD methodology?

Oh, is that the methodology where you never fix any bugs, no matter how incompetent they make your organization look?

It's not like this is some hacked-together website that they threw up in a hurry to cope with a disaster. Is it? If so, I apologize for all the nasty things I said about them. But it doesn't look likely.
posted by sfenders at 12:37 PM on September 27, 2005


Fuck off.
posted by SweetJesus at 1:44 PM EST on September 27 [!]


You were baiting me the entire thread with the conspiracy theory and "Alex" stuff, and when your argument collapsed and you had nothing left, you threw in Hitler.

If anybody's behavior lately deserves a Metatalk thread it was yours. I stayed, polite, reserved and on point, and I tried to carry a reasoned discussion with you. I know you have to throw in your hatred for me along with the rest of the cabal here, but next time don't make it so goddamned obvious.

You don't know a damn thing about software engineering, you still don't know what you're talking about, so shut your mouth already, and, in any case, enjoy a well-deserved "fuck off" right back at you. Sweet-fucking-Jesus, indeed, nimrod.
posted by Rothko at 1:06 PM on September 27, 2005


Some additional thoughts RAD and agile methodology, and their applicability to this circumstance:

If you showed this app to an exponent of agile methods, they would cringe. They might launch into a passionate discourse on how the FEMA application illustrated the need for formalized agile approaches.

sfenders, you're too right about what RAD is usually like, AFAICS. But RAD is not synonymous with "agile", of course; folks who are really into agile methods take great pains to make that clear. Indeed, this kind of application can be a great place for agile methods, and it may have started out that way: The team that built it may have produced what they thought was an internal prototype, that they intended to refine and refactor before taking it to a wider audience. That's a common agile method.

One of the common criticisms of agile methods is that the prototypes then get used as the final product. In all the formalized agile methods I've read about or had conversations about, it's not supposed to work that way. At the very least, the initial version gets fixed; it may get entirely refactored. (I've never used a formal agile methodology, myself, though we did use a home-brewed quasi-agile methodology on a community site I built a few years ago.)

Of course, an app often does go straight from prototype to production without bugfixes -- but I venture to say that it works that way more often in traditional methodologies than in agile methodologies. I can personally point to four examples: Three I was personally involved with, one that was a legend in my group. The last is the best story. The team was told they needed a "demo" in time for a sales conference. They burned the midnight oil to make one, by taking their unfinished interface and writing a bunch of canned simulations that they compiled into a DLL called "smokeAndMirrors" (I kid you not). Management saw it, pronounced it good, and said, "You'll be ready to release in a month, right?" For the next two release versions, that software was filled with calls to smokeAndMirrors....

This is just a bad piece of work, all around. Invoking agile methods in its vicinity is a disservice both to agile methodologies (since this clearly wasn't done in the spirit they intend) and to traditional methodologies (since it prevents them from learning anything from agile methodolgoies by having misrepresented them).

(In other words: Yes, SweetJesus, I have heard of "agile" and "RAD". You wanted anecdotes? You got 'em.)

(Rothko: Sorry; someone told me once that you were AlexReynolds, and that's why I went along with calling you Alex.)
posted by lodurr at 1:34 PM on September 27, 2005


Rothko is AlexReyonlds...
posted by SweetJesus at 1:57 PM on September 27, 2005


lodurr, I have in fact done (and still do) plenty of development that could be (and sometimes officially was) described as RAD. I like it. But IMO it does depend more heavily than some other methods on having good formal test procedures. If you don't keep track of the defects, and correct them, they just get worse with each iteration.

Anyway, yeah, it's pretty easy to put half-baked systems into production without sufficient testing using pretty much any methodology.
posted by sfenders at 3:28 PM on September 27, 2005


Lodor: Don't mean to drag this thing out any longer, but I've been busy today, so no substantial Metafilter comments until 1am. I'm going to answer you point by point.

Do you agree that 90% is a poor quality

No, I think 90% is damn good. 100% is ideal, but it depends on the software you're developing, the software you develop with, the level of skill in your development team, your time frame for construction, testing and deployment and the competence of your managers.

Do you agree that it was poor planning to release a web application which could be guaranteed to fail in at least 10% of cases?

Again, that depends on wether or not you think 90% is a poor target. It's safe to say that you do. IE 6 SP1 runs on Windows 98 and above, which is the most popular operating system family in the world for home and business use. Most OEM computers sold in the US come with Windows and IE pre-installed, and it stands to reason that most donated computers would be running Windows. So, in that respect, I don't think it's a poor development methodology.

Do you think, based on your experience, that there is a reason for using IE that justifies sacrificing 10% of the audience? (If your justification is cost savings, please provide supporting examples and arguments. Tokens will do.)

Time, money, skill... What's your backend running on, what's your database guy trained in - Oracle, SQL Server or Postgres? It depends on the project. It depends on thousands of small decisions and variables. I don't' have any IE-specific examples, but the software I write only runs on Linux, so it could be said that I ignore about 95% of the audience, but that wouldn't really be true.

But I can make some arguments for rapid prototyping and RAD methodology. Which is cheaper and easier - Writing GTK GUI's by hand in a text editor, or using something like Glade to make the GUI visually and generate a skeleton? The answer is overwhelmingly #2. This, to an extent, is the thinking of people who use FrontPage or NetObjects or even the shitty visual HTML editor KDE comes with.

Do you agree that it was poor planning to release a web app that was so poorly tested, it failed for 80% of cases under heavy use?

Yes, I'm a big fan of a QA engineering, specifically the use of regression testing tools. I think they should have anticipated the traffic they would get, and dealt with it. But I would think something like that would be a backend issue, something to do with their choice of database or web server. I think that's an independent issue.

Do you agree that we ought to expect the release of mission-critical applications, whether created by the government or not, to be well-planned? Should the fact that an app was produced by the gov't have any bearing on our expectations?

Well, I think it's pretty clear that FEMA is a mess. I would be interested to know if this was an inside FEMA job, or was sub-contracted out. I would be very, very interested in knowing the name of the sub-contractor if the app wasn't internal to FEMA, as there are many companies out there who take contracts and deliver shit. If it was internal to FEMA, I imagine it's shoddiness is a result of FEMA's disarray over the past 5 year.

Should mission critical applications written for or by the government work correctly, and above all work well? Yes, absolutely. Should mission critical web applications be designed for each and every browser according DOM 3 standards? There, I'm not with you so much. I think it depends on the factors I mentioned in #2.

I'm thinking we agree that selecting IE only wasn't a conscious subsidy, but do you agree that specifying IE only would constitute a de facto endorsement of MS technology?

No, I think, but that depends on what you mean by endorsement. Is it an endorsement (or rather acceptance) of Microsoft's dominance in the PC market, or is it a government "thumbs up" to Microsoft's software? I lean towards the first one.
posted by SweetJesus at 10:59 PM on September 27, 2005


Well, since this discussion is moving in a different direction, let's keep going as long as we want, eh?

E 6 SP1 runs on Windows 98 and above, which is the most popular operating system family in the world for home and business use. Most OEM computers sold in the US come with Windows and IE pre-installed, and it stands to reason that most donated computers would be running Windows.

Let's say I accept that 10% is an acceptable rate of failure. (I think that's terrible for a mission critical app, but nevermind that for now.) And let's say that the total marketshare for non-IE6 browsers is 10%. We then need to look at the share in the target demographic. You conjecture that "donated" microcomputers are most likely to be MS-OS, Win98 or above. Let's go ahead and stipulate that as 100% true. We still have to look at the market share in places where people are liable to be accessing the net to file FEMA applications.

Now, all of this is angels on a pinhead, to be sure, but my suspicion is that a large proportion of those applications (if not most of them) are going to be filed from private homes, public schools, churches and colleges. Forget about Linux: Think Mac. Those are all strong Apple markets. Apple's market share is increasing, and nobody using a Mac that's less than four years old uses IE. (And IE on mac is a nightmare, anyway.) That would mean that the failure rate is something over 10% in the target demo.

.... but the software I write only runs on Linux, so it could be said that I ignore about 95% of the audience, but that wouldn't really be true.

Right: Because your software is either running on the server, or being used by a very specialized group of people. The backend is irrelevant to this part of the discussion, basically. What's relevant is the front-end. If this app is developed as a purely internal app, there are then arguments for using modal dialogs (which again, AFAICS, is the only reason they've restricted the app). More typically, though, modal dialogs in IE are used as a means of routing around data validation and proper action-confirmation steps. Since I don't have the latitude of testing this application, I can't know for sure that's what they were doing here, but my gut tells me that's what I'd find out if I did.

Also, I can be 90% certain based on examining the code that the people making this didn't use FrontPage. They could have used the current incarnation of Interdev, but even so, they would have had to add the IE-only stuff by hand -- or at least, by conscious choice.

As for what visual editors are used in gov't, I don't know, but on the Windows side there are plenty that smooth the path to creating complex HTML without enforcing complex, browser-sensitive layouts. They range from very dear and very complex (e.g. Dreamweaver, Interdev, Fusion) to free (1stPage has a nice array of templates and wizards). Again, based on my examination of the code, rendering was never the problem: They used a third-party library for the menu system, which probably required hand-integration and customization (indicating an at least intermediate level HTML hand). Those libraries were JavaScript, not VBScript, and so they were probably cross-browser. It seemed to work fine in Opera.

On scalability: But I would think something like that would be a backend issue, something to do with their choice of database or web server. I think that's an independent issue.

Well, that's part of the point. It's not part of the browser discussion, but it becomes part of the whole discussion when it comes up. This app was badly designed and implemented, and we know that extends to at least two domains: The client, and the server. It's my contention, also, that the particular ways that it's bad are things that any tyro should have been able to recognize, and that any competent peer review would have caught.

... I would be interested to know if this was an inside FEMA job, or was sub-contracted out. I would be very, very interested in knowing the name of the sub-contractor if the app wasn't internal to FEMA, as there are many companies out there who take contracts and deliver shit.

No argument. I've cleaned up after a lot of outsourced shit. But we can still fault FEMA for not examinging what they got. There is always an element of caveat emptor in outsourcing. You don't always get what you pay for unless you make sure that you check the bag before you leave the store.

If it was internal to FEMA, I imagine it's shoddiness is a result of FEMA's disarray over the past 5 year.

I'm always fascinated by organizational culture. I don't think we'll ever know this, though, unless some career FEMA types defect and write books or get chairs at some school of emergency management. (Of course, Brownie himself is all set to blame it on any Democrat he can get into his gunsights...)

.... Is it an endorsement (or rather acceptance) of Microsoft's dominance in the PC market, or is it a government "thumbs up" to Microsoft's software? I lean towards the first one.

Actually, so do I; but I suspect we differ on whether that distinction matters much. I'm not interested, really, in the morality of that particular issue. I'm interested in what it does to the market ecology (particularly with reference to safety, but also with reference to disenfranchisement and discouraging competition) to have a broad tacit endorsement for one vendor.
posted by lodurr at 6:27 AM on September 28, 2005


It's as if they used pink text on a blood-red background, and SweetJesus is arguing "well, maybe that's the default colours in whatever shitty HTML editor they used."
posted by sfenders at 10:00 AM on September 28, 2005


It's as if they used pink text on a blood-red background, and SweetJesus is arguing "well, maybe that's the default colours in whatever shitty HTML editor they used."

I think that's a silly argument. Whatever problems the total application suffered were on the database or server-side scripting issues. HTML is not a programming language, as such, it's an interpreted descriptive language. You can't crash a browser through bad HTML (ok, so you can if you're using client side scripting language in addition to the HTML, but that's a different story), you can only fuck up the browser's interpretation of the HTML code.

The code not working in firefox is not a problem of the "business logic" of the software, it's a problem of support for a display device. It's as if you choose to support Nvidia and ATI graphics cards, but made a conscious choice not to support 3Dfx. Separate the presentation from the logic, and you see your analogy doesn't really fit.

Lodurr, I'm enjoying this conversation and I'll answer your questions sometime later tonight.
posted by SweetJesus at 10:26 AM on September 28, 2005


You can't crash a browser through bad HTML...

Wanna bet? I can't tell you how many times I've manage to do exactly that -- most recently it was Safari and the offending code had something to do with applying a style to a container that Safari didn't want to see that style applied to. But I used to find lots of fun ways to crash IE for the Mac and NN4.4 on Windows, no JavaScript required.

And let's not reduce this to Firefox. The code explicitly excludes everything except IE6. Right now, the most significant proportion of other users is using Firefox (or some other gecko-based browser), but significant (albeit unknowable*) proportions of the excluded population are also Safari, Opera, or a non-compliant version of IE (like IE for Mac or for Mobile Windows/WinCE).

--
* unknowable, because of user agent spoofing. Unless you perform an object-detection, for example, you won't be able to tell from the server that i'm using Opera for Mac, right now. Which is another argument for designing your code in such a way that it doesn't matter what the client is.

posted by lodurr at 10:46 AM on September 28, 2005


Wanna bet? I can't tell you how many times I've manage to do exactly that -- most recently it was Safari and the offending code had something to do with applying a style to a container that Safari didn't want to see that style applied to.

Ok, maybe I was being too general. You can crash a browser with bad HTML, but it's not the HTML's fault, it's a bug in the browser's ability to display the HTML or render CSS correctly. You don't compile HTML, you parse it. Any errors that crash the browser are parsing errors internal to the browser, not a problem with the HTML.

But in your example, you're not just using HTML, you're using a combination of DHTML, which is strapped to javascipt, and CSS. It's easy to crash a browser with poorly written javascript code (just code a three line infinite loop) because the JavaScript can generate both compile errors and run-time errors. But it's very hard to crash a browser with just mangled HTML these days.

And let's not reduce this to Firefox. The code explicitly excludes everything except IE6. Right now, the most significant proportion of other users is using Firefox (or some other gecko-based browser), but significant (albeit unknowable*) proportions of the excluded population are also Safari, Opera, or a non-compliant version of IE (like IE for Mac or for Mobile Windows/WinCE).

Why would you write an app like this for windows CE? Why would you even waste your time if your hard work is going to be used by .2% of your audience. This is why I don't give a shit about coding for WAP, because I don't think it's useful for situations such as this. Also, I'm using Firefox as a generic example as it's the most popular "other" browser, so when I mention it you can substitute browser X in for it.

I'll agree with you that isn't smart to force the browser to exclude non IE browsers. I'm a bigger fan of allowing anything to access my work, and just letting the code fail if it fails, rather than excluding all cases based off USER_AGENT strings.
posted by SweetJesus at 11:26 AM on September 28, 2005


The code not working in firefox is not a problem of the "business logic" of the software,

Well yes, that's sort of my point there. The client-side of this kind of web app is not some fancy sophisticated software project - it's a stupid web page. Web pages are not hard. There is absolutely no good reason to make it browser-specific. If they were doing something that actually required some fancy javascript trickery, that would be another matter. As far as I could tell, it's just a bunch of relatively simple forms.

The fact that "it's very hard to crash a browser with just mangled HTML these days" just makes it that much harder to find ways to make your website browser-specific.
posted by sfenders at 12:07 PM on September 28, 2005


I was being pedantic about teh HTML and browser crashes, and I'll be this much more pedantic: I've also managed to do it to Safari, and it's got nothing to do with JavaScript. The only browser (that I'm aware of) where CSS was inherently tied to JavaScript was NN4. (Which let you do some nifty detection for problematic CSS, but that's another story.) I'm just talking about linked stylesheets. But yes, it's a total fluke when it happens, and it generally goes away with the next update to the browser.

And my points would end up in synch with sfenders' point: There's no good reason to make the forms anythng but dumb forms....UNLESS: You don't know how to do things like handle sessions to prevent multiple page submissions. In that case, you might do something like set up detailed confirmation dialogs using JavaScript. Since those only work in IE, you'd be left open to failure in other browsers -- hence, blocking other browsers from using the site.
posted by lodurr at 12:37 PM on September 28, 2005


« Older For the Birds and Bird Lovers   |   Sephardim Newer »


This thread has been archived and is closed to new comments



Post