What purpose does browser-specific blocking serve?

[IainB]

Here's a puzzle that I don't quite understand yet:
At the blog on Paydirt they have a post:
We don't support Internet Explorer, and we're calling that a feature
- and this has been discussed at Hacker News - here.
- where they also give a link to a very interesting post about this by one Rey Bango: Hey Paydirt: Your Site Works Just Fine in IE

I can't quite understand the rationale here, which seems to be:

• We have spent and are spending way too much time/cost tweaking and maintaining our website to work around IE browser-specific idiosyncrasies.
• Therefore:
• (a) we're not going to spend that time/cost doing that work any more.
• (b) we're going to block IE browsers. (This looks like a non sequitur - a classic logical fallacy.)
• [Unspoken] This despite the fact that our website apparently works fine with current IE browser versions. (This would seem to defeat/contradict the initial statement.)

Trying to make sense of this, I went to the Paydirt website home page, where it has the slogan:

Honest-to-goodness time tracking and invoicing for browser-based freelancers, consultants and small teams.

- which probably does not say quite what they intended it to say.

Underneath that, near the bottom of the page, is mentioned:

• SSL Security
• Chrome Extension
• Firefox Extension

I would deduce from that that the real issue could possibly be related to the work required to build in SSL security and/or develop an extension that suits the IE browser, but if that is the case, then I do not see why they don't say so, rather than go through the seemingly absurd rationale (above).

This all leads me to questions:
What might be the point behind announcing this seemingly irrational blocking (or non-support) of their potential customers who might use IE browsers? Couldn't that effectively block 50% (or more?) of their potential market? (I don't know what market share the different browsers have.) Why limit themselves by turning their backs on that potential market?

This looks like it could be a case of "cutting off one's nose to spite one's face".
Is there some valid technical reason for this?
What purpose does browser-specific blocking serve?
Is it a business purpose or a technical purpose, or a combination of both?

[zridling]

In 2012, I'd almost think that was a joke from the 90s, when the day was the opposite -- you had to use IE for quite a number of commercial (or corporate) websites to work based on Microsoft's proprietary APIs and funky way of writing HTML3/4. But those days are long, long gone. Hell, the latest IE is a thousand times better than the arch-evil IE6. But bad reputations are hard to live down -- just ask Google right after they say, "Anything else, officer?" when they hand over your data to the authorities.

[db90h]

I can reason not supporting older versions of IE because they had such a problem with rogue extensions (Browser Helper Objects and Toolbars) *and* render content differently than other browsers, including newer versions of IE. Blocking newer versions of IE that render content the same as any other browser and that have essentially 'fixed' a lot of the extension problem by making users aware of what extensions are present is nonsensical though.

Also note that NOT ALL SSL ROOT CERTIFICATE AUTHORITIES ARE PRESENT IN ALL BROWSERS. The newer CA efforts for cheap or sometimes free SSL certs, for instance, certainly are not. The root CA not being present in IE may be the issue here. Without it, every SSL protected page would cause a huge warning. UPDATE: Not the case, at least not with latest IE.

[Renegade]

See here:

https://paydirtapp.com/ie

That's when you click to sign up.

IE is <insert tirade of profanity here /> miserable to code for. Still. And you need to suffer that torture for v6, 7, 8, and 9. 4x the pain.

I very well understand what they're saying --- we do not have ungodly amounts of time to waste on making things work for IE. IE is broken. Use something else.

The days of IE insanity are only worse now than before. Seriously.

[barney]

Wel-l-l ... yes and no  .

I've seldom had any problem(s) with Websites because of a particular browser.  But, then, I wasn't trying to do anything outre, Web-wise. 

If a site is designed for a particular browser, coded such that it only displays properly in one (1) or two (2) browsers, that burden/failing falls upon the Website developer, not upon the browser creator.  And the site owner should be cognizant of and forceful about inadequate design.

Yes, all browsers should adhere to all standards, but they don't  .  If a developer is too damned lazy [or ignorant] to make the site work in the four (4) or five (5) major browsers, then the site owner is the one (1) who gets hurt. 

Adding a disclaimer to the effect that a site only works in {ie|firefox|chrome|opera|safari|add your browser here} is, to my mind, just not being willing to put forth the effort to make it fully functional.  Yeah, it's extra work, but that's what developers get paid to do  .  There's most always a workaround that will fit that particular browser's engine  .

Now, another aspect is that since IE6 attained such a heinous reputation, it's sometimes considered a plus to publicly state that it is not supported.  I don't agree with that position, but 'tis not rare  .  'Course, all this may be moot considering the current controversy over the forthcoming -webkit debacle  .

[db90h]

Yes, but IE users make up too much of a market share. That's why I indicated to block older versions of IE makes some sense, but newer versions -- that's killing too much of your user base.

Anyway, I speculated about other causes because the OP suggested this was not their true reason.

EDIT: I had concluded that their SSL cert was signed by an untrusted root authority for IE, but am re-evaluating that for older versions. I posted too hastily. Turns out I was wrong, had my adrenaline pumping ;p. The root CA is a branch of Comodo, who is certainly trusted, but I don't know whether that branch is.

[db90h]

Fixed oopsie on my premature conclusion, although that may be accurate for *older* versions of IE. Also, this site seems to work fine for the latest version of IE, even /w SSL. They may simply be saying they aren't going to try to work around the issues with all the older builds of IE. No ulterior motive... entirely possible. Maybe this policy hasn't went into effect yet, I dunno.

Indeed, I myself quit caring about older IE versions too. Too much of a pain to create multiple CSS files, etc.. That said, my site *works* under those older versions, I did make sure of that.

[Renegade]

If a site is designed for a particular browser, coded such that it only displays properly in one (1) or two (2) browsers, that burden/failing falls upon the Website developer, not upon the browser creator.  And the site owner should be cognizant of and forceful about inadequate design.

Yes, all browsers should adhere to all standards, but they don't  .  If a developer is too damned lazy [or ignorant] to make the site work in the four (4) or five (5) major browsers, then the site owner is the one (1) who gets hurt. 

Adding a disclaimer to the effect that a site only works in {ie|firefox|chrome|opera|safari|add your browser here} is, to my mind, just not being willing to put forth the effort to make it fully functional.  Yeah, it's extra work, but that's what developers get paid to do  .  There's most always a workaround that will fit that particular browser's engine  .

-barney (May 08, 2012, 08:31 PM)

Sorry. Not on board there. And no, there isn't always a work-around. Sometimes you are forced to have entirely different behaviour for different browsers -- again, IE is the problem here.

One of the core issues is that HTML is a declarative, markup language, and conditional logic is at best completely insane for it. The conditional logic for IE for style sheets is just moronic. Trying to put conditional logic into JavaScript is similarly idiotic. But, you are FORCED to use conditional logic for what would apparently be trivial tasks, e.g. using a transparent PNG as a design element.

Taking that example, transparent PNGs, you end up with an entire world of pain. It's simply insane. Anyone who has ever done any web design knows that the IE filters used to "fix" the problem are voracious in their appetites for CPU. It's extremely easy to create an otherwise simple design that will grind almost any computer to a halt, and will certainly destroy an older computer.

Trying to code for IE is like swimming with a concrete cinderblock chained to your neck, both arms, both legs, and just to add insult to injury, to any other external appendage that you may have.

At some point you simply need to say, "Bugger this. We're not going to support that anymore."

Sure, you *can* support IE... But do you have the time and money to do it? Development costs money. For a primarily developer audience, there's ZERO reason to support IE 6, 7, or 8, and 9 is debatable.

The site there made the right choice to not support IE. Their audience doesn't use IE, so why bother? It's like manufacturing ski jackets for Mexicans.

Now, for general sites aimed at non-developer audiences, that's a different matter... You have to be careful in how you choose which browsers to support.

If you're doing government or corporate work... god... They're often still using IE6... So you're forced to do a massive amount of work so that the people that commissioned the work can see it, even if the audience for the site doesn't use IE6 at all. That's entirely about who's paying the bills... (IE6 was around for a dog's age, and there are a large number of applications built for it that are still in use.)

[barney]

(IE6 was around for a dog's age, and there are a large number of applications built for it that are still in use.)

-Renegade (May 08, 2012, 08:58 PM)

Yeah, that's an albatross tied 'round MS' neck.  There are so many in-house apps built on IE6 that MS has been having Hell's own time retiring it.  Those apps just won't work in a real browser  .

As to the developer/non-developer audience, I don't think choice is really a valid option.  With the inherently catholic audience that is the Web, you pretty much have to design for all. 

On the other hand, stats that I've seen in several different venues indicate the IE6 is less than one (1) percent of the market, so things have improved somewhat.  IE7&8, as you've stated, are still pretty crappy, but IE9 has gotten closer.  And from what I've read - no direct experience - the [putative] browser for Win8 (IE10?) toes the mark about as closely as any other extant browser.

But a bottom line here is that there'll always be browsers that do not hew to the line on standards, or perhaps interpret them differently.  If those browsers are/become popular, we have to design with them in mind.  Or find a different line of endeavour.  Browser bias simply is not viable in the long run, regardless the efforts required to accommodate.

[db90h]

I amended my statement above, as I wanted to clear one thing up .. although I quit caring much about older versions of IE, I do make sure my site *works* in them. It just is not as pretty .

[Renegade]

Browser bias simply is not viable in the long run, regardless the efforts required to accommodate.

-barney (May 08, 2012, 09:41 PM)

That's the thing -- if you are a small business, or have limited resources, and must have a web site, you can either piss away your budget on 1% of the market, or you can do a spectacular job for 99% of the market. (Or whatever the actual percentages are.)

It's like having a restaurant -- just because some people are allergic to some food doesn't mean that you can't serve it. They can go elsewhere. Serve your own customers and be content with that. Not everyone has to love/eat your food. But, the ones that do eat your food, you can certainly work at making their meals delicious.

At some point you MUST say that effort beyond a certain amount is too much. Limited fund. Limited resources. Limited browsers.

Taking your statement a bit further... Why not support Netscape as well? And Netscape versions down to v1 while we're at it. Obviously any site that does that is going to be total crap, or insanely expensive.

It would be better if developers/designers simply stopped supporting IE entirely. If nobody supported IE's insanity, MS would either have to finally come around to playing nicely, or people would have to get a browser that worked.

By forcing browser companies to adhere to standards, the web becomes a better place. Letting them run rampant is wasteful and counterproductive.

[barney]

Browser bias simply is not viable in the long run, regardless the efforts required to accommodate.

-barney (May 08, 2012, 09:41 PM)

That's the thing -- if you are a small business, or have limited resources, and must have a web site, you can either piss away your budget on 1% of the market, or you can do a spectacular job for 99% of the market. (Or whatever the actual percentages are.)

-Renegade (May 08, 2012, 10:21 PM)

OK, it seems there are two (2) different arguments here. 

From the development side, my argument stands.  It is the developer's job to develop for as many venues as possible/practicable.

From the business owner's side, it is not only feasible, but required, to try to attract the largest possible audience.

I don't see a real difference there.  (And I've been in both positions  .)

So we're basically talking apples 'n oranges in approach  .

However, I'm approaching this from the technical, not financial, standpoint. 

As developer, it is my job to design for the widest possible selection of browsers in order to provide the widest possible coverage for the Website owner.

As a site owner, I would, of course, want to pay for the minimum amount of developer time.

I don't see that the arguments on either side of this question are contrary.

I also think we've wandered far afield from IanB's original post, so I'll shut up now.

[Renegade]

@Barney - Got it! I know what you mean. I'm all for trying to get as much support for as many browsers as possible, but... limitations on time, etc. can pose problems. I coded the design for one site and had to hack through some IE issues. Wasn't optimal, but that's IE for ya!

[IainB]

It is odd though, isn't it?
Initially, I thought that maybe they couldn't cost-justify building an IE extension and reckoned that they could get away with not doing it, because their target market wouldn't be IE-dependent - as Renegade says:

Their audience doesn't use IE, so why bother?

But, that's not necessarily a safe assumption, as @db90h says:

...that's killing too much of your user base.

- and as @barney says:

With the inherently catholic audience that is the Web, you pretty much have to design for all.

If it was being purely business-driven, then presumably they would want to spread the net as wide as possible, but they don't seem to be doing that and are even making a "thing" out of not doing it, so it could be that there is a technical issue that is taking priority over the business driver.
It doesn't seem to be their website - that apparently works fine for current IE. Maybe there is some coding in those extensions that makes it a nightmare to develop one to work in IE?
Odd.

[TaoPhoenix]

I'll throw some stuff on all this from another angle.  Y'all are looking at this from a "nice innocent" angle, both on the business side and the developer side. "Let's develop for all the browsers!" "The widest reach possible is best for business".

But what if that's only "Level 1" type strategy? "Build the website for all the browsers and they will come."

What if you *don't* need to develop for all the browsers? This is the exact same argument over Flash on iPads. "The widest reach" is of course to put Flash on iPads. "It's the developer's job to make it work" etc etc.

But Apple made a "Level 2" strategy move not to. They wanted to rearrange the ecosystem.

So here's a snarky Level 2 strategy: Look at it from the Support Side. Maybe you don't want a customer base filled with the type of people who post comments on Youtube.  So anyone who can't read enough to know what the words browser and IE are, isn't going to read anything about your app either! "If you can't download Firefox, you're not worth my time to support on the help desk."

Edit: I'll say it even stronger. "Honest-to-goodness time tracking and invoicing for
browser-based freelancers, consultants..."

You're a *consultant*. This isn't a Zynga Facebook app. It's a tool for *working knowledge advice professionals*. If you as a consultant can't download another browser, then I wouldn't hire you, and if "the CEO did" then put me in a padded room now. I can't even fathom a conversation to a consultant that goes like this:
"Hmm. I can't use this website."
"Oh, that's okay, you have to use a different browser than IE."
"What's a browser? What's IE?"

[IainB]

Some more light on this: Supporting IE Is Too Much Work

Seems like the developers might possibly be having difficulty with supporting IE because...they are developing on a Mac. Now that at least could make some sense.

[IainB]

Here is an example where they are not blocking IE, but openly suggesting that you upgrade from the older version IE6:
From: Meet The Team | 6Wunderkinder

Marvin may look like the boy living next door, but really, this guy is a total machine! He is one hell of a front end developer and, as the coffee cups on the picture indicate, he works at lightning speed. Thanks to him, you can look at this page even with your old IE6 browser. However, he really suggests that you should upgrade!

Needing maintain backwards compatibility with older versions of IE (or any other browser?) would presumably be a widespread issue.