Hacker News Reader

Comment by bpatrianakos | Hacker News Reader

Front-ends that work in IE6 but not in modern browsers? Are you sure? That doesn't really make sense to me. I can see how maybe some outdated code or a bunch of hacks could make a site designed for IE6 look just a little funny here and there but I don't see how it would be a lot of work at all to make a site made for IE6 work in a modern browser. As far as I know, the only reason large companies don't upgrade is because they have a bunch of infrastructure that's totally unrelated to websites that stop them from upgrading. Basically, the IT department sets up the entire company's computer systems a certain way and then resist upgrading anything so they don't have to do a lot of testing, more patching, and generally adapt to the new tech after they've tuned their infrastructure just so.

View on HN

Large IT departments are really conservative for good reason. Hypothetically, consider that a company maintained a web app that accountants used to track revenue. Now, consider that the accountants needed to input interest earned from investments. Let's say there was a Regex bug in parsing the interest rate, and all values were interpreted to be NaN or 0 on non-IE6 browsers due to an implementation quirk. This would have disastrous consequences for a large company: they might face shareholder suits, government investigations, etc. And of course, the people maintaining the system would likely lose their jobs.

There are so many downsides, but what's the upside? So that FB loads a little faster? The incentives are stacked against adopting new technology. Any new browser has to be thoroughly tested against every one of the company's apps to make sure that nothing breaks. As anyone who's done QA before will tell you, that's a tall order.

I agree that the probability of such an incident is small. However, given the disastrous consequences that would happen for a large organization should such a bug occur, it's hard to make the case that an upgrade is a good idea for the sake of "better technology."

The upside? Javascript will run 100 times faster and your developers can build even more functional software, moving software that once had to run on the desktop to software that now runs in a browser.

That doesn't change that IE is still needed for those old applications. So the best case scenario is running whatever new browser like Firefox, Chrome, etc, and running IE6 alongside. Ditch IE6 only after the old apps have been replaced completely.

I'm a little confused. Why is there IE6 lockin? Apps weren't built so they could at least be easily upgraded to IE8? Some testing is required and no one wants to do the work?

My memory isn't that hazy. What happened was that people said the software only had to run in IE6 and Netscape wasn't important. Writing "good" web software wasn't important. Simply support Microsoft's IE6 and you're done. If you're still stuck on IE6/IE7 it really is your own fault.

michael_miller explained it adequately above:

1) IT departments are massively conservative (and this is not an intrinsically bad thing)

2) A lot of apps were built specifically for IE6. They still work with IE6. Many times they only work with IE6.

3) It hasn't been made worthwhile to upgrade.

> If you're still stuck on IE6/IE7 it really is your own fault.

A lot of us web developers have decided to just say, "Fuck it," and stop supporting IE6. If it works, cool; if it doesn't, we don't care. This is a luxury available to those of us who have clients who aren't supported by the aforementioned IT departments. Some of us are still required to support IE6 because we have clients who we can't drop and refuse to move off, and we don't have the leverage to persuade them otherwise.

For some companies it doesn't make sense to drop support for IE6. I work for a company who's product is sold completely online. If we dropped support for IE6 we would loose hundreds of thousands (if not millions) of dollars a year. Sure it's a tiny fraction of our business, but it's a considerable amount of money and pretty much covers the wages of our entire web development team.

The problem is JavaScript not (only) the CSS. For instance, "document.getElementById" in IE6 targets the name attribute, so those web sites can function without setting an id attribute. The IRS web application of my country does that.

We threw out a firewall appliance that had the admin page only able to run in IE6 (I think it did an explicit check for browser verion) Didn't replace it till we got tired of the P4/XP/IE6 box in the corner. We are tiny (<10ppl) and I can imagine tons of others in similar situations. The 90's was not a good time for browser code bases.

Here's an example of something that I do not believe would work in modern browsers:

http://msdn.microsoft.com/en-us/library/ms970405.aspx

If your entire application was created like that then you are talking about a significant overhaul.

Then you've never seen it happen. I've seen it happen lots. It's horrible. Absolutely horrible. :) I won't pretend I knew why those apps broke in IE6.

ActiveX

Reply on news.ycombinator.com