That is a very valid response to David’s comment. Some don’t quite understand that “popular” is a relative term. In this case it is relative to the website in question (i.e. “yahoo.com”). However, you lost me when you said “Our data.” Would you mind publishing some of that data with a brief explanation of how you collected it? Most of us could use that same procedure for our own websites to determine where to allocate our QA resources!

Thanks,
BrandonZ

First, I’d like to say that your (and Yahoo’s) efforts are greatly appreciated. However, your reply to the concerns about Linux and Chrome basically amounted to a long-winded, rambling non-answer.

Nobody asked about X-Grade support, nobody asked what A-Grade means in terms of Yahoo’s internal QA costs and mechanics, and nobody asked about the justification for ignoring those other more obscure browsers.

Specifically, the real issue at stake is the reason why the relatively common Linux and Chrome, the latter of which is proven to be more popular than other platforms on your A-Grade list, have thus far been excluded from said list. In other words, what exactly are the secret criteria that cause Opera to gain more importance than Chrome (or Firefox on Linux) despite fewer people using it?

Thanks.

In fact, the testing that Yahoo performs on YUI is one of the main reasons why I use it.

As I’ve written repeatedly in the past, we do not “ignore Linux.” Nor do we ignore Chrome, Camino, Flock, SeaMonkey, Maxthon, Bento, IceWeasel, K-Meleon, iCab, Konqueror, Shiira and the many-thousand others that visit Yahoo! every day.

In fact, the spark for the GBS framework was an urgent desire to support all users.

Browsers not on the A-grade list—nor identified as incapable—are provided X-grade support. The only practical difference between A- and X-grade is that QA tests A-grade browsers rigorously. In nearly every case (approaching 100% for direct relatives of A-grade browsers, e.g. Firefox 3 on Linux), the end-user experience for A-grade and X-grade support under GBS is identical.

About Chrome:

Our data does not support your claim about its relative popularity. But that’s barely relevant because the true answer is the same as above: X-grade support does not equate exclusion.

Why aren’t all great browser on the A-grade list? Because the list does not grade quality. It grades support. QA actively tests against A-grade browsers. That’s the difference. Even Chrome itself advises web developers that “If you’ve tested your website with Safari 3.1 then your site should already work well on Google Chrome.”

Of course there are differences between browsers. Undoubtedly, Firefox/Linux differs from Firefox/Win differs from Flock/Win, even though they share the same internals. That’s why we test. But testing is expensive and so we use GBS to align our testing resources to (a) the technical landscape and (b) users’ actual behavior.

To be crystal clear, Chrome’s home at Google is not a factor.

I believe sentiments like those expressed in the comments above are the same sentiments that I myself hold dear: a desire that all users are supported (because at the end of the day we’re in the business of supporting users, not browsers!). That is precisely why I developed the GBS strategy; GBS clarifies that support is not binary, but graduated, and that it’s unwise to exclude. Yes, it’s difficult and imprecise deciding how to deploy QA resources — which is really what the A-grade label is. There are countless factors and endless data.

I will continue to monitor the landscape carefully and welcome input and data from everyone.

Thanks,
Nate