@drfloob Thanks for the report and a screenshot, we will try to figure out why this happening.

Test example:

We could just rename grunt to dorkq to fix this problem, no?

@shama, no we need to rename to juxqw

@vladikoff aww, never mind then. I dont want to type that.

The issue is

-webkit-text-stroke: 4px;

fixing...

Fixed. Motorola Droid 4 users are safe from the evil dorkq!

After looking at 858d22a, it looks like the problem was actually the font (makes more sense than it being text-stroke)?

That is awesome, lol...

@zackbloom I think I had to choose to either keep text-stroke or keep the h1

@vladikoff You wouldn’t happen to have a reduced testcase or a minimal changeset? :)

Yea, I'd love an explanation on what the actual bug was.

It's some combination of text-stroke being inside of an h1. So you have to choose either have an h1 or text-stroke. I tried making an isolated test case but can't reproduce, I don't have any of the devices that had this problem.

(Really, just wanted to comment on epic issue)

There is no Dana, only dorkq.

I am sad that dorkq is taken on Twitter :(

Sent from my iPhone

On 29.10.2013, at 12:52, Rob Loach notifications@github.com wrote:

There is no Dana, only dorkq.

—
Reply to this email directly or view it on GitHub.

Is that a font issue? The string "dorkq" is not random:

d - e - f - g
o - p - q - r
r - s - t - u
k - l - m - n
q - r - s - t

@Boldewyn nice find!

And if we sum it up with random numbers... the result is 42!

For anyone interested, I've made a reduced test case here http://jsbin.com/aMuzAn/7

Browserstack screenshots: http://www.browserstack.com/screenshots/42f41255e13b5dce1cbdcae58097900eedf9c2ee

@Boldewyn @kahlil That correlation was mentioned in the original report. More likely I think they are off by 29: GHIJKLMNOPQRSTUVWXYZ[]^_`abcd.

Can bug reports on GitHub be "archived" or "frozen" for posterity purposes? I think this one seems mature and—somewhat hypocritically—I don't believe it needs more outside input to achieve its maximal historical interest and hilarity.

Well, someone had to do this https://npmjs.org/package/dorkq

@andrewsf D’oh! Thanks for the reminder, that one should read bug reports thoroughly!

For the "freezing": I’d love to see one last follow-up comment of the glorious person, that files the upstream bug with the Android browser, sharing the bug URL.

This is crazy

LOL I just saw this

Please make a tshirt parody of this bug. It would sell like crazy

I second the T-shirt idea

Just saw this, but I will buy a dorkq shirt for sure.

👍 For the tshirt

I'd seriously still like to know what happened here, because I know my OpenType, and unless the font somehow has GSUB rules specific to running a Caesar cypher, that is not a thing that will spontaneously happen.

There was a bug in Java where fonts were cached in way that used Glyph IDs, so when a font was updated and the GID order changed, you'd see a caesar cypher effect. Its possible there is such a bug in Android.

That's quite the bug. Would you happen to know the bug link for that on the java side?

https://bugs.openjdk.java.net/browse/JDK-8051892

https://bugs.openjdk.java.net/browse/JDK-7184798

Tee spring is a YC company that does exactly this. https://teespring.com

Grunt at web dev 🤣