WebKit Bugzilla
New
Browse
Log In
×
Sign in with GitHub
or
Remember my login
Create Account
·
Forgot Password
Forgotten password account recovery
RESOLVED INVALID
14291
Webkit applications send 7F/127 characters randomly.
https://bugs.webkit.org/show_bug.cgi?id=14291
Summary
Webkit applications send 7F/127 characters randomly.
Maolf
Reported
2007-06-21 13:05:48 PDT
I honestly don't know if this is a webkit problem or not, but it's insane, and has been driving me insane for about a month. Safari, Colloquy, and Adium all seem to share webkit in common, so I hope it's the problem. Things entered in input boxes occasionally have one or more 7F/127 characters inserted in them. It seems OS X for the most part will never render them, so they are invisible to me, but patrons on the receiving end will see hex boxes or ?'s when I'm speaking to them on Adium and Colloquy, in Safari I'll often see %7F in the URL after I do searches on Google and stuff. Here's what it looks like on the receiving end:
http://www.jpgclan.com/floamfailchars.png
What's even weirder, is I've done clean installs of OS X (granted, with likely some "stuff" moved over from previous ones), upgrades to Safari 3 beta, and it even occurs in 10.5 9a466! I'm going nuts here, certainly if I encounter it in all these situations, I can't be the only one!
Attachments
Add attachment
proposed patch, testcase, etc.
Maolf
Comment 1
2007-06-21 13:07:23 PDT
I'm using ASCII/UTF-8 everywhere, with regular US/English settings and no strange input setups.
Mark Rowe (bdash)
Comment 2
2007-06-21 15:08:49 PDT
0x7f is the UTF-8 representation of the "delete" codepoint. Text input in Colloquy and Adium is handled via standard Cocoa controls, not WebViews, so it seems unlikely that this is an issue with WebKit.
Maolf
Comment 3
2007-06-21 15:13:18 PDT
Okay. Any ideas what I should do next? Should I mail cocoa-dev, or what?
Maolf
Comment 4
2007-06-21 17:04:22 PDT
I tracked it down. It occurs when I hold shift + arrow up + delete at the same time. I don't know why, and I'd love to figure out how to kill it. It occurs apparently when I type a line, shift + arrow up to highlight it all, and then delete it and retype a new line.
Alexey Proskuryakov
Comment 5
2007-06-22 06:14:47 PDT
FWIW, I cannot reproduce this by typing Shift+Up Arrow+Delete in Safari's Google search.
Maolf
Comment 6
2007-06-22 08:15:59 PDT
Weird. You won't be able to see the character, but at least on both of my machines here, if you type shift+up+del followed by "asd" and hit enter, to your behest you should see ?q=%7Fasd in URL bar.
Maolf
Comment 7
2007-06-22 08:49:07 PDT
Also, while there are two keys labeled "Delete" on most Mac keyboards, I was referring to the one above return, ie. "backspace".
Dan Hendershot
Comment 8
2007-06-22 09:01:45 PDT
Aha! I'd just like to say I get the exact same thing, although I never did figure out the pattern that would generate these prior to finding this bug report. Likely not a WebKit bug, but it certainly is /someones/ bug. I also notice shift, up, and enter held at the same time will generate an invisible 0x0D (but not an actual linebreak), easy to check using textmate and a hex editor opened to the same file.
Alexey Proskuryakov
Comment 9
2007-06-22 09:21:17 PDT
(In reply to
comment #7
)
> Also, while there are two keys labeled "Delete" on most Mac keyboards, I was > referring to the one above return, ie. "backspace".
Aha, that's it! On my keyboard, only one key is labeled "del", and that's another one :-) I was now able to reproduce this in TextEdit, too, thus closing as INVALID. Please file a bug with Apple at <
http://bugreport.apple.com
> (free ADC Online membership required).
Maolf
Comment 10
2007-06-22 14:53:10 PDT
Okay, it's bug #5288750 over there. Thanks!
David Kilzer (:ddkilzer)
Comment 11
2007-06-23 10:52:22 PDT
(In reply to
comment #10
)
> Okay, it's bug #5288750 over there. > Thanks!
Thank you!
Aaron Gyes
Comment 12
2015-07-06 21:10:30 PDT
For those following along at home, this is reproducible in El Capitan. I'm sure they'll get to it any day now.
http://www.openradar.me/radar?id=5578055959445504
Alexey Proskuryakov
Comment 13
2015-07-06 22:19:33 PDT
> I'm sure they'll get to it any day now.
Hi Aaron, As you are the originator of
rdar://5288750
, you should be able to see it that the bug closed now. I think that this was done erroneously. Would you be willing to e-mail
devbugs@apple.com
, asking them to re-open the bug? If you don't have more time to spend on it, that's understandable, and I can take it from here - however in that case, if would be harder for you to see updates to the original Apple bug report.
Aaron Gyes
Comment 14
2015-07-06 23:57:57 PDT
Sure. I already did go ahead and a posted a reply to the rdar today saying that it's been 7 years, and, you know, the behavior is still present. Their last update seemed to imply it was fixed internally, in 2012. I was doing so hoping that may reopen the closed report as suggested in their documentation (
https://developer.apple.com/bug-reporting/using-bug-reporter/
see "How to reopen a closed report"). Should I also email that address?
Aaron Gyes
Comment 15
2015-07-07 00:20:25 PDT
I went ahead and emailed them anyway.
Aaron Gyes
Comment 16
2015-07-13 00:35:20 PDT
Asked to re-file due to age of the bug. It is now
rdar://21783531
.
Note
You need to
log in
before you can comment on or make changes to this bug.
Top of Page
Format For Printing
XML
Clone This Bug