Slashdot Mirror
Apple Updates All of Its Operating Systems To Fix App-crashing Bug (engadget.com)
It took a few days, but Apple already has a fix out for a bug that caused crashes on each of its platforms. From a report: The company pushed new versions of iOS, macOS and watchOS to fix the issue, which was caused when someone pasted in or received a single Indian-language character in select communications apps -- most notably in iMessages, Safari and the app store. Using a specific character in the Telugu language native to India was enough to crash a variety of chat apps, including iMessage, WhatsApp, Twitter, Facebook Messenger, Gmail and Outlook, though Telegram and Skype were seemingly immune.
I applaud your swift response patching this issue. Downloading the update now, since I don't have an Android and actually get updates for my 3+ year old phone.
It used to be just Little-Endian and Big-Endian, now it's also Single-Indian. Thanks Apple!
It really is time to replace Unicode with something more robust. These errors due to things like combinational characters and tricks like using the text flow control characters to mask file extensions keep coming up.
Programmers aren't language experts, there are no good libraries for handling Unicode, can't even agree on one sane encoding for it... And it's so bad that it's avoided in east Asia for the most part, or just some incompatible subset is used.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
PS. I'm thinking of starting work on this, but can't think of a good name. Any suggestions?
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
Hey, putting out alpha versions of software in major commercial releases, using us all as guinea pigs and then saying 'its okay we put out a patch!'
It is the american way ...
Which is why most people would rather fuck a vacuum cleaner full of razors than buy anything made in america
That's what you get for hiring H1Bs
My first Apple purchase was a Mac Mini with an Intel Core Duo CPU, 160Gb laptop HDD and (I think) 4Gb RAM. Since then I've purchased various other Apple devices (including iPods, iPads and an iPhone 7). According to my records, "Apple spend" has made up about 31% of my total hardware budget since the Mini.
One thing I've noticed change at Apple over that period of time is that, since the passing of Steve Jobs, there has been a slow but steady decline in quality and reliability from Apple products. That's not to say that they were immune before he left us, just that there appears, subjectively, to be deterioration in QA over at Apple.
I write this not as an Apple Fanboi nor an Apple Basher: my current iPad (Pro, 10") is probably the most-used piece of technology I've ever owned, but on the other hand last weekend saw me swearing in disbelief at my Mac Mini : having gone to it to update my iPhone and iPad software, I discovered that, somehow, iTunes had decided to unilaterally "lose" the artwork for about 20-25% of my music collection. Of 900 albums. I've already spent a good 90 minutes trying to repair that damage and have a *long* way to go yet...
My experience to date has been that when I made my first Apple purchase, the company had a reputation for high prices but excellent quality. Today, the high prices remain but the quality appears to be disappearing rapidly. Issues with iTunes Artwork, iCloud replication, corruption of the iOS Address Book, a Mac Mini update that bricked the machine, iTunes that can't cope with it's media database on a network-connected drive; the list goes on - and that's just since Christmas 2017...
Apple really needs to get back to basics. If it can't sell reliable product, then no matter how shiny it is, people won't buy.
I'm thinking of starting work on this, but can't think of a good name.
How about NIHcode.
Unicode has nothing to do with the encoding used. Just use UCS-4/UTF-32 if you don't like encoding.
Everything I write is lies, read between the lines.
The problem is, text is hard. The rules for text make no sense. Western text is easy - we're used to it, and have a generally controllable amount of characters. We can choose to encode it as individual letters (so accented characters are stored as individual characters) because there are a limited number of them.
But other cultures, not so much. Arabic can be hard and most are decorations that affect a base character. Plus, character pairings don't make sense - adding a character can make the entire word being displayed shorter and more compact than without that character (instead of longer).
It's bad enough that people keep wondering when /. will support Unicode. Internally, it already does, and has for over a decade (and probably since the turn of the millennium). Problem was, people realized the potential for chaos and trolls spent absurd amounts of time crafting Unicode text bombs that would cause the comment section to be displayed incorrectly or overwritten by characters that were thousands of pixels tall and unreadable. In the end it got so bad the only solution was a approved character whitelist - the only accepted characters for comments were on a whitelist, and basically was what you could represent n ASCII. Eventually they added a display filter that killed the crap comments in affected articles as well so the archives were usable.
Unicode is composed of codepoints. A character may be composed of one or more codepoints. Trolls have managed to generate characters that are composed of thousands of codepoints (imagine using 10kB of data which represents one character - how will you program that?).
Of course, I suppose it disappoints lots of people who were hoping to embed the character everywhere to crash iOS devices...
There is one already: '~'.
Could you be more specific? Except for the reverse file name problem, which is unique to windows, you made no point. ... and an asian automatically uses unicode on his android or iOS device, what else would he? Same for Window, Linux, Macs, switch the language to Thai or Japanese and you have everything in Unicode, heck plain english is unicode, too.
All modern languages and OSes have unicode built in. All is working fine
Cost free eBook I read (by iBook/Kobo/Amazon/ObookO/Gutenberg etc.): "The Green Odyssey" by Philip Jose Farmer.
Wish granted.
What about the 9681 other less publicized yet very annoying iOS bugs?
Slashdot, fix the reply notifications... You won't get away with it...
What about the 9753 other less publicized yet very annoying iOS bugs?
Slashdot, fix the reply notifications... You won't get away with it...
An amazing opportunity for writers in India to earn a decent living through freelancing! Freelancing Online writing jobs has now become the most preferred job profile and you can attain a prestigious preference from clients all over the world. If writing is what you excel in, do not think much and become a freelance writer at the earliest. Invest your skills with NerdyTurtlez and get paid at will.
Just so you know: UCS-4 and UTF-32 are not the same thing. You can hand in your Unicode Coders' Union card now.
Except for the reverse file name problem, which is unique to windows, you made no point.
If you're referring to RLO spoofing then Linux and macOS are not immune to that problem. And most web browsers probably aren't immune to it either (see the Russian "paypal" example in the same link).
Thank you, Unicode Consortium, for building goddam security vulnerabilities into a character set!
Well, maybe hand in your English comprehension/troll/geek card instead, I am not sure which one you may have.
What Does / Mean Between Words?
To Indicate Or
Often, when a slash is used in a formal or informal text, it is meant to indicate the word or. The examples below illustrate this meaning of the forward slash:
When leaving the classroom, the teacher noticed that a student had left his/her backpack.
College freshmen should bring a mattress and/or cot to sleep on during orientation.
If/when Mary ever shows up, we can all head out to the party together.
Burgers or pizza for dinner? Yeah, either/or is fine with me.
https://www.grammarly.com/blog...
Now guess what & between words mean to redeem yourself.
Everything I write is lies, read between the lines.
All of these problems are because Unicode to highly inconsistent and seemingly designed to cause these kinds of bugs and denial-of-service attacks.
Unicode should have focused just on encoding characters. No modifier characters - Unicode uses them inconsistently in some scripts but not others - which would fix the need for an impossibly complex set of crash-proofing, anti-trolling and grammar rules.
On screen layout, font selection, text flow direction and the like should have been provided by supplemental libraries that understand languages. Stuff like the on-screen width of a string is something you need to query the font rendering engine for anyway. Translation of input is done by the IME.
All the meta characters need to go too, like zero width spaces and bidirectional flow encodings. That stuff is metadata.
The app most likely only really cares about handling arbitrary strings (remember programmers don't know anything about languages) without crashing. And if it does care, the developer doesn't want to become an expert in Unicode as well as in the relevant languages.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
No reason why...until now. Kept shutting down, not responding to input, power cycling didn't do anything. Called Apple, and then it started working appropriately...seems like something was being changed...and voila Apple was doing something in the background. Ugh!
AmiMojo's Precise Karacters
I like that Slashdot has become Apple's new radar. If only there was a site I could report linux bugs, which are aplenty.
It has to be posted
https://xkcd.com/927/
hmmmmm....
Bravo. I googled and no-one else is using that.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
It's not that bad if you try to do one thing at a time. If you try to do all things at once and be all things to everybody, then yes, the whole thing quickly becomes unmanageable. Unicode has this as a designed-in goal.
All of these problems are because Unicode to highly inconsistent and seemingly designed to cause these kinds of bugs and denial-of-service attacks.
It's what you get when you "standardise" on the premise that competence doesn't matter since you're a committee now. (Borrowing loosely from Naggum's comments on the W3C and all it produces, all of which have similar problems.)
Unicode should have focused just on encoding characters. No modifier characters - Unicode uses them inconsistently in some scripts but not others - which would fix the need for an impossibly complex set of crash-proofing, anti-trolling and grammar rules.
Unicode doesn't currently encode the full set of data and meta-data for representation. See han-unification, where vital bits of what language is actually being used is now left to the font. This is fun because while the characters might mean the same, they don't look the same across those unified languages. So if you mix languages... you can't use just one font.
And as to whitelisting, it does have "regional variant" codepoints distinct from the main codepoint, but in many fonts those look identical. This besides the codepoints in different languages that look identical, but that you might filter out using some whitelist iff you know what language is being used.
So it sort-of does what you want but then does so inconsistently, and sometimes it is designed such that it plain cannot. Just like it does have codepoints for character-with-accent but then says you really should use two codepoints. IOW they changed their minds halfway through.
On screen layout, font selection, text flow direction and the like should have been provided by supplemental libraries that understand languages.
Then it's still part of unicode, you just make some parts undisplayable despite the software "supporting unicode".
And we already have that situation, actually: Something like fully half glibc code volume is there to "support unicode" but that isn't enough for decent unicode support, you need extra libraries for that. And then you find that some concepts aren't even defined. Not in code, not in libraries, not in standards or annexes, not anywhere. Such as
the concept of cannonicality, which is how I think it was spotify got breached through its use of unicode usernames. Alright, there was something they borrowed from somewhere else only defined for a unicode standard a few versions back, so they didn't have to make up the whole idea from scratch. It still got them breached.
Translation of input is done by the IME.
And what do you suppose it is to produce?
All the meta characters need to go too, like zero width spaces and bidirectional flow encodings. That stuff is metadata.
I never really understood the use of zero width spaces in unicode, except as some sort of attempt to be more correct than just plain correct. Its practical effect is more leeway to troll and a clear need to only use all those wonderful overly complex special unicode string handling functions.
The app most likely only really cares about handling arbitrary strings (remember programmers don't know anything about languages) without crashing. And if it does care, the developer doesn't want to become an expert in Unicode as well as in the relevant languages.
In Unicode, just about by definition, all languages are relevant. The ultimate diversity inclusivity.
Yes, nasty sneer right there. Justified through the diversity debate over fscking smileys, and needing ele
Sounds like it's time for Apple to start using the Big List of Naughty Strings (https://github.com/minimaxir/big-list-of-naughty-strings).
It's pretty new, what with it's first commit only being in 2015 (although the idea of it's been around a lot longer).
Some AC offered:
Wish granted.
The problem is that /. forbids the display of both certain Unicode characters and their HTML-entity counterparts. So, although that's a handy tool for other purposes, it doesn't solve the Slashdot-character-display-is-purposefully-borked problem.
Next ... ?
Check out my novel.
No, US English western is simple and we're used to it. Software has its fair share of problems with non English but western Latin based text as well.
Consider just one example: the IJ digraph in Dutch, It has 2 unicode forms, one for upper case and one for lower case. Yet spellcheckers will often be confused by this and not understand if you use them in a word. Office also has classic issues with this in that it converts them to the separate characters I and J. If you then proceed to write something in Word with auto-language detection enabled, if you start a sentence with that dygraph the first thing it will do is lower case the J for you (not recognising the dygraph) quite helpfully thinking you're writing English, and then after the end the sentence proceed to tell you you misspelled the first word since both I and J should be capital.
Then there's classical linguistic confusion around characters that have alternative forms. Ö in German being written as Oe, the strong ß in German being ss. Even the western languages are hard.
Cut & Paste, All your base are belong to us.
Harrison's Postulate - "For every action there is an equal and opposite criticism"
There was no patch released for El Capitan, nor one for Sierra to the best of my knowledge. Both are still maintained by Apple.
I expect those patches will come later, as did the Meltdown and Spectre fixes.
It’s too bad Apple is such a small shop... they should consider hiring a couple extra folks so they can keep on top of all their supported products. I certainly wouldn’t want something trivial, like development a security patch, to adversely impact the release schedule of the Animoji team.
#DeleteChrome
Do you use it by default?
"UTF" stands for "Unicode Transformation Format".
Unfortunately just using UTF-32 won't really help fix Unicode because a lot of the problems are unrelated to it. The Unicode Consortium's dislike of UTF-32 is actually quite telling.
For example, they say that there isn't much advantage to fixed size code point encodings because with UTF16 and UTF8 you can still work backwards by simply examining 2-3 previous bytes... But what does 95% of software do? Convert UTF8 to UTF16 internally and then assume 2 bytes = 1 character.
It's this kind of thing that makes Unicode software so unreliable and fragile enough that a single character can boot loop your phone.
Unicode is full of similarly terrible ideas, like the Basic Multilingual Plane (BMP). The pitch is that if you support the BMP you are good with 99% of the world's languages, only obscure stuff needs the extended parts. In reality millions of people can't type their name into your software. And all that because >2 bytes per codepoint is barely supported by anything. It's like going back to the 7 bit ASCII days on Usenet.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
On screen layout, font selection, text flow direction and the like should have been provided by supplemental libraries that understand languages.
Then it's still part of unicode, you just make some parts undisplayable despite the software "supporting unicode".
I'm thinking about something like this.
Say you have a range dedicated to Japanese, 0x00010000 to 0x000FFFF. In your font description metadata you can have all the glyph mapping information. Your word processor software will know about bidirectional text and Japanese grammar.
You might have an official table that states that characters in that range are all Japanese language. They are typically rendered left to right, or alternatively top to bottom. It's up to the software to do what it will with that information. At least it can select a reasonable font, unlike now where there is no way of knowing if it should use Japanese or Chinese characters.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
A/UX? System 7? Mac OS 9? Puma?
People should be a little more deliberate when considering headlines. This is deceptive.
Tirpitz... Hide all time. Single time main battery is fired 140 norwegians die. One day it gets sank.
The problem is, as usual, other cultures. What is inconsistent to us may be completely consistent to them - "That character is really X, plus Y and Z modifiers" and "That character is X, but has two A modifiers applied to it", with the result being "Both are X in our language, but you use the first when the phase of moon is waxing, and the second is when your spouse committed adultery".
And then there's no real "capital" representation for languages with case - if you represent everything as characters, capitals are separate and distinct from lowercase, and when you need to do a case-insensitive search, you have to handle both. Or rely on Unicode-aware versions of tolower/toupper that don't always work on all languages and encodings.
Unicode has turned into a giant monster. It's primary goal now is to be the one code that you can then translate into other codes - so you can take any script in the world, and it'll map into Unicode. You can do the reverse too - take Unicode and map it out into a script, handling non-script characters as appropriate. Then you process your text in the chosen script.
At least, that was the justification for including emoji into Unicode as well - because the Japanese scripts include emoji, Unicode needs to as well so you can continue to map Japanese script into Unicode without having issues where some characters will not map into Unicode.
This is not good enough. Apple needs to issue updates for all the older affected OSs too. Not all hardware can run the new OSs. Not everyone wants the new OSs. Not all legacy software works with the old OSs. The result is there are a lot of older devices out there that need continued legacy support. The cost of fixing the older OSs is trivial. Apple should do it.
Like Sierra v10.12.6 and El Capitan v10.11.6?
Ant(Dude) @ Quality Foraged Links (AQFL.net) & The Ant Farm (antfarm.ma.cx / antfarm.home.dhs.org).
ASCII
Why the fuck hasn't /. removed the iOS shit character from the whitelist? Fucking hell.
Even back in the day of the mini- and first micro-computers there already were "Three Little Endians".
The third was DEC Endian, on the PDP-11 (at least initially - I don't recall if it propagated to other things like the VAX).
Due to the 16-bit word size and a byetwise addressing mode that treated the lower byte as zero LSB address (for byte-wise iteration of bignum arithmetic), peripheral I/O operations loaded the record with the even and odd bytes swapped. (ABCDEFGH -> BADCFEHG)
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way
...no problem.