• anonny (unregistered) in reply to James Schend
    James Schend:
    What's even worse is when Linux/open source systems "alphabetize" by ASCII value instead of, you know, actually alphabetizing things.
    Interesting point. The answer, I think, comes down to robustness.

    Sorting by character value may not always do exactly what you want, but it's entirely predictable in speed and extremely robust. There are no grey cases. Sort by alphabetic letter, though, and suddenly you have to start worrying about character sets and all the brain-damage they involve. Throw in having to guess file sequences and suddenly a robust, predictable algorithm becomes slow and unpredictable.

    The same goes for filesystems, sort of. Case-sensitivity is simple and robust; a different name is a different name, no ifs ands or buts. Implementing it as case insensitive creates a whole minefield of problems, special cases, workarounds, loopholes to close...

    These are the kind of design conditions MS makes constantly, sprinkling uncountable special cases everywhere to make things "easier" for one specific kind of user instead of programming for the general case. This is bad enough in a UI design, let alone an operating system.

  • Mr Steve (unregistered)

    I HATE APPLE!!!!

    Overpriced sh1t that breaks ALOT. We have a chick who uses a mac in the office and the thing just keep breaking / cant install new things, constantly patching OSX, blah blah. The thing is fussier then Linux. And the people who use these things go on about how windows is so bad but their own operating system is soooo much worse

    Look if you use apple you should realise something

    Apple = best marketing on the planet. Could sell ice cubes to eskimos

    You = just got sold buddy. wake the fork up

  • (cs) in reply to Mr Steve
    Mr Steve:
    I HATE APPLE!!!!

    Overpriced sh1t that breaks ALOT. We have a chick who uses a mac in the office and the thing just keep breaking / cant install new things, constantly patching OSX, blah blah. The thing is fussier then Linux. And the people who use these things go on about how windows is so bad but their own operating system is soooo much worse

    Look if you use apple you should realise something

    Apple = best marketing on the planet. Could sell ice cubes to eskimos

    You = just got sold buddy. wake the fork up

    Funny, I usually say the same thing about Windows and Microsoft, although I'm a bit more... articulate.

  • (cs) in reply to anonny
    anonny:
    James Schend:
    What's even worse is when Linux/open source systems "alphabetize" by ASCII value instead of, you know, actually alphabetizing things.
    Interesting point. The answer, I think, comes down to robustness.

    Sorting by character value may not always do exactly what you want, but it's entirely predictable in speed and extremely robust. There are no grey cases. Sort by alphabetic letter, though, and suddenly you have to start worrying about character sets and all the brain-damage they involve. Throw in having to guess file sequences and suddenly a robust, predictable algorithm becomes slow and unpredictable.

    The same goes for filesystems, sort of. Case-sensitivity is simple and robust; a different name is a different name, no ifs ands or buts. Implementing it as case insensitive creates a whole minefield of problems, special cases, workarounds, loopholes to close...

    Care to enumerate any of those?

  • Hilbilly Geek (unregistered) in reply to Franz Kafka

    Actually, that's the punchline to one of the Apple commercials that's getting all your panties in their proverbial wads. heh. There's quite enough smugness around on both sides, actually.

  • Ivan (unregistered) in reply to anonny
    anonny:
    James Schend:
    What's even worse is when Linux/open source systems "alphabetize" by ASCII value instead of, you know, actually alphabetizing things.
    Interesting point. The answer, I think, comes down to robustness.

    Sorting by character value may not always do exactly what you want, but it's entirely predictable in speed and extremely robust. There are no grey cases. Sort by alphabetic letter, though, and suddenly you have to start worrying about character sets and all the brain-damage they involve. Throw in having to guess file sequences and suddenly a robust, predictable algorithm becomes slow and unpredictable.

    Maybe, but it's a moot point anyway. Linux has locales, which define everything from how dates/times are formatted, through the sort order of strings, to which language to use. The locale is used by the command line tools as well as the desktop environments (Gnome/KDE/etc). You can even mix and match things if you want (eg display dates like europeans, but spell like americans).

    Basically you get customize how it all behaves in one central place, if the fancy strikes you. Generally you just use the settings for your region.

  • thimblehead (unregistered) in reply to gkdada

    OK, pinheads, what does MAC stand for? Message Authentication Code? Media Access Control? Oh, you're talking about an Apple Macintosh? It's not a fucking acronym! It's spelled "Mac," short for "Macintosh." See how that works? Now you try...

    And the irony that that original post was related to case-sensitivity is not lost on me.

    The real WTF is what a character from an old, crappy sitcom has to do with effective communication.

  • (cs) in reply to snoofle
    snoofle:
    KattMan:
    Especially when oracle actually has a specification for tri-state booleans with four states: http://www.oracle.com/technology/products/jdev/esdk/api1013/oracle/ide/util/TriStateBoolean.html
    Oh. My. Dear. Lord.

    The doc for that 4-state tri-state boolean looks like the javadoc parser blew a fuse.

    TriStateBoolean or(TriStateBoolean b) OR true false tri true true true tri false true false tri tri tri tri tri

  • Bert (unregistered) in reply to boolean
    boolean:
    When you get into case-insensitivity, you have to define what ASCII values are equivalent to one another. What happens when you get into UTF encoding, and different languages or dialects that may or may not have upper-/lower-case equivalents to certain letters? Purely hypothetical, as I'm not aware of any language that behaves this way, but it seems that case-sensitivity leaves less room for error.

    That was exactly my thought, and I can give you an example. I started studying Turkish, and they have some extra letters. In particular, they have 4 i's; with and without dot, both upper and lower case: i I i I.

    As far as I know, the I and i share the same unicode (and thus utf-8) code as the English I and i, while actually being a different case of totally different letters. I do not know how this is handled in computers supporting Turkish.

  • Anonymous (unregistered) in reply to James Schend
    James Schend:
    What's even worse is when Linux/open source systems "alphabetize" by ASCII value instead of, you know, actually alphabetizing things.

    Files: apple.png, pear.png, raspberry.png, Artichoke.png, Peach.png

    Think about this one, open source programmers: Which is easier to remember? That the filename starts with an A or that the filename starts with a capital/lowercase?

    That's just a matter of locale settings. The default locale POSIX sorts strings using ASCII order, which is NOT alphabetical order. If you change it to en_US, it'll sort alphabetically.

    Here is a simple experiment:

    $ /tmp$ mkdir /tmp/foo
    $ /tmp$ cd !$
    cd /tmp/foo
    $ /tmp/foo$ touch a A b B c C
    $ /tmp/foo$ ls
    A  B  C  a  b  c
    $ /tmp/foo$ locale
    LANG=
    LC_CTYPE="POSIX"
    LC_NUMERIC="POSIX"
    LC_TIME="POSIX"
    LC_COLLATE="POSIX"
    LC_MONETARY="POSIX"
    LC_MESSAGES="POSIX"
    LC_PAPER="POSIX"
    LC_NAME="POSIX"
    LC_ADDRESS="POSIX"
    LC_TELEPHONE="POSIX"
    LC_MEASUREMENT="POSIX"
    LC_IDENTIFICATION="POSIX"
    LC_ALL=
    $ /tmp/foo$ LANG=en_US ls 
    a  A  b  B  c  C
    $ /tmp/foo$ LANG=en_US locales
    -bash: locales: command not found
    $ /tmp/foo$ LANG=en_US locale 
    LANG=en_US
    LC_CTYPE="en_US"
    LC_NUMERIC="en_US"
    LC_TIME="en_US"
    LC_COLLATE="en_US"
    LC_MONETARY="en_US"
    LC_MESSAGES="en_US"
    LC_PAPER="en_US"
    LC_NAME="en_US"
    LC_ADDRESS="en_US"
    LC_TELEPHONE="en_US"
    LC_MEASUREMENT="en_US"
    LC_IDENTIFICATION="en_US"
    LC_ALL=
    

    Linux (actually, it's GNU libc's i18n library) is that flexible. If you aren't smart enough to handle that flexibility, then it's not Linux's fault. The problem is YOU.

  • jdj (unregistered)

    Well duh, the Mac mouse has only one button too, so why would you need more than one button on the screen! :)

  • (cs) in reply to Opie
    Opie:
    4-state logic can be achieved with 2 bits and can be accomplished with standard logical operators and no WTFs whatsoever. Who the bloody hell at Oracle allowed someone to create that?

    That actually upsets me and raises questions in my mind about what is going on under the hood in Oracle.

    Oh, you don't want to go there ...

    I'm frankly puzzled about the OP. I think Apple need to work this idea through and incorporate the "Keanu Reeves" concept directly into their product range, and not just in their marketing.

    I mean, "Yes" and "No" are pretty authoritarian, when it comes down to it. And "OK" might well have its roots in the military. What this dialog needs, I think, is a "Cool" button. Or maybe even a "Dude!" Yes, "Dude!" would work. In fact it's the only button you will ever need. Throughout the entire operating system.

    I also look forwards to Apple replacing the "Error" dialog, which is a tad accusatory, with something more appropriate, like a "Bogosity" dialog.

    Now, wouldn't that make you want to buy a Mac?

  • Thilo (unregistered)

    Fix The Fucking Finder™ ;)

  • (cs)

    My guess is (to get back to the original topic) that closing the dialog box rather than clicking 'Ok' will cancel the operation (i.e. 'No').

    I've seen at least one Windows app that was as shoddily coded as that, with exactly the same behaviour.

    So there...

  • sol (unregistered) in reply to iMalc
    iMalc:
    The strangest thing I ever saw on a Mac (many years ago) was this error that popped up for no apparent reason and said: "An error occurred because: An error Occurred" and it had only an okay button.

    But I wasn't really even doing anything.

    I love this.

  • (cs) in reply to boolean
    boolean:
    When copying files and there are existing files with the same name, you choose between Yes | Yes to All | No | Cancel. Where is the "No to All" button? Why isn't there a check box for "All" like in OSX?

    Umm... Wouldn't "Cancel" be the same as "No to All"? And doesn't "Yes to All" do the same as the OSX checkbox?

    boolean:
    How about when you're moving a large amount of files, and one of them has an error?

    How does OSX handle this?

    boolean:
    What about when you delete/move/copy/rename files, and then you're somewhere else (Desktop, explorer, etc), possibly hours later and you accidentally hit Undo or Ctrl-Z? Whoops! Whatever file operation you did last, anywhere on the system, has now been undone. Does Windows tell you what that operation was? No. Does it let you redo the action? No. Hope it wasn't anything important.

    I've been using Windows since 3.0, and have never had this happen. Never. Can you provide the exact steps to replicate this problem? (Something less stupid than doing a delete/move/copy/rename in Explorer, changing focus to the Desktop, changing focus back to Explorer, and doing a Ctrl-Z, please.)

  • Alfred (unregistered)

    Seriously, if you're going to post such an easily verifiable WTF, do verify it. I don't know what software, if any, produced this, but no version of OS X I have access to (10.2, 10.3, 10.4) gives this message.

  • Diego (unregistered) in reply to EvanED

    what about the "Class object;" idiom? e.g: Widget widget(10,10); or Widget widget = new Widget(10,10);

    captcha: tesla, the magic who made a duplicating machine

  • Jesse (unregistered) in reply to gkdada

    iHawk? Will he be you switch him to 'Irreverent' or 'Maudlin?'

  • (cs) in reply to KenW
    KenW:
    boolean:
    When copying files and there are existing files with the same name, you choose between Yes | Yes to All | No | Cancel. Where is the "No to All" button? Why isn't there a check box for "All" like in OSX?

    Umm... Wouldn't "Cancel" be the same as "No to All"? And doesn't "Yes to All" do the same as the OSX checkbox?

    Cancel = stop copying No to all = keep copying, but only files that don't conflict

  • Icelight (unregistered)

    Right clicks out of topic

  • (cs) in reply to KenW
    KenW:
    boolean:
    When copying files and there are existing files with the same name, you choose between Yes | Yes to All | No | Cancel. Where is the "No to All" button? Why isn't there a check box for "All" like in OSX?
    Umm... Wouldn't "Cancel" be the same as "No to All"?
    No, "Cancel" would stop the operation. "No to All" would continue the operation, ignoring conflicting filenames.
    KenW:
    And doesn't "Yes to All" do the same as the OSX checkbox?
    Yes. The second sentence was an extension of the complaint that there was no option to answer no to all. The check box and the extra button would be viable solutions, not both.
    KenW:
    boolean:
    How about when you're moving a large amount of files, and one of them has an error?
    How does OSX handle this?
    I don't know; my gripe wasn't that OSX did it better, but that the current solution is extremely confusing and unhelpful to the user. The operation that was requested was neither fully unsuccessful nor fully successful, and there was no option to continue with unaffected files. A user-friendly way to handle this would be one of the options I mentioned, while keeping the user informed of what's going on.
    KenW:
    boolean:
    What about when you delete/move/copy/rename files, and then you're somewhere else (Desktop, explorer, etc), possibly hours later and you accidentally hit Undo or Ctrl-Z? Whoops! Whatever file operation you did last, anywhere on the system, has now been undone. Does Windows tell you what that operation was? No. Does it let you redo the action? No. Hope it wasn't anything important.

    I've been using Windows since 3.0, and have never had this happen. Never. Can you provide the exact steps to replicate this problem? (Something less stupid than doing a delete/move/copy/rename in Explorer, changing focus to the Desktop, changing focus back to Explorer, and doing a Ctrl-Z, please.)

    First of all--the wrong button/menu item can always be clicked by both stupid and non-stupid users. I'm not stupid, and have been using Windows since 3 as well, so this isn't an experience issue.

    I'll provide the simplest possible example, however:

    1. In explorer, rename a file.
    2. Close explorer. Go to desktop.
    3. Right click the desktop, as if you're going to click "Properties". "Undo rename" is an option. Note that you can't tell what was renamed, or when. (Same with move/copy/delete). Clicking it, perhaps by accident, yields no redo option. Hitting CTRL-Z, perhaps by thinking another window was in focus when the desktop was in focus, will result in the same action.

    My point isn't that this mistake is common, it's that the interface doesn't properly inform you of what will be undone, nor does it allow you to redo a mistaken undo. But I suppose you don't have to worry about it, since you evidently have never made any mistakes while using computers.

  • ropata (unregistered)

    Dude! Like, this copy operation is gonna totally wipe out some of your files!!! What ya wanna do man? [forget it dude] [waste the old files, dude!] [keep the old files someplace else dude]

  • Nico (unregistered)

    Hmm, i guess in this case he must take the blue pill!

  • noise (unregistered)

    someday copy dialogs will actually list the files that the copy had problems with, in a format that i can cut and paste or otherwise get easy access to - so that i can actually do something about the problems it encountered.

  • myname (unregistered)

    "Their names are the same, except for the case of some characters."

    I thought that next filenames were conflicting: abcdefgh.txt, axcdefgh.txt. They are almost the same, but there are a few characters in which case they differ.

  • myname (unregistered) in reply to boolean
    boolean:
    However, when I look at it from the computer's standpoint, why should "FILE" and "file" be the same? They're composed of completely different ASCII characters. When you get into case-insensitivity, you have to define what ASCII values are equivalent to one another. What happens when you get into UTF encoding, and different languages or dialects that may or may not have upper-/lower-case equivalents to certain letters? Purely hypothetical, as I'm not aware of any language that behaves this way, but it seems that case-sensitivity leaves less room for error.
    Turkish has some letters which is said to cause many software bugs where case insensitivity/case alteration is needed:

    i (ascii) <-> dotted I i (dotless) <-> I (ascii)

    http://en.wikipedia.org/wiki/Turkish_dotted_and_dotless_I

  • (cs)

    The "revised" dialog box with Stop and Continue doesn't make any sense either.

    The question is not clear (Do you want to do A and B?). Yes, I want to do A and B. Or No, I don't want to do A and B.

    The question "Do you want to skip copying these items and continue copying the other items?" calls out for a Yes/No answer, as far as I understand my native language of American English.

  • Anonymous Coward (unregistered) in reply to anonny
    anonny:
    Interesting point. The answer, I think, comes down to robustness.

    Sorting by character value may not always do exactly what you want, but it's entirely predictable in speed and extremely robust. There are no grey cases. Sort by alphabetic letter, though, and suddenly you have to start worrying about character sets and all the brain-damage they involve. Throw in having to guess file sequences and suddenly a robust, predictable algorithm becomes slow and unpredictable.

    The same goes for filesystems, sort of. Case-sensitivity is simple and robust; a different name is a different name, no ifs ands or buts. Implementing it as case insensitive creates a whole minefield of problems, special cases, workarounds, loopholes to close...

    Ahh, the old "worse is better" design philosophy (see the "Unix haters handbook" for details). I disagree. The computer should adjust to the real world, not the other way around. The correct way to handle the complexity you describe is to create a library function for it that any program can use, not to ignore it.
    These are the kind of design conditions MS makes constantly, sprinkling uncountable special cases everywhere to make things "easier" for one specific kind of user instead of programming for the general case. This is bad enough in a UI design, let alone an operating system.
    I consider correct alphabetical sorting (taking all language-dependant special cases into proper account) to be the general case, and using the ASCII order as special case (the case where the programmer is too lazy to do it right). Which is why I do embedded programming, where the issue doesn't come up =8^)
  • Henrik (unregistered) in reply to foxyshadis
    foxyshadis:
    chrismcb:
    While not quite in the same vein, I am unhappy that Vista changed the save as dialog in notepad:

    "Do you want to save changes?" Choices: "Save" "Don't Save" "Cancel"

    In previous versions of windows it was "Yes" "No" "Cancel".

    Not only do I now I have to hit 'S' to save (instead of 'Y'). it doesn't make sense. You asked me a yes/no question, let me respond yes or no.

    This is one of my bigger gripes with mac/linux HCI, but not because of whether it makes logical sense or not. Basically just because I don't give a damn what stupid verb the designer came up with for the action, I'd much rather use the muscle memory without having to read & parse the question and answers everytime.

    The idea is that you don't have to parse the question--the dialog doesn't even need to contain a question. Instead of asking a yes/no question (or often something like: "Installation finished. Click yes to restart your computer or no to look at the README file."), the dialog describes the situation and you select a course of action.

    For example: Situation: The file you are about to close has unsaved changes. Choices: Save, Don't Save

    Instead of: Question: Do you want to save the changes? Answers: Yes, No

    ...and in another program: Question: Do you want to discard the changes? Answers: Yes, No

    ...and perhaps in yet another program: Question: Are you sure you don't want to avoid discarding the changes? Answers: Yes, No

    No matter how the question is worded, "Save" and "Don't Save" would mean the same thing, unlike "yes" and "no". Even if you switch the logic around, "Discard" and "Don't Discard" are easily understood no matter the wording of the question.

  • Zoki (unregistered)

    This error appears mostly when copying files with names containing special characters to a network share (Samba comes to mind), e.g. "Mon fichier méga cool d'année dernière & de cette année.txt".

    Doesn't necessarily mean the file really exists, but most probably that Finder doesn't trap all errors properly.

    Cheers.

  • OldNeXTer (unregistered) in reply to jaserlet

    "Yep. Ever since NeXTstep 1.0"

    Good to see an old NeXTer here. Ah, the memories. The beauty of Webster on a 4 grayscale monitor (the "cat" entry was an all time favorite). The need to open two workspace window to perform a copy. The 8 minutes boot time on optical disc.

    NeXTstep 1.0 (well, 0.9 for me). Mmmmmm.

  • mct (unregistered) in reply to EvanED
    EvanED:
    Opie:
    shift + no = no to all in windows.

    Enjoy

    Wow, that rules. I'll have to remember that one.

    Here's another... um, innovative UI feature from our friends at Redmond: in Excel, shift-click the edit menu. A new item will appear, copy picture, which is very useful to get a chart or something into a JPG/GIF/PNG/whatever. Saves printscrn, paste, crop.

    And in Word, Shift-File gives you a Close All option...

  • Clark Cox (unregistered) in reply to pinkduck

    Even Better:

    "Some items already exist in the destination folder with the same names but differing case. To continue, you must overwrite or skip copying the offending items"

    Cancel|Skip|Overwrite

  • Hackaback (unregistered)

    Really Funny thing. Haven't used mac yet but this is awesome

  • Pecos Bill (unregistered) in reply to Opie
    Opie:
    anne:
    the Mac guy is so irritatingly smug.

    Congratulations, you just described 90% of Mac users. Odd how a lot of them generally can't seem to handle tasks through which their hand is not held, when working with a computer.

    Funny, most Mac users feel that of 90% Windows users. (The sea of corporate users and home users easily qualify)

  • Pecos Bill (unregistered) in reply to EvanED
    EvanED:
    I'm going to catch a lot of flak for this I know, but the Real WTF is that most filesystems support case-sensitive file names.

    The only flak deserved is where people are wrong. Apple file systems (the built in ones) vary. All of them are case preserving. The default, HFS+, is case insensitive as most users don't want Files.txt and files.txt in the same folder with differing content.

    The dialog is a giant WTF but John Q User will never see it as others pointed out. The default file system is HFS+. The biggest WTF is there is no cancel button closely followed by the horrid wording. None of the people thus far actually fixed that either unless you fix the underlying code or file system API. The code wants to copy all the files that do not have case issues. I would hope the issue only arises when copying from a case sensitive to case insensitive fs. Anyone who formatted with a case sensitive fs would understand the horribly worded dialog.

    Windows XP's NTFS (or at least its code) is also case insensitive based on my test just now. Unix is the only case sensitive OS / fs I know of. Intuitively, why should files with the same name (speaking English that is) but differing case be different?

  • JM (unregistered) in reply to gilleain
    gilleain:
    Those ads are the smuggest ever. I know that's not a word, shutup.

    Smuggest is indeed a word.

  • Mookie (unregistered) in reply to guesto

    Haha! Zing! I was thinking the same thing.

  • Pecos Bill (unregistered) in reply to boolean
    boolean:
    How about when you're moving a large amount of files, and one of them has an error? It stops the entire move operation, presents you with an error box and an OK button, then dumps you back to whatever you were doing. Was the rest of the move after the problem file completed? No. Was the partially-successful operation rolled back, moving the files back to their original location? No. You're left with a half-finished move operation, and now you have to pick through which files weren't moved manually, or just repeat the whole operation. With large or deeply-nested folders, this can be a real PITA.

    What about when you delete/move/copy/rename files, and then you're somewhere else (Desktop, explorer, etc), possibly hours later and you accidentally hit Undo or Ctrl-Z? Whoops! Whatever file operation you did last, anywhere on the system, has now been undone. Does Windows tell you what that operation was? No. Does it let you redo the action? No. Hope it wasn't anything important.

    I use both XP and OS X and am in the power user realm but not an expert on either. OS X has flaws but it is by far my favorite. That being said, there's a LOT of FFTF complaints. FFTF Fix the F'd/F'ing Finder. IIRC, if there is a catastrophic error which usually DOES result in a cryptic error message (-36 anyone?). The copy fails with a partially complete state. Copying files under Classic was far superior with CopyDoubler.

    I suspect that you could have the exact same problem with OS X Finder undo though there is a redo option.

    OS X flaw/peeve: allowing apps to steal focus. Gawd I despise that. It's bad enough when they bring to front-but-not-focused. Well behaved applications never steal focus but some things are worth suffering: NeoOffice vs paying any money to M$. At least Apple tries to maintain app focus while launching an app in the background.

  • Room (unregistered)
    Comment held for moderation.

Leave a comment on “Apple, Simple and Elegant”

Log In or post as a guest

Replying to comment #:

« Return to Article