Microsoft Word WTF



  • I've been having an issue with Microsoft Word.  I've been tasked with creating some documentation and putting it on our shared drive.  The shared drive is running out of space.  When I try to save it using Word, I get this message:

    If you don't want to click, it's giving a WTF-style "Out of disk space" error.  It says to "free more memory" as opposed to "free more disk space."  The last bit of the message is part of the filename.

    The WTF is that I can save the file to my own hard drive, then copy it to the shared drive using Windows Explorer with no issues.


     



  • Word for Mac has the exact same problem.  It's in the knowledge base, something to do with corrupt document templates IIRC.



  • I'm not really surprised honestly, the latest WMP on my computer will "run out of memory" and go insane (missing graphics etc.) before crashing. Now considering I have 4 gigs of RAM and this can happen whether I have 100 programs or 1 program running I'm thinking someone over at MS screwed up their memory management.



  • @teqman said:

    Word for Mac has the exact same problem. It's in the knowledge base, something to do with corrupt document templates IIRC.

    Forget which version but it would ask you if you wanted to save your current document as the 'normal template' every. Single. Time.



  • @belgariontheking said:

    I've been having an issue with Microsoft Word.  I've been tasked with creating some documentation and putting it on our shared drive.  The shared drive is running out of space.  When I try to save it using Word, I get this message:

    If you don't want to click, it's giving a WTF-style "Out of disk space" error.  It says to "free more memory" as opposed to "free more disk space."  The last bit of the message is part of the filename.

    The WTF is that I can save the file to my own hard drive, then copy it to the shared drive using Windows Explorer with no issues.


     

     

    Any smart application (debatable if that applies to Word) will save a document by creating a new file with a temporary name, spit out the data into it, and then delete the old file if the save was successful. Otherwise, as is happening with you, if the save fails for some reason, you're not left with a corrupt partial copy on disk.

    Basically it sounds like the shared disk has somewhat less than (size of original file + size of new temp copy) worth of space left.

    But yep, Word's got moronic error messages.
     



  • @Lingerance said:

    @teqman said:
    Word for Mac has the exact same problem. It's in the knowledge base, something to do with corrupt document templates IIRC.
    Forget which version but it would ask you if you wanted to save your current document as the 'normal template' every. Single. Time.

    That'd happen if you opened the Normal.dot as if it was a regular file, doing some edits, and hitting File->Save, instead of having started out with File->New. 



  • Free more memory - as in: add more ram -> use less swap space on your disk -> have more disk space

     



  • + Since it's some kind of shared drive my bet is that it's a problem with permissions and that the disk is not really full, just write protected. Hence copying it to your drive first might be circumventing the underlying permission problem

    rtfm?
     



  • @GuntherVB said:

    + Since it's some kind of shared drive my bet is that it's a problem with permissions and that the disk is not really full, just write protected. Hence copying it to your drive first might be circumventing the underlying permission problem

    rtfm?

    Nah, I can write to the drive plenty.  I can copy the file there with Windows Explorer, just not Word



  • I have the odd feeling that someone at Microsoft wrote code that calls GetDiskFreeSpace() (which will only report at most 2 GB free) instead of GetDiskFreeSpaceEx()--or missed updating the code when Microsoft dropped support for Win95 SP1.



  • Having read the title of this thread I came in here to say "yes, it is."

    Then I realised there was an actual POST and got all disappointed. 



  • I never have memory issues in Notepad.  Hell, it's probably the least bug-ridden Windows application.



  • @purge said:

    I never have memory issues in Notepad.  Hell, it's probably the least bug-ridden Windows application.


    Really?  You've never opened a file too big for Notepad?



  • @belgariontheking said:

    The WTF is that I can save the file to my own hard drive, then copy it to the shared drive using Windows Explorer with no issues.

    Word creates a big hidden temp file in the same directory as your file, for "recovery" purposes.  It also has undo information, etc if I'm not mistaken.  That's why it gets so big.

    If you're going to say that file's the WTF, remember that vi does the same thing..... 



  • @MarcB said:

    @Lingerance said:

    @teqman said:
    Word for Mac has the exact same problem. It's in the knowledge base, something to do with corrupt document templates IIRC.

    Forget which version but it would ask you if you wanted to save your current document as the 'normal template' every. Single. Time.

    That'd happen if you opened the Normal.dot as if it was a regular file, doing some edits, and hitting File->Save, instead of having started out with File->New. 


    I opened it from the little icon we got with our school user accounts' desktop.



  • Heh, don't get me started on word and disk space.  I have a 32 page document here, that's ballooned to 3.5MB in size.  Not a complex document mind, it's mainly text, with a few hyperlinks, a 74KB PDF embedded in it, and a couple of e-mails.

    But then 3.5MB wasn't enough for Microsoft, oh no.  If I have a look at the folder it's saved in and show hidden files, I can see that word has created twenty four temporary files as I've been saving my work.  So that single document is currently taking up 57.7MB of disk space.  57MB!  For a 32 page document?  The sooner we can get Linux and OpenOffice in this building the better.

     



  • Speaking of Office WTFs, why won't Excel let me save a document with [ or ] in the name?



  • @myxiplx said:

    Heh, don't get me started on word and disk space.  I have a 32 page document here, that's ballooned to 3.5MB in size.  Not a complex document mind, it's mainly text, with a few hyperlinks, a 74KB PDF embedded in it, and a couple of e-mails.

    But then 3.5MB wasn't enough for Microsoft, oh no.  If I have a look at the folder it's saved in and show hidden files, I can see that word has created twenty four temporary files as I've been saving my work.  So that single document is currently taking up 57.7MB of disk space.  57MB!  For a 32 page document?  The sooner we can get Linux and OpenOffice in this building the better.

     

    Step 1:  FIle -> Save As

    Step 2:  RTFM



  • @ShadowWolf said:

    @myxiplx said:

    So that single document is currently taking up 57.7MB of disk space.  57MB!  For a 32 page document? 

    Step 1:  FIle -> Save As

     ITYM "File -> Versions -> select all -> Delete", don't you? 

     



  • @ender said:

    Speaking of Office WTFs, why won't Excel let me save a document with [ or ] in the name?
    Or open two files with the same filename, but in different directories...



  • @DaveK said:

    @ShadowWolf said:
    @myxiplx said:

    So that single document is currently taking up 57.7MB of disk space.  57MB!  For a 32 page document? 

    Step 1:  FIle -> Save As

     ITYM "File -> Versions -> select all -> Delete", don't you? 

     

    That would probably be the "best practice" method of doing it.  The Save As trick just comes to mind first.



  • @mare said:

    @ender said:
    Speaking of Office WTFs, why won't Excel let me save a document with [ or ] in the name?
    Or open two files with the same filename, but in different directories...

    Why is any file open in Office locked?  Read-only makes sense, but prohibiting read?  I can't think of anything else that does that.  "I'm sorry, you can't send the file you just saved to your colleagues without closing it in your office suite, even though that's a giant pain in the ass."  I especially like it when it's totally not obvious that's the reason...  Have you ever done an ad-hoc distributed query from SQL Server via oledb to an excel file?  And had it fail without giving an error message? (I think that's actually how it words it too, 'the provider did not give an error message')  And spent 5 minutes trying to debug it before realizing it's because you have the file open?  Gahhhh!

    I wish it was a Microsoft wtf, but OpenOffice does that too.  Drives me up the wall!  You're already saving auto-saves to a temp file, what possible need is there to LOCK the destination file? 



  • @misguided said:

    I wish it was a Microsoft wtf, but OpenOffice does that too.  Drives me up the wall!  You're already saving auto-saves to a temp file, what possible need is there to LOCK the destination file?
    I think this happens because while you're editing a file, it's marked as open, and if another user tries to load it, they'll get a "User XXX is editing this file." You can then open it as read-only, and choose to be notified when the other user closes the file.



  • @misguided said:

    I wish it was a Microsoft wtf, but OpenOffice does that too.  Drives me up the wall!  You're already saving auto-saves to a temp file, what possible need is there to LOCK the destination file? 

    It's Windows, not the application. 



  • @Cap'n Steve said:

    @purge said:
    I never have memory issues in Notepad. Hell, it's probably the least bug-ridden Windows application.


    Really? You've never opened a file too big for Notepad?

     

    Notepad is awful and has plenty of bugs for a simple friggin' text editor.  First of all, there this ridiculous bug as described on the snopes message board: http://msgboard.snopes.com/message/ultimatebb.php?/ubb/get_topic/f/56/t/002928.html.  How in the hell do you even pull something like that off?

     
    Secondly, it doesn't have any clue about how to render Unix-style text files properly.  Each line needs to be delimited by a carriage return and a newline character. If you just have a newline character, it will barf and display a rectangle indicating "non-printable symbol."  Any text editor worth a crap should be able to figure out that a single newline character indicates a friggin' new line!

    Plus,  as Steve said, the file size limitations are ridiculous.  I can't think of a single good use for Notepad when there are so many better text editors out there.



  • @bighusker said:

    Notepad is awful and has plenty of bugs for a simple friggin' text editor.  First of all, there this ridiculous bug as described on the snopes message board: http://msgboard.snopes.com/message/ultimatebb.php?/ubb/get_topic/f/56/t/002928.html.  How in the hell do you even pull something like that off?
    This is not a bug in the notepad, but simply the way IsTextUnicode function works. And the limitation's been known for a long time.



  • @ender said:

    @bighusker said:
    Notepad is awful and has plenty of bugs for a simple friggin' text editor. First of all, there this ridiculous bug as described on the snopes message board: http://msgboard.snopes.com/message/ultimatebb.php?/ubb/get_topic/f/56/t/002928.html. How in the hell do you even pull something like that off?
    This is not a bug in the notepad, but simply the way IsTextUnicode function works. And the limitation's been known for a long time.

     

    Do other text editors have this problem?  I'd imagine there has to be some reasonable workaround for this.

     



  • @bighusker said:

    @ender said:

    @bighusker said:
    Notepad is awful and has plenty of bugs for a simple friggin' text editor. First of all, there this ridiculous bug as described on the snopes message board: http://msgboard.snopes.com/message/ultimatebb.php?/ubb/get_topic/f/56/t/002928.html. How in the hell do you even pull something like that off?
    This is not a bug in the notepad, but simply the way IsTextUnicode function works. And the limitation's been known for a long time.

     

    Do other text editors have this problem?  I'd imagine there has to be some reasonable workaround for this.

     

    Working exclusively with unicode text comes to mind. Or, of course, working exclusively in the system codepage. In the end I would guess most other editors use their own algorithm to identify the encoding used for a file. And, regardless, notepad's "bug" only happens when the file does not have a line break at the end, which text files are supposed to always have (but, then, the fact that notepad can also save a text file with no line break, without being explicitly told to, makes this doubly notepad's WTF) - in the end, we're expecting too much from what is basically a demo of the multiline text editor control; I'm just glad they finally modernized the key shortcuts.



  • @bighusker said:

    Notepad is awful and has plenty of bugs for a simple friggin' text editor.  First of all, there this ridiculous bug as described on the snopes message board: http://msgboard.snopes.com/message/ultimatebb.php?/ubb/get_topic/f/56/t/002928.html.  How in the hell do you even pull something like that off?

     
    Secondly, it doesn't have any clue about how to render Unix-style text files properly.  Each line needs to be delimited by a carriage return and a newline character. If you just have a newline character, it will barf and display a rectangle indicating "non-printable symbol."  Any text editor worth a crap should be able to figure out that a single newline character indicates a friggin' new line!

    Plus,  as Steve said, the file size limitations are ridiculous.  I can't think of a single good use for Notepad when there are so many better text editors out there.

    Gee, let's see. You think it's a bug that a WINDOWS text editor doesn't understand UNIX line endings. So you're another "let's bash MS for everything that happens, even when it's not their fault" /.er, huh?

    For the purpose it's designed for (reading/editing text files on Windows sytems), Notepad is fine. If the file size limitations or inability to deal with other than MS-DOS/Windows line feeds are an issue, you're obviously using it for more than what it was designed to do, and therefore are the one who is doing something wrong. Your choice to use the wrong tool for the job doesn't make the tool bad - it shows you don't have enough sense to choose the right tool.

    As usual, there's nothing to see here, folks. Move along.
     



  • @KenW said:

    Gee, let's see. You think it's a bug that a WINDOWS text editor doesn't understand UNIX line endings. So you're another "let's bash MS for everything that happens, even when it's not their fault" /.er, huh?
     



    Well Wordpad has no problem with it.  I wonder why they haven't just replaced notepad with wordpad, actually.



  • @Cap'n Steve said:

    @KenW said:

    Gee, let's see. You think it's a bug that a WINDOWS text editor doesn't understand UNIX line endings. So you're another "let's bash MS for everything that happens, even when it's not their fault" /.er, huh?



    Well Wordpad has no problem with it. I wonder why they haven't just replaced notepad with wordpad, actually.

    Not to mention the major Unix text editors (whether console or graphical) have the decency to support Windows line endings. But then, decency is not MS strong point.



  • @m0ffx said:

    Not to mention the major Unix text editors (whether console or graphical) have the decency to support Windows line endings. But then, decency is not MS strong point.

    STOP YOUR WHINING B*TCH AND JUST USE CRLF LIKE EVERYONE ELSE.  THATS WHY LINUX WILL NEVER BE AS GOOD AS WINDOWS 



  • @ender said:

    @bighusker said:
    Notepad is awful and has plenty of bugs for a simple friggin' text editor. First of all, there this ridiculous bug as described on the snopes message board: http://msgboard.snopes.com/message/ultimatebb.php?/ubb/get_topic/f/56/t/002928.html. How in the hell do you even pull something like that off?
    This is not a bug in the notepad, but simply the way IsTextUnicode function works. And the limitation's been known for a long time.

    Repeat after me:  It is a bug.

    Let me tell you what the most important user story for a text editor is:

        User types in stuff.  User saves this to a file.  Later, use opens file and sees exactly what he typed before.  If you wrote notepad and you saw this, would you say, "it's not a bug...  I've known about this for a long time."   Or would you say, "gee, this is a tricky issue, lets figure out how to get around it."

     

    PS.  Saying that text files are only valid if they have a newline at the end is about the dumbest thing I've ever heard.  A text file is a text file and is valid so long as it contains characters defined in the desired standard (ASCII/unicode) 



  • @purge said:

    STOP YOUR WHINING B*TCH AND JUST USE CRLF LIKE EVERYONE ELSE.  THATS WHY LINUX WILL NEVER BE AS GOOD AS WINDOWS 

     

    wow, purge, having a bad day or something?  Is it because your caps lock key is stuck on?



  • @bighusker said:

    @ender said:
    @bighusker said:
    Notepad is awful and has plenty of bugs for a simple friggin' text editor. First of all, there this ridiculous bug as described on the snopes message board: http://msgboard.snopes.com/message/ultimatebb.php?/ubb/get_topic/f/56/t/002928.html. How in the hell do you even pull something like that off?
    This is not a bug in the notepad, but simply the way IsTextUnicode function works. And the limitation's been known for a long time.
    Do other text editors have this problem?  I'd imagine there has to be some reasonable workaround for this.

    There is. It's called; Letting the user decide which encoding to use. Seriously; What's so hard about popping up a dialog box that states the application isn't sure what format the file has and asking the user to decide for it?

    Windows and its bundled applications are made to be usable by braindead idiots, but even that category of user wouldn't have problems with something like that, would they?



  • @Ragnax said:

    @bighusker said:

    @ender said:
    @bighusker said:
    Notepad is awful and has plenty of bugs for a simple friggin' text editor. First of all, there this ridiculous bug as described on the snopes message board: http://msgboard.snopes.com/message/ultimatebb.php?/ubb/get_topic/f/56/t/002928.html. How in the hell do you even pull something like that off?
    This is not a bug in the notepad, but simply the way IsTextUnicode function works. And the limitation's been known for a long time.
    Do other text editors have this problem?  I'd imagine there has to be some reasonable workaround for this.

    There is. It's called; Letting the user decide which encoding to use. Seriously; What's so hard about popping up a dialog box that states the application isn't sure what format the file has and asking the user to decide for it?

    Windows and its bundled applications are made to be usable by braindead idiots, but even that category of user wouldn't have problems with something like that, would they?

    Because when a user opens a text file and they see a dialog that comes up saying "This file is saved in an unrecognizable file format, please select the one you want" they are obviously going to very familiar with the encoding of that file.

    If you knew much about language encoding issues, I doubt you'd say this.  If you do know much about language encoding issues and you're saying this then you're just being elitist.

    There is no "right" solution because every solution is either an annoyance (every time you open File "X" you recieve a warning about how it may not have opened correctly despite that it worked fine) or a presumption of knowledge.



  • @bighusker said:

    @ender said:
    @bighusker said:
    Notepad is awful and has plenty of bugs for a simple friggin' text editor. First of all, there this ridiculous bug as described on the snopes message board: http://msgboard.snopes.com/message/ultimatebb.php?/ubb/get_topic/f/56/t/002928.html. How in the hell do you even pull something like that off?
    This is not a bug in the notepad, but simply the way IsTextUnicode function works. And the limitation's been known for a long time.
    Do other text editors have this problem?  I'd imagine there has to be some reasonable workaround for this.

    Yeah it's a really dumb bug, and the solution is simply not to support UTF-7.  That's a deprecated encoding that, as far as I know, isn't used anywhere and has never really been used either.  It was originally conceived as a 7-bit encoding for email, but everybody decided that base64-encoded UTF-8 or UTF-16 was a superior solution to that problem, so UTF-7 never took off.



  • @seaturnip said:

    Yeah it's a really dumb bug, and the solution is simply not to support UTF-7.  That's a deprecated encoding that, as far as I know, isn't used anywhere and has never really been used either.  It was originally conceived as a 7-bit encoding for email, but everybody decided that base64-encoded UTF-8 or UTF-16 was a superior solution to that problem, so UTF-7 never took off.
    Notepad doesn't actually support UTF-7. It just interprets the file as if it was UTF-16.



  • @ender said:

    @seaturnip said:
    Yeah it's a really dumb bug, and the solution is simply not to support UTF-7. That's a deprecated encoding that, as far as I know, isn't used anywhere and has never really been used either. It was originally conceived as a 7-bit encoding for email, but everybody decided that base64-encoded UTF-8 or UTF-16 was a superior solution to that problem, so UTF-7 never took off.
    Notepad doesn't actually support UTF-7. It just interprets the file as if it was UTF-16.

    Ah, okay, I see, I misread your link.  So this "statistical analysis" determines the file is UTF-16 even though there are no null bytes, no bytes with the high bit set and it is a series of common ASCII English dictionary words separated by spaces.  Brilliant.  This must be the same kind of analysis my mutual fund uses.



  • @seaturnip said:

    So this "statistical analysis" determines the file is UTF-16 even though there are no null bytes

    Pretty much guaranteed if the text is in a different alphabet (Chinese, Japanese, etc.) without Latin glyphs for numbers. 

    , no bytes with the high bit set

    There doesn't have to be if you allow for non-BOM text. (that's possibly the REAL WTF in this case, but let's ignore that for now)

     and it is a series of common ASCII English dictionary words separated by spaces.

    ...because obviously, every such algorithm should not only be biased towards a specific language, but also think that a space will always be a word separator, even though it doesn't have to be in UTF-16.



  • I think actually you'd be hard-pressed to find a Chinese or Japanese UTF-16 document with no null bytes.  In plain text, there would be one after a carriage return I think.  Richer formats like HTML would have them all over the place in the ASCII-based markup.

    And 0x20 does not always have to be a space yes, but if there are five in 25 or so bytes, a non-WTF statistical analysis should figure out that it probably is.  I don't think adding dictionary checks with several common languages would be an unreasonable addition either.



  • @seaturnip said:

    I think actually you'd be hard-pressed to find a Chinese or Japanese UTF-16 document with no null bytes.  In plain text, there would be one after a carriage return I think.  Richer formats like HTML would have them all over the place in the ASCII-based markup.

    Let me clarify: with a text of this length, THEN it is pretty much guaranteed - we don't have that in this case. Obviosuly, the longer the text, the better your chance is of judging it accurately, but that doesn't help with these shorter ones.

    And 0x20 does not always have to be a space yes, but if there are five in 25 or so bytes, a non-WTF statistical analysis should figure out that it probably is.

    Why? There are plenty of Unicode-characters characters where 0x20 is a byte, and when it, as in this case, all maps to the same character segment, there's no way of knowing. Had some of them generated, say, Arabic, then I might agree, but I believe IsUnicodeText already takes this into account.

    I don't think adding dictionary checks with several common languages would be an unreasonable addition either.

    Then you need to define "several common languages". Where do you draw the limit without inconveniencing people using "uncommon" languages even further? While not the case with the well known "Bush hid the facts" example, what if the bytes used for a specific English text file just happen to a valid sentence in Chinese as well? How do you tell these cases apart without prompting the user (that will just confuse many people), and more importantly, how will you know you did the right thing if your dictionary only turns up matches for English with a Chinese text?



  • @Cap'n Steve said:

    @KenW said:

    Gee, let's see. You think it's a bug that a WINDOWS text editor doesn't understand UNIX line endings. So you're another "let's bash MS for everything that happens, even when it's not their fault" /.er, huh?
     



    Well Wordpad has no problem with it.  I wonder why they haven't just replaced notepad with wordpad, actually.

    For what it's worth (and from what I can remember), Windows 3.1 came with Write (what Wordpad was known as then) as standard, and Notepad as an optional accessory. No, I have no idea why.



  • @ShadowWolf said:

    @Ragnax said:

    @bighusker said:

    @ender said:
    @bighusker said:
    Notepad is awful and has plenty of bugs for a simple friggin' text editor. First of all, there this ridiculous bug as described on the snopes message board: http://msgboard.snopes.com/message/ultimatebb.php?/ubb/get_topic/f/56/t/002928.html. How in the hell do you even pull something like that off?
    This is not a bug in the notepad, but simply the way IsTextUnicode function works. And the limitation's been known for a long time.
    Do other text editors have this problem?  I'd imagine there has to be some reasonable workaround for this.

    There is. It's called; Letting the user decide which encoding to use. Seriously; What's so hard about popping up a dialog box that states the application isn't sure what format the file has and asking the user to decide for it?

    Windows and its bundled applications are made to be usable by braindead idiots, but even that category of user wouldn't have problems with something like that, would they?

    Because when a user opens a text file and they see a dialog that comes up saying "This file is saved in an unrecognizable file format, please select the one you want" they are obviously going to very familiar with the encoding of that file.

    If you knew much about language encoding issues, I doubt you'd say this.  If you do know much about language encoding issues and you're saying this then you're just being elitist.

    There is no "right" solution because every solution is either an annoyance (every time you open File "X" you recieve a warning about how it may not have opened correctly despite that it worked fine) or a presumption of knowledge.

    They don't need to be familiar with the actual encoding of the file. Just provide a short decoded sample for all possible encodings. A similar solution exists for the BitTorrent client Azureus and it works wonderfully well.
     



  • @Ragnax said:

    @ShadowWolf said:
    @Ragnax said:

    @bighusker said:

    @ender said:
    @bighusker said:
    Notepad is awful and has plenty of bugs for a simple friggin' text editor. First of all, there this ridiculous bug as described on the snopes message board: http://msgboard.snopes.com/message/ultimatebb.php?/ubb/get_topic/f/56/t/002928.html. How in the hell do you even pull something like that off?
    This is not a bug in the notepad, but simply the way IsTextUnicode function works. And the limitation's been known for a long time.
    Do other text editors have this problem?  I'd imagine there has to be some reasonable workaround for this.

    There is. It's called; Letting the user decide which encoding to use. Seriously; What's so hard about popping up a dialog box that states the application isn't sure what format the file has and asking the user to decide for it?

    Windows and its bundled applications are made to be usable by braindead idiots, but even that category of user wouldn't have problems with something like that, would they?

    Because when a user opens a text file and they see a dialog that comes up saying "This file is saved in an unrecognizable file format, please select the one you want" they are obviously going to very familiar with the encoding of that file.

    If you knew much about language encoding issues, I doubt you'd say this.  If you do know much about language encoding issues and you're saying this then you're just being elitist.

    There is no "right" solution because every solution is either an annoyance (every time you open File "X" you recieve a warning about how it may not have opened correctly despite that it worked fine) or a presumption of knowledge.

    They don't need to be familiar with the actual encoding of the file. Just provide a short decoded sample for all possible encodings. A similar solution exists for the BitTorrent client Azureus and it works wonderfully well.
     

    Also bear in mind the language the system is using. If it uses the Roman alphabet, there's a good chance files will too. If it's Chinese, the file is more likely to be also.


Log in to reply