About Word on Macintosh, you are talking about a *near full* disk, right?
There isn't enouugh *space* for a new file, it *has* to destroy the old file before writing to the disk.
Frankly, yes, it *should* make some check that the size_of_free_space - size_of_old_file >= size_of_new_file, and inform you about it, but that isn't quite what you are talking about.
Maybe on 9x, but certainly not on 2K.
Regarding devices' names, Windows usually gives you the providers' name for it, like Maxtor 5t040h4 for my HD, but internally it uses names like \Device\HardDisk0, etc, very like *nix.
http://www.sysinternals.com/ntw2k/freeware/winob j. shtml
Is a small utility that will shows you what Windows uses internally.
For example, I've a \Device\Floppy0 , \Device\HarddiskVolume1, \Device\Scsi\, etc.
Giving the devices the manufactors names is usually a good thing, because those are usually more meaningful.
You are only partially correct, the solution is to do it on very low level.
Why should sed be aware of the forking in the file, that is a matter for:
A> Programs that use it. (Pretty high level ones, usually, like shells or explorer-clones.)
B> The file system driver itself.
On NT, the data is stored on one place, and irrelevent stuff on another.
IE, $DATA contains the data itself, and other streams contain the meta data.
Sed would read the $DATA steam unless you spesify otherwise.
And you spesify otherwise by using a special delimiter to the filename, leaving sed totally unaware of it.
A good example of how it should work is NTFS, if I want to edit myfile.txt in notepad, I enter the file name.
If I want to edit the metadata in notepad, I will open "myfile.txt:metadata".
You need to change just the API.
The problem is, as you mentioned, when you are trying to transfer files from stream-aware FS to non stream-aware FS.
Or to many stream FS (NTFS) to signle stream FS (Mac's resource forks, I believe that OSX has unlimited streams, though).
I think it's possible that is the above happens, then a stuffit (Mac people would know what it is) or a tar (the file format support stream as it is, but the programs don't) solution should occur automatically. (And vice versa, of course).
You would probably need something like cpyfrk() as well as copy() alone, but I think that it should be enough.
No, that isn't what it was.
C was supposed to be a better (portable, mostly) assembly, a lot of its contracsts can be translated directly to proccessor instructions.
The most obvious thing is, of course, ++ and += type of thing.
You must also remember that C as we know it today (C99) isn't at all what K&R C was.
C actually got a lot of things from Pascal (function signatures, and variable types, frex) that made it a language much more suitable to write applications than it used to be. There was a reason why C was used to low level stuff up until the 90s, and why Pascal (or Borland's version of it) was so popular for so long.
As a note, Ada programmers have contests for *clear* code.
In a language that was designed to be as clear as possible.
Now, show me the place where I can buy a Ford car "out of the box" which is like this?
Custom stuff is irrelevent, because you can do that for Windows as well.
And just *why* would I want to do that if I can just go and *hire* programmers?
That way, I wouldn't have to *hope* that someone would take my offer.
Yes, it may be more expensive, but at the end, it will be something that I need, not what somebody else think I need.
Slashdot Mirror
Actually, that isn't what you'll call a multidimensional array, you know.
It works the same, though.
Yes, indeed.
That is a remant from Word 1.0 or so, I believe.
No, it had it since 95 (ver 6 or so, I think) at least, and probably before it.
They just never advertised it so much.
About Word on Macintosh, you are talking about a *near full* disk, right?
There isn't enouugh *space* for a new file, it *has* to destroy the old file before writing to the disk.
Frankly, yes, it *should* make some check that the size_of_free_space - size_of_old_file >= size_of_new_file, and inform you about it, but that isn't quite what you are talking about.
14,517$, as a matter of fact
1 pound == 1.45$
A word of warning, EveryThing Works Likes This.
That isn't just about IIS, it's about every part of software making.
Maybe on 9x, but certainly not on 2K.
b j. shtml
Regarding devices' names, Windows usually gives you the providers' name for it, like Maxtor 5t040h4 for my HD, but internally it uses names like \Device\HardDisk0, etc, very like *nix.
http://www.sysinternals.com/ntw2k/freeware/wino
Is a small utility that will shows you what Windows uses internally.
For example, I've a \Device\Floppy0 , \Device\HarddiskVolume1, \Device\Scsi\, etc.
Giving the devices the manufactors names is usually a good thing, because those are usually more meaningful.
You are only partially correct, the solution is to do it on very low level.
Why should sed be aware of the forking in the file, that is a matter for:
A> Programs that use it. (Pretty high level ones, usually, like shells or explorer-clones.)
B> The file system driver itself.
On NT, the data is stored on one place, and irrelevent stuff on another.
IE, $DATA contains the data itself, and other streams contain the meta data.
Sed would read the $DATA steam unless you spesify otherwise.
And you spesify otherwise by using a special delimiter to the filename, leaving sed totally unaware of it.
A good example of how it should work is NTFS, if I want to edit myfile.txt in notepad, I enter the file name.
If I want to edit the metadata in notepad, I will open "myfile.txt:metadata".
You need to change just the API.
The problem is, as you mentioned, when you are trying to transfer files from stream-aware FS to non stream-aware FS.
Or to many stream FS (NTFS) to signle stream FS (Mac's resource forks, I believe that OSX has unlimited streams, though).
I think it's possible that is the above happens, then a stuffit (Mac people would know what it is) or a tar (the file format support stream as it is, but the programs don't) solution should occur automatically. (And vice versa, of course).
You would probably need something like cpyfrk() as well as copy() alone, but I think that it should be enough.
No, that isn't what it was.
C was supposed to be a better (portable, mostly) assembly, a lot of its contracsts can be translated directly to proccessor instructions.
The most obvious thing is, of course, ++ and += type of thing.
You must also remember that C as we know it today (C99) isn't at all what K&R C was.
C actually got a lot of things from Pascal (function signatures, and variable types, frex) that made it a language much more suitable to write applications than it used to be. There was a reason why C was used to low level stuff up until the 90s, and why Pascal (or Borland's version of it) was so popular for so long.
As a note, Ada programmers have contests for *clear* code.
In a language that was designed to be as clear as possible.
How do you know that it didn't?
--
Two witches watched two watches.
Now, show me the place where I can buy a Ford car "out of the box" which is like this?
Custom stuff is irrelevent, because you can do that for Windows as well.
--
Two witches watched two watches.
I'm not sure if you are aware of it, but that is what 98 did.
--
Two witches watched two watches.
Actually, they never said that you can't add icons to the desktop.
What they refused to do is to allow you to take their own icons out.
--
Two witches watched two watches.
That is the real bummer.
I might try it, but it has very low chance of recognizing my hardware, so I can't event try.
--
Two witches watched two watches.
Only on 9x.
It's unlimited size (as well as much more freindly) on NT
--
Two witches watched two watches.
I mean, Microsoft has Atom Bombs, and so many Open Source people at one place...
I'm not sure that Bill Gates can handle the temptation.
--
Two witches watched two watches.
It was Cain!
Cain did it!
--
Two witches watched two watches.
And just *why* would I want to do that if I can just go and *hire* programmers?
That way, I wouldn't have to *hope* that someone would take my offer.
Yes, it may be more expensive, but at the end, it will be something that I need, not what somebody else think I need.
--
Two witches watched two watches.
You might have meant it as a joke, but there *are* several instances where BOOL doesn't equal FALSE or TRUE in Windows!
--
Two witches watched two watches.
No, bugs are exactly what he said.
The computer does *exactly* what the programmer told it to do.
--
Two witches watched two watches.
I think he should counter sue Adobe for causing his arrest.
If nothing else, they will compromise outside of court to avoid the bad PR.
--
Two witches watched two watches.
PocketPC is CE, AFAIK.
And not even MS is stupid enough to base something on 9x.
CE started as a NT based project.
--
Two witches watched two watches.
Nope. You *can* redistrubute it, in original or modified from.
You just can't sell it.
--
Two witches watched two watches.
Just as long as it wouldn't be smart enough to insult my driving.
--
Two witches watched two watches.
No, it isn't.
You aren't forced to Share Source your own code.
--
Two witches watched two watches.