time_t

This is a discussion on time_t within the C++ Programming forums, part of the General Programming Boards category; Hi, I cant think of a single reason why is in Unix based systems variable containing time stamp (time_t) type ...

  1. #1
    Registered User
    Join Date
    Apr 2007
    Posts
    162

    time_t

    Hi,

    I cant think of a single reason why is in Unix based systems variable containing time stamp (time_t) type SIGNED.
    Why wasnt this variable declared unsigned so that it could last twice as long as now. Why would time need to have a negative value at all?

  2. #2
    and the hat of wrongness Salem's Avatar
    Join Date
    Aug 2001
    Location
    The edge of the known universe
    Posts
    32,761
    -1 is returned when there is no time available.
    If you dance barefoot on the broken glass of undefined behaviour, you've got to expect the occasional cut.
    If at first you don't succeed, try writing your phone number on the exam paper.
    I support http://www.ukip.org/ as the first necessary step to a free Europe.

  3. #3
    Frequently Quite Prolix dwks's Avatar
    Join Date
    Apr 2005
    Location
    Canada
    Posts
    8,048
    I was always under the impression that the designers chose a signed type so that one could represent differences in time if one wanted to. I mean, you could conceivably want to figure out the difference between two times (end - start or whatever), and what if the result was negative . . . ?

    But that was just uninformed guessing.

    (And of course, one should really use difftime() to get the difference between time_t's instead of subtracting directly.)
    dwk

    Seek and ye shall find. quaere et invenies.

    "Simplicity does not precede complexity, but follows it." -- Alan Perlis
    "Testing can only prove the presence of bugs, not their absence." -- Edsger Dijkstra
    "The only real mistake is the one from which we learn nothing." -- John Powell


    Other boards: DaniWeb, TPS
    Unofficial Wiki FAQ: cpwiki.sf.net

    My website: http://dwks.theprogrammingsite.com/
    Projects: codeform, xuni, atlantis, nort, etc.

  4. #4
    Just Lurking Dave_Sinkula's Avatar
    Join Date
    Oct 2002
    Posts
    5,006
    Fwiw
    there was originally some controversy over whether the unix time_t should be signed or unsigned. If unsigned, its range in the future would be doubled, postponing the 32-bit overflow (by 68 years). However, it would then be incapable of representing times prior to 1970. Dennis ritchie, when asked about this issue, said that he hadn't thought very deeply about it, but was of the opinion that the ability to represent all times within his lifetime would be nice. (ritchie's birth, in 1941, is around unix time −893,400,000.) the consensus is for time_t to be signed, and this is the usual practice.
    7. It is easier to write an incorrect program than understand a correct one.
    40. There are two ways to write error-free programs; only the third one works.*

  5. #5
    Registered User
    Join Date
    Apr 2007
    Posts
    162
    thx for replies, very interesting... I like the most that the guy who declared it actually wanted it to be able to cover his whole lifetime, clever dude :-)

  6. #6
    C++まいる!Cをこわせ!
    Join Date
    Oct 2007
    Posts
    22,915
    Meh, this whole thing is backwards. They could just start the counting earlier and use an unsigned type. This smells like a hack.
    time_t seems like a poor way of dealing with time and date anyway, due to its limited range.
    Quote Originally Posted by Adak View Post
    io.h certainly IS included in some modern compilers. It is no longer part of the standard for C, but it is nevertheless, included in the very latest Pelles C versions.
    Quote Originally Posted by Salem View Post
    You mean it's included as a crutch to help ancient programmers limp along without them having to relearn too much.

    Outside of your DOS world, your header file is meaningless.

  7. #7
    and the hat of wrongness Salem's Avatar
    Join Date
    Aug 2001
    Location
    The edge of the known universe
    Posts
    32,761
    Around 1970, when memory was priced at 1 cent per byte (think about that as you plug a 1GB DIMM module into your PC next time).

    Also that 1 cent would be nearer $1 in today's money, after 40 years of inflation.

    Also that in 1970, the machine that C was first developed on had a stunning 64K of memory.

    Spending lots of money on a problem that was 30 years away wasn't a priority.
    If you dance barefoot on the broken glass of undefined behaviour, you've got to expect the occasional cut.
    If at first you don't succeed, try writing your phone number on the exam paper.
    I support http://www.ukip.org/ as the first necessary step to a free Europe.

  8. #8
    C++まいる!Cをこわせ!
    Join Date
    Oct 2007
    Posts
    22,915
    Aye, hack it is. The good old days when future compatibility simply wasn't considered.
    Quote Originally Posted by Adak View Post
    io.h certainly IS included in some modern compilers. It is no longer part of the standard for C, but it is nevertheless, included in the very latest Pelles C versions.
    Quote Originally Posted by Salem View Post
    You mean it's included as a crutch to help ancient programmers limp along without them having to relearn too much.

    Outside of your DOS world, your header file is meaningless.

  9. #9
    Super Moderator
    Join Date
    Sep 2001
    Posts
    4,913
    This smells like a hack.
    Of course... It's UNIX...

  10. #10
    Registered User
    Join Date
    Oct 2008
    Posts
    1,262
    In 300 years...

    Quote Originally Posted by Salem View Post
    Around 2020, when memory was priced at 1 cent per gigabyte (think about that as you plug a 1PB DIMM module into your PC next time).

    Also that 1 cent would be nearer $1 in today's money, after 300 years of inflation.

    Also that in 2020, the machine that C was used on had a stunning 2GB of memory.

    Spending lots of money on a problem that was 300 years away wasn't a priority.

  11. #11
    C++まいる!Cをこわせ!
    Join Date
    Oct 2007
    Posts
    22,915
    Hang on.
    By these calculations, the inflation should be a lot more! "Fixed" quote below.
    Quote Originally Posted by Salem
    Around 2020, when memory was priced at 1 cent per gigabyte (think about that as you plug a 1PB DIMM module into your PC next time).

    Also that 1 cent would be nearer $7.5 in today's money, after 300 years of inflation.

    Also that in 2020, the machine that C was used on had a stunning 2GB of memory.

    Spending lots of money on a problem that was 300 years away wasn't a priority.
    Quote Originally Posted by Adak View Post
    io.h certainly IS included in some modern compilers. It is no longer part of the standard for C, but it is nevertheless, included in the very latest Pelles C versions.
    Quote Originally Posted by Salem View Post
    You mean it's included as a crutch to help ancient programmers limp along without them having to relearn too much.

    Outside of your DOS world, your header file is meaningless.

  12. #12
    Frequently Quite Prolix dwks's Avatar
    Join Date
    Apr 2005
    Location
    Canada
    Posts
    8,048
    Wouldn't it be at least a QIMM by then? (And wow, who fixed the problem of inflation?) [I see Elysia caught that one too.]

    Well, if you're going to use a single fixed-size integer to represent time, you'll always be running into this problem at some point. It would probably be best to use an insanely wide integer to represent it, but of course that uses lots of resources (and dealing with 16-byte numbers or whatever is a pain anyway). time_t was probably the best thing the designers of UNIX could have done at the time.

    Consider this: by having a signed time_t, we have to deal with the problem sooner, and maybe we'll come up with a better solution this time around. Imagine if it were unsigned, and the UNIX epoch ended about 2106. We'd get around to dealing with it by the time hovercars depended on time_t software . . . .
    dwk

    Seek and ye shall find. quaere et invenies.

    "Simplicity does not precede complexity, but follows it." -- Alan Perlis
    "Testing can only prove the presence of bugs, not their absence." -- Edsger Dijkstra
    "The only real mistake is the one from which we learn nothing." -- John Powell


    Other boards: DaniWeb, TPS
    Unofficial Wiki FAQ: cpwiki.sf.net

    My website: http://dwks.theprogrammingsite.com/
    Projects: codeform, xuni, atlantis, nort, etc.

  13. #13
    C++まいる!Cをこわせ!
    Join Date
    Oct 2007
    Posts
    22,915
    Quote Originally Posted by dwks View Post
    Wouldn't it be at least a QIMM by then? (And wow, who fixed the problem of inflation?) [I see Elysia caught that one too.]
    Yeah, I just couldn't be bothered to draw a proper chart to see how the memory size would expand.
    I can't think of a really proper formula either.
    We know that:
    - In 40 years, our memory size expanded from 64 KB to 2 GB (that's 2 GB - 64 KB).
    - Memory sizes are increasing exponentially.
    - Memory capacity is doubled every 2 years or so.

    So if anyone wants to come up with a proper function to represent a graph that we could use to determine the actual (theoretical) memory size at that point, feel free to do it!

    Well, if you're going to use a single fixed-size integer to represent time, you'll always be running into this problem at some point. It would probably be best to use an insanely wide integer to represent it, but of course that uses lots of resources (and dealing with 16-byte numbers or whatever is a pain anyway). time_t was probably the best thing the designers of UNIX could have done at the time.
    Well, I think they could have started counting earlier and made it unsigned. The whole negative time is like a hack: they added it to represent earlier dates after the date functions had been finalized.
    But I'm more of a fan of the tm struct anyway. It's a very nice and structured way of dealing with time and date.
    Quote Originally Posted by Adak View Post
    io.h certainly IS included in some modern compilers. It is no longer part of the standard for C, but it is nevertheless, included in the very latest Pelles C versions.
    Quote Originally Posted by Salem View Post
    You mean it's included as a crutch to help ancient programmers limp along without them having to relearn too much.

    Outside of your DOS world, your header file is meaningless.

  14. #14
    Frequently Quite Prolix dwks's Avatar
    Join Date
    Apr 2005
    Location
    Canada
    Posts
    8,048
    Well, I think they could have started counting earlier and made it unsigned. The whole negative time is like a hack: they added it to represent earlier dates after the date functions had been finalized.
    Well, sure, using signed numbers solely to represent dates before 1970 is not a good idea. But signed time_ts have other benefits, as described above. It allows -1 to be an error value (though -1u could be in case of unsigned numbers), and differences in time (even negative differences) to be represented more easily.
    dwk

    Seek and ye shall find. quaere et invenies.

    "Simplicity does not precede complexity, but follows it." -- Alan Perlis
    "Testing can only prove the presence of bugs, not their absence." -- Edsger Dijkstra
    "The only real mistake is the one from which we learn nothing." -- John Powell


    Other boards: DaniWeb, TPS
    Unofficial Wiki FAQ: cpwiki.sf.net

    My website: http://dwks.theprogrammingsite.com/
    Projects: codeform, xuni, atlantis, nort, etc.

  15. #15
    C++まいる!Cをこわせ!
    Join Date
    Oct 2007
    Posts
    22,915
    I don't find that to be much of benefit.
    First off, all functions should really take an argument and return bool, or some other type to indicate failure instead of making it signed and having all negative numbers mean failure (waste of bits!).
    Secondly, signed mathematics on unsigned types should work:

    typedef unsigned int uint;
    uint SomeNum1, SomeNum2;
    int Diff = (int)(SomeNum2 - SomeNum1);

    I've left out the initialization on purpose.
    Quote Originally Posted by Adak View Post
    io.h certainly IS included in some modern compilers. It is no longer part of the standard for C, but it is nevertheless, included in the very latest Pelles C versions.
    Quote Originally Posted by Salem View Post
    You mean it's included as a crutch to help ancient programmers limp along without them having to relearn too much.

    Outside of your DOS world, your header file is meaningless.

Page 1 of 2 12 LastLast
Popular pages Recent additions subscribe to a feed

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21