Re: PersianComputing Digest, Vol 21, Issue 6 (fwd)
On Fri, 25 Feb 2005, Skip Tavakkolian wrote: I'm not sure how the date data type can be representation agnostic. What ever the OS provides (via a system call) is in reference to a starting point in some calendar. On UNIX systems, this is traditionally the number of seconds since January 1, 1970, i.e. Gregorian. GetSystemTime on Win32 returns a structure, which represents the Gregorian date. The UNIX epoch can easily defined as the number of seconds since 11 Dey 1348. The important data is a date in the sense of a point in the axis of time. How you write it out, depends on one speific representation though. --behdad http://behdad.org/ ___ PersianComputing mailing list PersianComputing@lists.sharif.edu http://lists.sharif.edu/mailman/listinfo/persiancomputing
Re: PersianComputing Digest, Vol 21, Issue 6 (fwd)
On Wed, 23 Feb 2005, mohsen ali momeni wrote: Now something else , For AddDate and DateDiff functions, I need an algorithm which calculates the number of leap years between two given Date. Is there any such algorithm or at least a documentation for the above algorithms (jalali.c) so that i can find it in the code myself? (Or AddDate, DateDiff functions ready in ideal case) Regards, Mohsen A. Momeni Well, that's why I'm saying your implementation is not what MySQL people expect. The date data type is representation-agnostic itself, and AddDate, DateDiff, etc work with the date data type (at least in MySQL). What you need is functions to covert from internal date representation to Iranian calendar string, and vice versa. You don't need (and should not) implement all date functions again. --behdad http://behdad.org/ ___ PersianComputing mailing list PersianComputing@lists.sharif.edu http://lists.sharif.edu/mailman/listinfo/persiancomputing
Re: PersianComputing Digest, Vol 21, Issue 6 (fwd)
On Wed, 23 Feb 2005, mohsen ali momeni wrote: Now something else , For AddDate and DateDiff functions, I need an algorithm which calculates the number of leap years between two given Date. Is there any such algorithm or at least a documentation for the above algorithms (jalali.c) so that i can find it in the code myself? (Or AddDate, DateDiff functions ready in ideal case) Regards, Mohsen A. Momeni Well, that's why I'm saying your implementation is not what MySQL people expect. The date data type is representation-agnostic itself, and AddDate, DateDiff, etc work with the date data type (at least in MySQL). What you need is functions to covert from internal date representation to Iranian calendar string, and vice versa. You don't need (and should not) implement all date functions again. --behdad http://behdad.org/ I'm not sure how the date data type can be representation agnostic. What ever the OS provides (via a system call) is in reference to a starting point in some calendar. On UNIX systems, this is traditionally the number of seconds since January 1, 1970, i.e. Gregorian. GetSystemTime on Win32 returns a structure, which represents the Gregorian date. ___ PersianComputing mailing list PersianComputing@lists.sharif.edu http://lists.sharif.edu/mailman/listinfo/persiancomputing
Re: PersianComputing Digest, Vol 21, Issue 6 (fwd)
Hi , I talked to Roozbeh Pournader about your mail. Seems like you are using an old version of the code, which has the problem as you mentioned. Our latest code is available from address below and doesn't have that problem: http://www.farsiweb.info/jalali/jalali.c Thanks for your attention. I will apply this new algorithm now and will send a patch again to mysql in the hope they accept it. Now something else , For AddDate and DateDiff functions, I need an algorithm which calculates the number of leap years between two given Date. Is there any such algorithm or at least a documentation for the above algorithms (jalali.c) so that i can find it in the code myself? (Or AddDate, DateDiff functions ready in ideal case) Regards, Mohsen A. Momeni ___ PersianComputing mailing list PersianComputing@lists.sharif.edu http://lists.sharif.edu/mailman/listinfo/persiancomputing
Re: PersianComputing Digest, Vol 21, Issue 6
On Sat, 19 Feb 2005, mohsen ali momeni wrote: No. Wrong. So you say we should still fight about our calender name? I mean yes, if we have not come up with a name yet, we can continue discussion, of course you are free to call it fight or whatever. No. They simply are not interested in your functions. They were interested, as their first email showed that. They accept it but i got no answer after that. Then your implemention has been poor. Again no. Calendars does not belong to databases. I didn't say they belong to databases but having calender functions in mysql will make these calcultions much faster in programs instead of doing them in php. No. A C module for PHP is as fast, if not faster, than doing them in MySQL, which is apparently the wrongest place you can implement them. What about in the kernel then? Faster? --behdad http://behdad.org/ ___ PersianComputing mailing list PersianComputing@lists.sharif.edu http://lists.sharif.edu/mailman/listinfo/persiancomputing
