This is getting interesting ;-)

Morocco doesn't use Lisbon time.

After spending all morning looking at zic and the files at
ftp://elsie.nci.nih.gov/pub
I have compiled a new Africa/Casablanca file.


Then I decided to look at what is on my PC.
It looks like the dst values are correct on my Ubunutu PC:
$ zdump -v Africa/Casablanca
[SNIP]
Africa/Casablanca  Sat Apr  2 23:59:59 2011 UTC = Sat Apr  2 23:59:59 2011 WET
isdst=0 gmtoff=0
Africa/Casablanca  Sun Apr  3 00:00:00 2011 UTC = Sun Apr  3 01:00:00 2011 WEST
isdst=1 gmtoff=3600
Africa/Casablanca  Sat Jul 30 22:59:59 2011 UTC = Sat Jul 30 23:59:59 2011 WEST
isdst=1 gmtoff=3600
Africa/Casablanca  Sat Jul 30 23:00:00 2011 UTC = Sat Jul 30 23:00:00 2011 WET
isdst=0 gmtoff=0
[SNIP]

So, for Casablanca the variable is going to have to be:
TZ_TIMEZONE ="WET0WEST,M4.1.0/0,M7.5.6/23"
(tell me if I got it wrong - this is not intuitive when you're doing it for the
first time, like me).

WET = designation for standard time when daylight saving is not in force
0 = offset in hours = 0 hours west of Greenwich meridian (i.e. UTC)
WEST = designation when daylight saving is in force
, = no offset number between code and comma, so default to one hour ahead for
daylight saving
M4.1.0 = when daylight saving starts = the 0th day (Sunday) in the first week of
month 4 (April)
/0, = the local time when the switch occurs = 00:00 in this case
M7.5.6 = when daylight saving ends = the 6th day (Saturday) in the last week of
month 7 (July).
/23, = the local time when the switch occurs = 23:00 in this case


Since the files on "elsie" are up-to-date, someone who knows what they are doing
should be able to create a new Africa/Casablanca file and since this is such an
interesting case I'm sure you'd be glad to add this to the list in the GUI for
the next release [wouldn't you? ;-)]

IMPORTANT:
As strange thing is that it doesn't look like /etc/TZ can be created as a simple
tail -n1 of /usr/share/zoneinfo/Africa/Casablanca because as I said, on my
Ubuntu machine, the zdump (above) looks correct.
If I tail -n1 /usr/share/zoneinfo/Africa/Casablanca the result is still "WET0".


Thanks again!

-Graham-


Lonnie Abelbeck wrote on 11/05/11 03:35:
> Hi Graham,
> 
> This timezone stuff normally just works, but there appears to be a problem 
> with the Africa/Casablanca zoneinfo file.
> 
> A little background...
> 
> The timezone files are in "/usr/share/zoneinfo/" directory tree.  For example:
> --
> # ls -l /usr/share/zoneinfo/Europe/Berlin 
> -rw-r--r--    1 root     root         2309 May 10 14:24 
> /usr/share/zoneinfo/Europe/Berlin
> 
> # tail -n1 /usr/share/zoneinfo/Europe/Berlin
> CET-1CEST,M3.5.0,M10.5.0/3
> --
> In this case the /etc/TZ file is automatically set (at startup) to 
> "CET-1CEST,M3.5.0,M10.5.0/3"
> 
> Now look at the Africa/Casablanca file.
> --
> # tail -n1 /usr/share/zoneinfo/Africa/Casablanca 
> WET0
> --
> which appears incorrect, since daylight savings time seems to be supported.
> 
> In AstLinux you can define the TZ_TIMEZONE variable to override the automatic 
> "WET0" value set to /etc/TZ .
> 
> You might try: (add in Network Tab -> User System Variables: { Edit User 
> variables } )
> --
> TZ_TIMEZONE ="WET0WEST,M3.5.0/1,M10.5.0"
> --
> Which is what Europe/Lisbon uses.
> 
> Hope this helps, maybe some googling you can find the correct /etc/TZ value.
> 
> Lonnie
> 
> PS: Don't forget this rule in AstLinux:
> http://doc.astlinux.org/userdoc:tt_editable_files
> 
> 
> On May 10, 2011, at 6:39 PM, Graham S. Jarvis wrote:
> 
>> Thanks Lonnie,
>>
>> 'thought so.
>>
>> I'm actually building a system for Africa/Casablanca but this file didn't 
>> work
>> because it doesn't have the dst changes for summertime.
>> So I used WET which seems to set the time OK - even if not on the status 
>> page of
>> the GUI (see my other post).
>> BUT, it doesn't have the correct change over dates for Casablanca.
>> (See http://www.timeanddate.com/worldclock/city.html?n=60)
>> BTW neither does Africa/Ceuta (a bit of Spain in the Mahgrib).
>>
>> I now need to find a good timezone file with the correct changeover dates . 
>> . .
>> (Others have the same problem:
>> http://lists.mandriva.com/bugs/2010-04/msg02890.php)
>>
>> Anyone know of a website that does on-line compilation of timezone files???
>>
>> We'll get there!
>>
>> Thanks,
>>
>> -Graham-
>>
>>
>> Lonnie Abelbeck wrote on 10/05/11 14:44:
>>>
>>> On May 10, 2011, at 3:50 AM, Graham S. Jarvis wrote:
>>>
>>>> Hello all,
>>>>
>>>> Under the Network Tab of the GUI there are number of Timezones listed.
>>>> Where does this list come from?
>>>> How do I change this list?
>>>
>>> You can't, but I made an effort to cover all timezones with major cities 
>>> around the world.  Both Brussels and Berlin should be the same as 
>>> Luxembourg.
>>>
>>>
>>>> Is there really a relationship between the Timezone and the choice of NTP 
>>>> server?
>>>
>>> No.
>>>
>>>
>>>> I thought the NTP server is chosen based on it's closeness simply so that 
>>>> the
>>>> workload is spread.
>>>
>>> In general, the closer the NTP server, the more accurate the time sync.
>>>
>>> Lonnie
>>>
>>>
>>>>
>>>> Thanks,
>>>>
>>>> -Graham-
>>>
>>> ------------------------------------------------------------------------------
>>> Achieve unprecedented app performance and reliability
>>> What every C/C++ and Fortran developer should know.
>>> Learn how Intel has extended the reach of its next-generation tools
>>> to help boost performance applications - inlcuding clusters.
>>> http://p.sf.net/sfu/intel-dev2devmay
>>> _______________________________________________
>>> Astlinux-users mailing list
>>> Astlinux-users@lists.sourceforge.net
>>> https://lists.sourceforge.net/lists/listinfo/astlinux-users
>>>
>>> Donations to support AstLinux are graciously accepted via PayPal to 
>>> pay...@krisk.org.
>>>
>>
>> ------------------------------------------------------------------------------
>> Achieve unprecedented app performance and reliability
>> What every C/C++ and Fortran developer should know.
>> Learn how Intel has extended the reach of its next-generation tools
>> to help boost performance applications - inlcuding clusters.
>> http://p.sf.net/sfu/intel-dev2devmay
>> _______________________________________________
>> Astlinux-users mailing list
>> Astlinux-users@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/astlinux-users
>>
>> Donations to support AstLinux are graciously accepted via PayPal to 
>> pay...@krisk.org.
>>
>>
> 
> 
> ------------------------------------------------------------------------------
> Achieve unprecedented app performance and reliability
> What every C/C++ and Fortran developer should know.
> Learn how Intel has extended the reach of its next-generation tools
> to help boost performance applications - inlcuding clusters.
> http://p.sf.net/sfu/intel-dev2devmay
> _______________________________________________
> Astlinux-users mailing list
> Astlinux-users@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/astlinux-users
> 
> Donations to support AstLinux are graciously accepted via PayPal to 
> pay...@krisk.org.
> 

------------------------------------------------------------------------------
Achieve unprecedented app performance and reliability
What every C/C++ and Fortran developer should know.
Learn how Intel has extended the reach of its next-generation tools
to help boost performance applications - inlcuding clusters.
http://p.sf.net/sfu/intel-dev2devmay
_______________________________________________
Astlinux-users mailing list
Astlinux-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/astlinux-users

Donations to support AstLinux are graciously accepted via PayPal to 
pay...@krisk.org.

Reply via email to