Bug 10011

Created attachment 3854 [details]
patch adding first_weekday and first_workday setting to de_AT locale

General policy for these changes is to require pointers proving this - can you
please provide URL of some official institution showing this or so?

It's hard to prove with URLs from official institutions that Austrian calendars
should start in Mondays.



There are „Ö-Normen” (ie. Austrian norms) dictating the start of the week. But
first_weekday is not the start of the week -- week[2] would do that job.

But even if ... those Austrian norms are not available for free. I could not
provide you with a copy of such a norm. The only URLs I can provide are URLs to
pages, where you can buy those norms.



I could take pictures of calendars used in Austria to prove that calendars start
in Mondays. But I assume you do not want me to mess up the bug tracker with
several binary attachments.



Would it do, if I can mobilize some Austrian people to add a comment to the bug
saying that they support the patch of changing first_weekday?



For first_workday it's even more of a problem. There, I do not even know of any
norm dictating this. I searched Austrian norms and laws but could not find anything.

However, the work week of every Austrian worker that I know starts on monday
(except for shift workers).

Again, would it do, if I can mobilize some Austrian people to add a comment to
the bug, saying that they support the patch of changing first_weekday?



How many supporting comments would it take to get this patch approved?

Subject: Re:  wrong week/first_weekday/first_workday for de_AT locale

On Wed, Apr 01, 2009 at 08:35:09AM -0000, flyer dot REMOVE dot SPAM at gSmPaAiMl dot com wrote:
> 
> ------- Additional Comments From flyer dot REMOVE dot SPAM at gSmPaAiMl dot com  2009-04-01 08:35 -------
> 
> It's hard to prove with URLs from official institutions that Austrian calendars
> should start in Mondays.
> 
> 
> 
> There are „Ö-Normen” (ie. Austrian norms) dictating the start of the week. But
> first_weekday is not the start of the week -- week[2] would do that job.
> 
> But even if ... those Austrian norms are not available for free. I could not
> provide you with a copy of such a norm. The only URLs I can provide are URLs to
> pages, where you can buy those norms.
> 
> 
> 
> I could take pictures of calendars used in Austria to prove that calendars start
> in Mondays. But I assume you do not want me to mess up the bug tracker with
> several binary attachments.
> 
> 
> 
> Would it do, if I can mobilize some Austrian people to add a comment to the bug
> saying that they support the patch of changing first_weekday?
> 
> 
> 
> For first_workday it's even more of a problem. There, I do not even know of any
> norm dictating this. I searched Austrian norms and laws but could not find anything.
> 
> However, the work week of every Austrian worker that I know starts on monday
> (except for shift workers).
> 
> Again, would it do, if I can mobilize some Austrian people to add a comment to
> the bug, saying that they support the patch of changing first_weekday?
> 
> 
> 
> How many supporting comments would it take to get this patch approved?

I agree that Monday is the first weekday in Austria.
You should then list the day as 1 for monday, and define monday as the 
day one in the week. The current patch lists the day as 2, which is not
what is intended by the system.

Best regards
keld

FYI:
1. EN 28601, which Austria follows, makes a statement to the effect that Monday
is the preferred first day of the week.

2. CLDR says Austria uses Monday as the first day of the week. See the
"firstDay" element. 
<firstDay day="mon"  territories="001" />
001 is the default territory, so since AT doesn't have an entry for any other
day, "mon" applies.

Hello,

this bug is taking off in an unwanted direction.
This bug is _not_ about correcting the first day of a week in de_AT (which would
correspond to week[2]).
This bug is about correcting first_weekday (the day calendars should start a
week in) in de_AT.

Some commentors said, that in Austria weeks are Mondays based (note the
difference to first_weekday). This is true.
Currently de_AT bases weeks on Sundays.
This bug bug however does not address this mismatch. Setting de_AT to Monday
based, breaks the output of several applications (date, cal, ...), as for
example strftime does not work properly with Monday based weeks. In the end I
want to overcome this limitation and base Austrian weeks on Monday. But that's
out of scope of this bug. Getting things right requires three steps:
1. Provide a short-term fix to get calendars showing the correct start of a week
(This is the purpose of this bug).
2. Correct strftime to treat Monday based weeks nicely (This will become a bug
on its own).
3. For countries having Monday based weeks (just as Austria), adapt
corresponding locales (This will become other, separate bugs).

This bug's sole purpose is to provide a short-term fix for calendars.
Let's keep it at that.

Kind regards,
Paul

Hello,

according to Pesky's comment in [1] I'd argue that, comment #4 of this bug does not meet the current 
policy on weeks not starting on Sundays and can therefore be left alone. Same for comment #6.
According to [1], the week specifier should be set as well.
The new patch takes care of this.

Kind regards,
Paul 


[1]
http://sourceware.org/ml/libc-locales/2009-q2/msg00032.html

Created attachment 3943 [details]
patch adding week, first_weekday and first_workday setting to de_AT locale

I've added the patch to git.