[Info-vax] [OT] Linux vs Windows vs OS X. Was Re: Unix on A DEC Vax?

[Bill Gunshannon]

In article <kdjvuj$1eg$1 at dont-email.me>,
	Simon Clubley <clubley at remove_me.eisner.decus.org-Earth.UFP> writes:
> On 2013-01-21, Bill Gunshannon <billg999 at cs.uofs.edu> wrote:
>> In article <kdh5nh$pte$1 at dont-email.me>,
>> 	Simon Clubley <clubley at remove_me.eisner.decus.org-Earth.UFP> writes:
>>> 
>>> Only an American could have made that comment. :-)
>>
>> How about an American who has spent a considerable amount of his
>> life living outside the US? :-)
>>
> 
> With your later comments, I cannot tell any longer if you don't know
> about these routines or just don't consider them a part of the OS.

What routines?  The ones that display time and currency symbol?
I see those as application specific.  

> 
>>
>>> 
>>> The country specific attributes (currency symbols, date format, etc) are
>>> set at operating system level and the application is expected to acquire
>>> those values from the machine it is running on.
>>
>> Hmmmm...  Linux (which is what we were talking about) stores time as 
>> seconds since the epoch.  All displays are at the application level.
>> Timezone (which controls the display) is a file well outside the 
>> kernel (and thus the OS).  I can change the timezone and display of
>> time without making any change to the OS, re-compilling or even re-
>> booting.   I know for a fact that I can set my Currency Symbol in a
>> wordprocessor or spreadsheet program to be different than other users
>> of the same machine.
>>
>>> 
>>> These settings, while they can be changed at application level in some
>>> applications, most certainly are not application specific.
>>
>> I know of no kernel structure in Linux or any version of Unix I have ever
>> worked with that contained "Currency Symbol".  That has always been an
>> application setting. (Even COBOL let you set that yourself in the source
>> to your programs.)
>>
> 
> So you don't consider the C RTL and environment variables to be part of
> the OS ?

No, not really.  Let's look at the layout of the man pages.  Section 2
is system calls.  Section 3 is library calls.  I consider section 2 as
part of the OS, section 3 not so much.  Think of it like rings.  Or,
mayber use the OSI model.  :-)

> 
> So something is not a part of the OS unless it needs a kernel structure
> to support it ?

No, didn't say that.  What I said was the date is tored exactly the same
in a Unix machine in Germany as it is in a Unix machine in the US.  How
that date is presented tot he user is an application thing.  The "date"
command is not part of the OS it is a user level application.

> 
> That's so out of touch with reality, I don't really know how to respond. :-)

OK.  You wouldn't be the first one to tell me I am out of touch with
reality.

> 
> In my world, I consider a application specific package to be, for example,
> a word processor, a email client or a game. 

And each of those gets to decide on its own how it will present dates
and currency symbols, regardless of what number is stored in that counter
called time_t.

>                                              I do not consider it to be
> the base OS supplied support routines used to implement those applications.

time_t is the only time value returned by the OS.  What is done with that
value is determined by various programs which I don't consider to be part
of the OS as the OS really doesn't require any of them in order to function.

But, getting back to the original question which had to to with the
difficulty or lack thereof changing this under Ubuntu, I found it
as a simple selection from a dropdown in "Language Support".

bill

-- 
Bill Gunshannon          |  de-moc-ra-cy (di mok' ra see) n.  Three wolves
billg999 at cs.scranton.edu |  and a sheep voting on what's for dinner.
University of Scranton   |
Scranton, Pennsylvania   |         #include <std.disclaimer.h>