Why don't European highway systems use shields a kin North American systems?

[agentsteel53]

At work, I've been saving reports as filename MM-DD-YYYY, e.g. 04-06-2012 for April 6, 2012.   Now that we've turned over the calendar to 2013, I wish I had been saving them as YYYY-MM-DD; I  might go back and rename them when I'm bored at work some day.  However, I find a disadvantage in that, with the year first, it's not necessarily intuitive that the next number is the month.  2013-01-02 could just as easily be February 1 as January 2, because nobody actually writes the date that way in everyday life.

my photos tend to be sorted as "20121118", as that allows Windows to sort folders in the intended order.

[J N Winkler]

At work, I've been saving reports as filename MM-DD-YYYY, e.g. 04-06-2012 for April 6, 2012.   Now that we've turned over the calendar to 2013, I wish I had been saving them as YYYY-MM-DD; I  might go back and rename them when I'm bored at work some day.

You don't need to do them one by one--it is fairly straightfoward to write a batch file to handle that for you.  It just needs code to parse the filenames, extract the part that is the date, and then make the appropriate reversal of digit sequences.  This fairly simple script should work as long as you can guarantee that the date suffix, when present, always occurs at the end of the filename, and always follows the MM-DD-YYYY syntax:

Code Select Expand

SETLOCAL ENABLEDELAYEDEXPANSION

dir /a:-d /o /b > temp1.txt
findstr /r [0-9][0-9]-[0-9][0-9]-[0-9][0-9][0-9][0-9] temp1.txt > temp2.txt
FOR /F "tokens=1 delims=" %%A IN (temp2.txt) DO (
set FNBase=%%~nA
set FNBase=!FNBase:~0,-10!!FNBase:~-4,4!-!FNBase:~-10,3!!FNBase:~-7,2!
move "%%A" "!FNBase!%%~xA"
)
del temp*.txt

ENDLOCAL

However, I find a disadvantage in that, with the year first, it's not necessarily intuitive that the next number is the month.  2013-01-02 could just as easily be February 1 as January 2, because nobody actually writes the date that way in everyday life.

I don't think that confusion actually occurs in practice.  When a date string is incorporated into otherwise identically named files and the year is the first part of the string, that is a strong signal that the person who generated the files intends a name sort to result in a correct ordering by date, which obviously does not happen if the day rather than the month is the second digit group.

[agentsteel53]

I don't think that confusion actually occurs in practice.  When a date string is incorporated into otherwise identically named files and the year is the first part of the string, that is a strong signal that the person who generated the files intends a name sort to result in a correct ordering by date, which obviously does not happen if the day rather than the month is the second digit group.

indeed - I don't know of any locality which uses YYYY-DD-MM.  "2012, 6th March" sounds awful.

[kphoger]

Right.  I know it's never said like that.  But my brain always takes a few seconds to sort it out when the year comes first.

[Duke87]

Whenever I put the date in any filename I always do it YYYYMMDD or YYYY-MM-DD since that is what you have to do to make sorting files in alphabetical order give you dates in chronological order. So it doesn't seem at all unnatural to me and I have no trouble parsing it.

I do have to mentally adjust come fall, though, since after having spent the last nine months putting a zero before the number of the month I tend to instinctively want to type October as "010" when it should just be "10".
I mean hey, when mentally you say to yourself "oh one, oh two, oh three, oh four, oh five, oh six, oh seven, oh eight, oh nine...", "oh ten" seems to logically come next.

[mgk920]

For several years now I have been writing all of my dates in the YYYY-MM-DD format as my default.  It just makes life easier for me.  And yes, I have little real use for the USA's oddball customary 'MM-DD-YYYY' format.

Mike

[kkt]

At work, I've been saving reports as filename MM-DD-YYYY, e.g. 04-06-2012 for April 6, 2012.   Now that we've turned over the calendar to 2013, I wish I had been saving them as YYYY-MM-DD; I  might go back and rename them when I'm bored at work some day.  However, I find a disadvantage in that, with the year first, it's not necessarily intuitive that the next number is the month.  2013-01-02 could just as easily be February 1 as January 2, because nobody actually writes the date that way in everyday life.

The trouble with mm/dd/yyyy is that half the world uses dd/mm/yyyy and the other half uses mm/dd/yyyy, and you can't generally tell which was intended.  So when handwriting or in text, I spell out the name of the month, or use three letter abbreviations.  Non-English speakers may not know which month I mean, but at least they realize it's the month and not the day.

For filenames, yyyymmdd seems pretty clear to me because it's in logical progression from larger units to smaller, much like place value or hours, minutes, and seconds.  Also, as noted, when sorted lexically it puts the files in date order.  I don't know anyplace where yyyyddmm is in use.

[Molandfreak]

Decreasing order makes the most sense to me, because that's how we measure the time of day.

12-12-12; 12:12. In my mind, that should read 2012, December 12; 12:12 PM.

Second would be the British system because, though it does defy this rule, it is not completely random and is an increase in units.

I hate the American system, because it is totally random, and doesn't follow a pattern...

[spmkam]

But does any operating system offer sort by date making this moot

[Duke87]

But does any operating system offer sort by date making this moot

No, for several reasons.

First, the date the file was created or last modified is not necessarily the date desired in the filename. Files might be created or uploaded on a date other than the day they are current as of. And any time a file is moved, patched, or has to be restored from a backup, the date created/last modified becomes inaccurate.

Second, there may be multiple files in a folder that I want sorted by something other than the date before the date.
Basic example: I might have in a folder, sorted by name, the following:
Blue20090110
Blue20110720
Blue20120914
Green20100424
Green20120501
Red20091116
Red20110603
Red20121206

Assuming the dates created/modified line up correctly (which I guarantee you in any real world example they will not), sorting the folder by date I will get this:
Blue20090110
Red20091116
Green20100424
Red20110603
Blue20110720
Green20120501
Blue20120914
Red20121206

which is a mess.

[Scott5114]

Linux offers a program called "touch" that can be used to alter the last modified date to whatever you'd like it to be, although it's much easier to just keep it in the file name if it's that important.

I usually use the ISO format (2013-02-08) wherever I can, although I am required not to do so at work, since accounting sent my boss an email several years back along the lines of "Please tell Scott to stop putting the year first on dates on jackpot paperwork." In that case, I tend toward the format 02-08-2013, which happens to be the same format used on Oklahoma driver licenses, which is a source of some of the date information I have to record. I would use "8 Feb 13" or something like that, except since most of the dates I work with are in numerical form to begin with, and I don't want to deal with the mental overhead of translating between words and numbers, minimal though it may be.

[Sykotyk]

Every file I make that I'm naming by date is YYYYMMDD. I never use dashes. I also catalog a lot of things for business records, and I use YYYMMDD-# (increasing in number for each occurrence).

The reason Americans don't use the British Style is partly anti-imperialism, but mostly the year isn't considered important.

If we're talking about many different dates in 2013 coming up... which makes more sense:

02/10
03/04
05/08
06/05
06/07
08/02
09/01
09/03
10/02
11/04

Most important part of the date would come first (month), followed by it's secondary characteristic (day).  The year is just thrown on last as a means of keeping it situated if there's more than one year being used.

To me, I always felt the British way was maddening.

You wouldn't have 48:02 for your clock, would you?

[J N Winkler]

There is a related issue--formatting count numbers to be compatible with different ordering systems.  Versions of Windows up to about XP used straight character collation, so 1, 11, 111, 1111, etc. all come before 2.  Versions of Windows from XP onward use string comparison in Explorer, which mimics natural-number ordering (1, 2, 3, . . ., 11,  . . ., 111) but requires the OS to identify which parts of a given filename are sequence numbers.  In all cases both MS-DOS batch and NT batch (the languages encountered at the command prompt in older and newer versions of Windows respectively) use character collation even if Explorer uses string comparison.

What this means is that character collation and string comparison will produce identical sorts only if all sequence numbers are zero-padded to an uniform width.  I usually do this by using a consistent syntax for all date expressions and zero-padding all other numbers to four-character width.  It is possible to accomplish zero padding using only the functionality available through NT batch (by parsing filenames one character at a time to identify sequence numbers), but personally I take the easy way out and use GnuWin sed.

Code Select Expand

SETLOCAL ENABLEDELAYEDEXPANSION

REM  This is a zerofill script which converts ALL digit strings to four-digit width

FOR /F "tokens=1 delims=" %%A IN ('dir /o /b /a:-d') DO (
echo "%%A" > !temp!\input.txt
sed "s/\([^^0-9]\)\([0-9]\)\([^^0-9]\)/\1000\2\3/g" !temp!\input.txt > !temp!\outtemp1.txt
sed "s/\([^^0-9]\)\([0-9][0-9]\)\([^^0-9]\)/\100\2\3/g" !temp!\outtemp1.txt > !temp!\outtemp2.txt
sed "s/\([^^0-9]\)\([0-9][0-9][0-9]\)\([^^0-9]\)/\10\2\3/g" !temp!\outtemp2.txt > !temp!\output.txt
FOR /F "tokens=1 delims=" %%B IN (!temp!\output.txt) DO set OutFN=%%B
move "%%A" !OutFN!
del !temp!\outtemp*.txt
del !temp!\input.txt
del !temp!\output.txt
)

ENDLOCAL

[Duke87]

The thing that maddens me about that is that there are some cases where you want a numerical sort and others where you'd rather have an alphabetical sort.

Any date or number at the end of a filename, naturally, you want sorted in chronological order. And any numbering at the beginning of a file you probably do as well, if the numbers are meaningful.

My problem is, I save a lot of crap off the internet and I usually don't bother to change the filename. This leaves me with folders full of files with names like "32b6a4e9.jpg". Ideally, when it's gibberish like this, I'd want a straight by character sort. But Windows will put that file before "226f1c7d.jpg" rather than after because it sorts the former as "32" and the latter as "226". To a human eye this is ridiculous - but the computer of course cannot determine when the number is meaningful and when it isn't.

Personally, if I have a meaningful number in a filename, I always ensure that it's zero-padded so each file has the same number of digits. I do this out of obsessive-compulsiveness rather than functional purpose, but if I could get windows to just do character collation like it used to it would be a net improvement to my computing experience.

[Desert Man]

European road systems are under different countries' authorities, they each have their ideas of numbering or lettering their highways. For them to use the North American or US highway standard might be easier to get around, like for finding the "north-south" route to the closest "east-west" route by a geographical number system can be a time-saver. In Europe, drivers are used to the way their highways are numbered or lettered, therefore I don't see them having an issue getting around on their highway networks.

[kphoger]

For them to use the North American or US highway standard might be easier to get around, like for finding the "north-south" route to the closest "east-west" route by a geographical number system can be a time-saver.

Err... huh?  You look at a map, find your two cities, find the set of highways that most closely resembles a straight line, and read the route numbers.  I don't see how that process would change much by using a "˜north American or US highway standard'.

[albertocsc]

There are some European countries that use shields, although not as much as in the US. Here you have examples of Romania and Moldova.

Shields in Moldova:


Shields in Romania:

[agentsteel53]

astonishing how much - apart from the one point missing of course - the Romanian shield resembles USA '70 spec.