Some Problems with Special Charakters

[thuter]

Hi There,

We noticed some Problems with the german special characters  like ä, ö, ü, ß
If we use them in iptc info and read them in other programms (photoshop, fotostation, exifreader - like the most newspapers do) the will shown as:

ä = Š
ö = š
ü = Ÿ
ß = §

Its a local problem for us in germany, but i think other europeans like french, sweden or danmark will have the same problems with their special charakters like the french " ê " for example - that becomes a " ?"

[Kirk Baker]

We noticed some Problems with the german special characters  like ä, ö, ü, ß
If we use them in iptc info and read them in other programms (photoshop, fotostation, exifreader - like the most newspapers do) the will shown as:

ä = Š
ö = š
ü = Ÿ
ß = §

Its a local problem for us in germany, but i think other europeans like french, sweden or danmark will have the same problems with their special charakters like the french " ê " for example - that becomes a " ?"

IPTC specifies character encoding to be defined by the ISO 2022 standard which is IMHO the most convoluted standard for specifying character sets/encodings.  It is so complex that basically nobody implements it.  The one encoding I have seen used is the Unicode (UTF-8) which in ISO 2022 format is specified by ESC%G.  I have seen no other encodings.

Programs like Photoshop just put the text you enter into their Image Info dialog into IPTC as is.  So on my Windows PC the Euro symbol is 0x80, but on my Macintosh it is 0xDB.  Now when I open an image captioned on the opposite platform with the Euro in it, why does it look correct?  Because Photoshop now writes out an XMP record as well as an IPTC record and XMP specifies its text in UTF-8.  Photoshop prefers the XMP record if it exists over that of the IPTC record.  The UTF-8 character stream for the Euro character is 0xE2 0x82 on all platforms, regardless of locale.

Photo Mechanic writes out IPTC characters in MacRoman which is what most of our customers use.  We do that on both Mac OS X and Windows because for people who choose to use IPTC only this provides the capability for IPTC to be shared among Windows and Mac users.  We used to do what Photoshop did, but that made users of Photo Mechanic discouraged to find that when they captioned on their PC and went to a Mac and viewed it then the characters were wrong.  We decided that since there was no de-facto standard for character encoding we went with what was the most popular encoding at that time: MacRoman.

Photo Mechanic writes out properly encoded UTF-8 in its XMP records.  So if you want maximum compatibility, use XMP.  It is the direction that the industry is headed.  IPTC is obsolete, and even the IPTC folks know this and that is why they have created IPTC4XMP.

So these other apps that you describe will hopefully be upgraded to use XMP, because not only can XMP support western/roman languages, it can support *all* languages.  If you can configure these other apps like Fotostation to use XMP, all will be well.  If they do not yet support XMP then hopefully they will in the future.

-Kirk

[thuter]

It's not my problem - it's the problem of our customers. It's not easy to convince a international company changing their software cause we use PM ...

So for now it seems to be the only solution to change the programm... 

[Kirk Baker]

It's not my problem - it's the problem of our customers. It's not easy to convince a international company changing their software cause we use PM ...

So for now it seems to be the only solution to change the programm... 

Why can't they use XMP?  It is the only true international standard.  If we went and just put in the characters from the encoding on your system then it would be no more interoperable outside your country than anything else.

Are companies just afraid of XMP?  I don't see why.  Adobe gives out a free SDK with source code to anyone who wants it.  And it is quite easy to use.  XMP is the future.  IPTC is the present but will fade away over time.

Future versions of PM will let users type in Japanese, Korean, Chinese and any roman-based languages, all through XMP.

-Kirk

[soren steffen]

I totally agree with Kirk. Why can't Fotostation use XMP ?

And why don't developers agree on standards for color labelling and rating so I can have my colorlabels and ratings follow the file or synchronize them?

 

[thuter]

Why can't all streets have 5 lanes so i can drive 250km/h without traffic jam ?
Sometimes you have to arrange with the circumstances - if you want to sell pictures 

[Kirk Baker]

Why can't all streets have 5 lanes so i can drive 250km/h without traffic jam ?
Sometimes you have to arrange with the circumstances - if you want to sell pictures 

Since it is unlikely that all of the other apps are going to deal with MacRoman or are going to adopt XMP anytime soon, then probably what we should add is a way for users to choose a default encoding for IPTC.  Those that want to use ISO-8859-1 or ANSI or MacRoman or MacCyrillic or Windows-1252 can do so.  But do you see what a mess this could be?

That's why I like the UTF-8 variant of Unicode.  It is very simple to work with and convert to/from different encodings.  If only it had been invented before ASCII then the world would be a better place. 

-Kirk

[thuter]

Why can't all streets have 5 lanes so i can drive 250km/h without traffic jam ?
Sometimes you have to arrange with the circumstances - if you want to sell pictures 

Since it is unlikely that all of the other apps are going to deal with MacRoman or are going to adopt XMP anytime soon, then probably what we should add is a way for users to choose a default encoding for IPTC.  Those that want to use ISO-8859-1 or ANSI or MacRoman or MacCyrillic or Windows-1252 can do so.  But do you see what a mess this could be?

That's why I like the UTF-8 variant of Unicode.  It is very simple to work with and convert to/from different encodings.  If only it had been invented before ASCII then the world would be a better place. 

-Kirk

I see the mess clearly. But i stuck between a rock and a hard place - I understand all your arguments and PM is a nearly perfect program for our shooters and we want to use ist right now to deliver picures to our customers....but we can't cause this damned special charakter problem (depending on the customers software).

The Soccer WC 2006 is near...and thats a big thing in germany - maybe you get a little patch for germans untill the end of may in your magic fingers...

[Kirk Baker]

Why can't all streets have 5 lanes so i can drive 250km/h without traffic jam ?
Sometimes you have to arrange with the circumstances - if you want to sell pictures 

Since it is unlikely that all of the other apps are going to deal with MacRoman or are going to adopt XMP anytime soon, then probably what we should add is a way for users to choose a default encoding for IPTC.  Those that want to use ISO-8859-1 or ANSI or MacRoman or MacCyrillic or Windows-1252 can do so.  But do you see what a mess this could be?

That's why I like the UTF-8 variant of Unicode.  It is very simple to work with and convert to/from different encodings.  If only it had been invented before ASCII then the world would be a better place. 

-Kirk

I see the mess clearly. But i stuck between a rock and a hard place - I understand all your arguments and PM is a nearly perfect program for our shooters and we want to use ist right now to deliver picures to our customers....but we can't cause this damned special charakter problem (depending on the customers software).

The Soccer WC 2006 is near...and thats a big thing in germany - maybe you get a little patch for germans untill the end of may in your magic fingers...

How near?  June?

Which character set/encoding do you need?

-Kirk

[thuter]

How near?  June?

Which character set/encoding do you need?

-Kirk

WC starts on June 9th - one Week testing before would be enough 
For me it's enough to solve the problem with the german charakters "ä,ö,ü,ß". Our Caption text will only be in german and english.

[soren steffen]

Why can't all streets have 5 lanes so i can drive 250km/h without traffic jam ?
Sometimes you have to arrange with the circumstances - if you want to sell pictures 

Oh, I am used to arrange withe circumstances.
As you can see I even spell my own name without the danish  " ø "
And I never use the danish characters  æ -ø -å in my captions. Maybe your german clients could read your captions even if you spell Flügelflitzer as Fluegelflitzer
 

[thuter]

Why can't all streets have 5 lanes so i can drive 250km/h without traffic jam ?
Sometimes you have to arrange with the circumstances - if you want to sell pictures 

Oh, I am used to arrange withe circumstances.
As you can see I even spell my own name without the danish  " ø "
And I never use the danish characters  æ -ø -å in my captions. Maybe your german clients could read your captions even if you spell Flügelflitzer as Fluegelflitzer
 

Of course they could read it with "ue" instead of "ü" - but they wont find it in their database, cause they don't search for "Bayern Muenchen" or "Bayern Munic" but rather "Bayern München" . My problem are only Names of persons, citys, clubs etc. not "Skål!" or "Skal!"

Farvel sålænge  ( in german tschüß ! )

or better

Farvel salaenge or tschuess

[atf300]

it would be nice ,if you could chose a character set in preferences . my main language slovenian has special characters like š č ž , so now im forced to go trough all my iptc in photoshop after ive written it in PM .  i belive a selectable character set might not be the way in standards but it certanly would help people in europe.

[miguelr]

I am captioning pictures in spanish language.

Using the IPTC pad, the reproduction of the spanish caracters are correct.

On preview mode I inserted the {caption} variable to the info text and the caracters are not reproduce correctly.

Miguel
PM 4.4.3 on Windows XP

[Kirk Baker]

I am captioning pictures in spanish language.

Using the IPTC pad, the reproduction of the spanish caracters are correct.

On preview mode I inserted the {caption} variable to the info text and the caracters are not reproduce correctly.

Miguel
PM 4.4.3 on Windows XP

I've just fixed that.  It will be corrected in 4.4.3.1.

-Kirk