preview RAW file from digital camera (CANON, NIKON,etc..)

Post by **admin** » 15 Apr 2008 17:16

infimum wrote:OK

If v7.00.0026 did not work for you, then I really don't know what's going on. How do other apps save their "ö"s???

infimum · Post by **infimum** » 15 Apr 2008 18:25

How do other apps save their "ö"s???

OK, here's one way. Why don't you take a look at 7-zip portable.
http://portableapps.com/apps/utilities/7-zip_portable
And extract it and take a look at 7-ZipPortable\Data\settings\7-zip_portable.reg. Do you see that FolderShortcuts and FolderHistory are specified as hex values? Can you do the same thing?

Post by **admin** » 15 Apr 2008 20:20

infimum wrote:
How do other apps save their "ö"s???
OK, here's one way. Why don't you take a look at 7-zip portable.
http://portableapps.com/apps/utilities/7-zip_portable
And extract it and take a look at 7-ZipPortable\Data\settings\7-zip_portable.reg. Do you see that FolderShortcuts and FolderHistory are specified as hex values? Can you do the same thing?

Okay, that's the registry way. It's one possible way, but I don't like it's unfriendlyness: the data are not human-readable.

I could simply save all data converted to UTF-8 and solve all problems. But I want to know the conditions... I have other users from Japan and Korea (all DBCS) and they have none of the problems you describe. I need to know exactly what's going on.

If you type or paste an "ö" into Notepad and save it to disk. Could you do this and send me the TXT file, wrapped in a zip. I'd like to see what's happening to the "ö".

infimum · Post by **infimum** » 15 Apr 2008 20:57

admin wrote:I have other users from Japan and Korea (all DBCS) and they have none of the problems you describe. I need to know exactly what's going on.

It's just that they don't bother to tell you (or they don't need "ö" or they have German Windows). It's not specific to my condition. When I look at the INI file, it's encoded in Windows 932 aka Shift-JIS, the default encoding of Japanese Windows. There's just no way to save "ö" in this encoding. Believe me.

If you type or paste an "ö" into Notepad and save it to disk. Could you do this and send me the TXT file, wrapped in a zip. I'd like to see what's happening to the "ö".

Notepad is quite good at handling various letters from various languages. Of course it shows "ö". Other text editors are not so good at handling them. They need to be explicitly told that the encoding is UFT7, 8, etc. Otherwise "ö" would be garbled. Just copy and paste some random Chinese characters on a txt file encoded in Windows 1252, which is your default, they will be garbled (even if you have a fond that covers Chinese characters). It's the same deal.

Look, if you don't like hex values, then the only way is to encode the INI in UTF7, 8, etc. There is no other way.

Post by **admin** » 15 Apr 2008 21:26

infimum wrote:Look, if you don't like hex values, then the only way is to encode the INI in UTF7, 8, etc. There is no other way.

Okay, I'm ready for UTF8. But in 7.00.0026 I DID encode strings containing "ö" in UTF8 ("ö" should be "Ã¶"), and you told me it did not work...

Please try again with the current beta version. There's an undocumented INI Key:
EscapeUpperANSI=1
Set it to 1 and restart XY. UTF8 encoding should set in for paths containing "ö".

infimum · Post by **infimum** » 15 Apr 2008 21:52

admin wrote:But in 7.00.0026 I DID encode strings containing "ö" in UTF8 ("ö" should be "Ã¶"), and you told me it did not work...

Please try again with the current beta version. There's an undocumented INI Key:
EscapeUpperANSI=1
Set it to 1 and restart XY. UTF8 encoding should set in for paths containing "ö".

I don't know the exact mechanics of this. But the entire INI file is still encoded in default encoding of Windows (diffrent from language to language). I am no programmer, but I can tell that with this encoding, nothing works. Well, except hex values because numbers aren't encoded differently from language to language and each character from various languages is given a unique number. This is why I am telling the entire INI should be in UTF (if you want to avoid using hex) and XY should be equipped with the ability of reading off unicode settings.

Post by **admin** » 15 Apr 2008 22:00

infimum wrote:
admin wrote:But in 7.00.0026 I DID encode strings containing "ö" in UTF8 ("ö" should be "Ã¶"), and you told me it did not work...

Please try again with the current beta version. There's an undocumented INI Key:
EscapeUpperANSI=1
Set it to 1 and restart XY. UTF8 encoding should set in for paths containing "ö".
I don't know the exact mechanics of this. But the entire INI file is still encoded in default encoding of Windows (diffrent from language to language). I am no programmer, but I can tell that with this encoding, nothing works. Well, except hex values because numbers aren't encoded differently from language to language and each character from various languages is given a unique number. This is why I am telling the entire INI should be in UTF (if you want to avoid using hex) and XY should be equipped with the ability of reading off unicode settings.

EscapeUpperANSI only sets in on DBCS systems, but you are on DBCS, so it should work.

Okay, looks like my UnicodeToUTF8 function is not okay! Let's test it: create an empty file called "köln.txt", select it and do menu File | Rename Special | Unicode to UTF-8. It should become kÃ¶ln.txt. But I bet it does not, right?

infimum · Post by **infimum** » 15 Apr 2008 22:31

admin wrote:Okay, looks like my UnicodeToUTF8 function is not okay! Let's test it: create an empty file called "köln.txt", select it and do menu File | Rename Special | Unicode to UTF-8. It should become kÃ¶ln.txt. But I bet it does not, right?

Your prediction is correct. I am not sure what "kÃ¶ln" is supposed to be encoded in, but "Ã" and "¶" are garbled. They shouldn't be garbled if they are really encoded in UTF8 no matter what the default encoding is.

Post by **admin** » 15 Apr 2008 22:33

infimum wrote:
admin wrote:Okay, looks like my UnicodeToUTF8 function is not okay! Let's test it: create an empty file called "köln.txt", select it and do menu File | Rename Special | Unicode to UTF-8. It should become kÃ¶ln.txt. But I bet it does not, right?
Your prediction is correct. I am not sure what "kÃ¶ln" is supposed to be encoded in, but "Ã" and "¶" are garbled. They shouldn't be garbled if they are really encoded in UTF8 no matter what the default encoding is.

I'll upload a version with an improved UTF8 conversion in a minute. Please try the same "Köln" rename again with it...

infimum · Post by **infimum** » 15 Apr 2008 23:06

No improvement.

I am afraid this kind of detour tactics will not work.

Post by **admin** » 15 Apr 2008 23:25

infimum wrote:No improvement.

I am afraid this kind of detour tactics will not work.

detour tactics ? I'm right in the center of it. Later...

Post by **admin** » 16 Apr 2008 07:25

infimum wrote:No improvement.

I prepared a little zip containing 2 files named köln.txt and kÃ¶ln.txt. The also contain the words köln resp. kÃ¶ln.
http://www.xyplorer.com/download/cologne.zip

Here are 2 screenshots of the files and their contents. I wonder how this comes out on your system, infimum.

infimum · Post by **infimum** » 16 Apr 2008 13:11

Zip doesn't support unicode! So the file names are garbled as in my previous picture. Use 7z if you want to compress unicode-named (at least in my system) files/folders. Hex values appear as seen in you picture. The text itself is encoded in Windows 1252, which is, as I repeatedly said, not the default of my system. Does UnicodeToUTF8 function ever involve the encoding of Windows 1252? If so, forget it. It's not going to work.

Post by **admin** » 16 Apr 2008 13:22

infimum wrote:Zip doesn't support unicode! So the file names are garbled as in my previous picture. Use 7z if you want to compress unicode-named (at least in my system) files/folders. Hex values appear as seen in you picture. The text itself is encoded in Windows 1252, which is, as I repeatedly said, not the default of my system. Does UnicodeToUTF8 function ever involve the encoding of Windows 1252? If so, forget it. It's not going to work.

Sorry, here's a 7z version:
http://www.xyplorer.com/download/cologne.7z

No, the UnicodeToUTF8 converts Unicode to UTF8 -- independent of any codepage like Windows 1252. If I understand it right, codepages control the relation between bytes and screen pixels. However, the conversion we need here is just between bytes.

infimum · Post by **infimum** » 16 Apr 2008 13:45

OK, everything appears fine as seen in your picture. FYI, WinZip recently decided to support unicode. Other unicode-consious de/compression programs will follow WinZip's decision. In fact, the recent alpha version of 7-zip support unicode in the Zip format.

admin wrote:No, the UnicodeToUTF8 converts Unicode to UTF8 -- independent of any codepage like Windows 1252. If I understand it right, codepages control the relation between bytes and screen pixels. However, the conversion we need here is just between bytes.

OK, I jumped the gun. Sorry. But, in that case, why is conversion like "ö" to "Ã¶" involved at all?