Discussion:
Problem with "/" character and Japanese
(too old to reply)
Jeffrey Ellis
2003-12-24 21:19:22 UTC
Permalink
Hi, All--

I seem to having a pretty significant problem with the way the Japanese
localized version of our application is behaving as regards the "/"
character.

It turns out that in a Japanese system, JIS or Shift-JIS, when this
character is input or received from a Western encoded source, the system
reinterprets it as a "¥" instead.

This is obviously a pretty important character, and I haven't found any way
to correct this in our code.

I have looked up the "/" character and it's ASCII 47, hex 2f, and 002f in
Unicode, UTF8, JIS and Shift-JIS.

Does anyone know why this is being interpreted as a "¥", and what we can do
to make sure when a "/" is input in Western encoding that it will always
come out a "/" in Japanese?

All My Best,
Jeffrey
Louise Bremner
2003-12-24 22:45:01 UTC
Permalink
Post by Jeffrey Ellis
I seem to having a pretty significant problem with the way the Japanese
localized version of our application is behaving as regards the "/"
character.
It turns out that in a Japanese system, JIS or Shift-JIS, when this
character is input or received from a Western encoded source, the system
reinterprets it as a "¥" instead.
This is obviously a pretty important character, and I haven't found any way
to correct this in our code.
I have looked up the "/" character and it's ASCII 47, hex 2f, and 002f in
Unicode, UTF8, JIS and Shift-JIS.
Does anyone know why this is being interpreted as a "¥", and what we can do
to make sure when a "/" is input in Western encoding that it will always
come out a "/" in Japanese?
You do indeed seem to have a problem there.... I'm aware that the "\"
(backslash) character appears as "¥" in Japanese systems (because
Japanese fonts need to include the yen sign, and the backslash character
was deemed to be sufficiently lacking in importance so was replaced),
but I've never seen the "/" one doing so too.

________________________________________________________________________
Louise Bremner (log at gol dot com)
If you want a reply by e-mail, don't write to my Yahoo address!
Jeffrey Ellis
2003-12-24 23:19:04 UTC
Permalink
Post by Louise Bremner
Post by Jeffrey Ellis
I seem to having a pretty significant problem with the way the Japanese
localized version of our application is behaving as regards the "/"
character.
It turns out that in a Japanese system, JIS or Shift-JIS, when this
character is input or received from a Western encoded source, the system
reinterprets it as a "¥" instead.
This is obviously a pretty important character, and I haven't found any way
to correct this in our code.
I have looked up the "/" character and it's ASCII 47, hex 2f, and 002f in
Unicode, UTF8, JIS and Shift-JIS.
Does anyone know why this is being interpreted as a "¥", and what we can do
to make sure when a "/" is input in Western encoding that it will always
come out a "/" in Japanese?
You do indeed seem to have a problem there.... I'm aware that the "\"
(backslash) character appears as "¥" in Japanese systems (because
Japanese fonts need to include the yen sign, and the backslash character
was deemed to be sufficiently lacking in importance so was replaced),
but I've never seen the "/" one doing so too.
________________________________________________________________________
Louise Bremner (log at gol dot com)
If you want a reply by e-mail, don't write to my Yahoo address!
Hi, Louise--

Obviously, you're correct. That was an error on my part, sorry. It is the
"\" that I meant. Do you know what we need to do to correct input and output
so that everyone is getting/seeing the right character?

All My Best,
Jeffrey
Louise Bremner
2003-12-24 23:51:27 UTC
Permalink
Post by Jeffrey Ellis
Obviously, you're correct. That was an error on my part, sorry. It is the
"\" that I meant. Do you know what we need to do to correct input and output
so that everyone is getting/seeing the right character?
You can't, if the user is using a Japanese font--there is no backslash
character in the one-byte sections of those fonts. You could fake it
with the two-byte equivalent, or by forcing the text to appear in an
English font, I suppose.

But why do you want the backslash character to appear as a backslash,
might I ask? In my (admittedly limited) experience, I've only
encountered its use in pathnames, but I assure you that the few Japanese
users nowadays who view pathnames do expect to see a yen-sign there, not
a backslash.

________________________________________________________________________
Louise Bremner (log at gol dot com)
If you want a reply by e-mail, don't write to my Yahoo address!
Jeffrey Ellis
2003-12-25 02:36:32 UTC
Permalink
Post by Louise Bremner
Post by Jeffrey Ellis
Obviously, you're correct. That was an error on my part, sorry. It is the
"\" that I meant. Do you know what we need to do to correct input and output
so that everyone is getting/seeing the right character?
You can't, if the user is using a Japanese font--there is no backslash
character in the one-byte sections of those fonts. You could fake it
with the two-byte equivalent, or by forcing the text to appear in an
English font, I suppose.
But why do you want the backslash character to appear as a backslash,
might I ask? In my (admittedly limited) experience, I've only
encountered its use in pathnames, but I assure you that the few Japanese
users nowadays who view pathnames do expect to see a yen-sign there, not
a backslash.
________________________________________________________________________
Louise Bremner (log at gol dot com)
If you want a reply by e-mail, don't write to my Yahoo address!
Hi, Louise--

Good question...and thanks for letting me know our users don't mind seeing
the ¥ in the path.

Unfortunately, our program uses the paths of files to send to servers, so we
need to have that character be correct. What we are seeing is that the
communications with the servers breaks down in the Japanese version, and it
has taken us this long to figure out what the problem was. Now we need to
figure out what to do about it.

All My Best,
Jeffrey
Louise Bremner
2003-12-25 03:46:23 UTC
Permalink
Post by Jeffrey Ellis
Post by Louise Bremner
But why do you want the backslash character to appear as a backslash,
might I ask? In my (admittedly limited) experience, I've only
encountered its use in pathnames, but I assure you that the few Japanese
users nowadays who view pathnames do expect to see a yen-sign there, not
a backslash.
Good question...and thanks for letting me know our users don't mind seeing
the ¥ in the path.
Unfortunately, our program uses the paths of files to send to servers, so we
need to have that character be correct. What we are seeing is that the
communications with the servers breaks down in the Japanese version, and it
has taken us this long to figure out what the problem was. Now we need to
figure out what to do about it.
Now I'm totally confused. How on earth does your program manage to send
an actual yen character instead of a code that is interpreted as a
pathname delimeter by the recipient system?

________________________________________________________________________
Louise Bremner (log at gol dot com)
If you want a reply by e-mail, don't write to my Yahoo address!
Jeffrey Ellis
2003-12-25 06:00:26 UTC
Permalink
Post by Louise Bremner
Post by Jeffrey Ellis
Post by Louise Bremner
But why do you want the backslash character to appear as a backslash,
might I ask? In my (admittedly limited) experience, I've only
encountered its use in pathnames, but I assure you that the few Japanese
users nowadays who view pathnames do expect to see a yen-sign there, not
a backslash.
Good question...and thanks for letting me know our users don't mind seeing
the ¥ in the path.
Unfortunately, our program uses the paths of files to send to servers, so we
need to have that character be correct. What we are seeing is that the
communications with the servers breaks down in the Japanese version, and it
has taken us this long to figure out what the problem was. Now we need to
figure out what to do about it.
Now I'm totally confused. How on earth does your program manage to send
an actual yen character instead of a code that is interpreted as a
pathname delimeter by the recipient system?
________________________________________________________________________
Louise Bremner (log at gol dot com)
If you want a reply by e-mail, don't write to my Yahoo address!
LOL. That's what we're trying to figure out! All we can see is that the
Japanese version is failing in many ways, whereas the other languages are
operating well. This is the only difference we can see.

Are there any others?

All My Best,
Jeffrey
Paul Blay
2003-12-25 08:27:06 UTC
Permalink
"Jeffrey Ellis" wrote ...
Post by Jeffrey Ellis
Post by Louise Bremner
Post by Jeffrey Ellis
Good question...and thanks for letting me know our users don't mind seeing
the ¥ in the path.
Unfortunately, our program uses the paths of files to send to servers, so we
need to have that character be correct. What we are seeing is that the
communications with the servers breaks down in the Japanese version, and
****
Post by Jeffrey Ellis
Post by Louise Bremner
Post by Jeffrey Ellis
it
has taken us this long to figure out what the problem was.
****
Post by Jeffrey Ellis
Post by Louise Bremner
Post by Jeffrey Ellis
Now we need to
figure out what to do about it.
Now I'm totally confused. How on earth does your program manage to send
an actual yen character instead of a code that is interpreted as a
pathname delimeter by the recipient system?
LOL. That's what we're trying to figure out! All we can see is that the
Japanese version is failing in many ways, whereas the other languages are
operating well. This is the only difference we can see.
Sounds like you need to take a little bit longer and be _sure_ you've found
out what the problem is.
Post by Jeffrey Ellis
Are there any others?
It's your program, how would we know what differences there are between
Japanese and non-Japanese versions?
Jeffrey Ellis
2003-12-25 19:30:58 UTC
Permalink
Post by Paul Blay
"Jeffrey Ellis" wrote ...
Post by Jeffrey Ellis
Post by Louise Bremner
Post by Jeffrey Ellis
Good question...and thanks for letting me know our users don't mind seeing
the ¥ in the path.
Unfortunately, our program uses the paths of files to send to servers, so we
need to have that character be correct. What we are seeing is that the
communications with the servers breaks down in the Japanese version, and
****
Post by Jeffrey Ellis
Post by Louise Bremner
Post by Jeffrey Ellis
it
has taken us this long to figure out what the problem was.
****
Post by Jeffrey Ellis
Post by Louise Bremner
Post by Jeffrey Ellis
Now we need to
figure out what to do about it.
Now I'm totally confused. How on earth does your program manage to send
an actual yen character instead of a code that is interpreted as a
pathname delimeter by the recipient system?
LOL. That's what we're trying to figure out! All we can see is that the
Japanese version is failing in many ways, whereas the other languages are
operating well. This is the only difference we can see.
Sounds like you need to take a little bit longer and be _sure_ you've found
out what the problem is.
Post by Jeffrey Ellis
Are there any others?
It's your program, how would we know what differences there are between
Japanese and non-Japanese versions?
Ahem...I actually meant are there any other replaced characters in Japanese
we should know about.

All My Best,
Jeffrey
Louise Bremner
2003-12-25 22:50:12 UTC
Permalink
Post by Jeffrey Ellis
Post by Louise Bremner
Post by Jeffrey Ellis
Unfortunately, our program uses the paths of files to send to servers, so
we need to have that character be correct. What we are seeing is that the
communications with the servers breaks down in the Japanese version, and it
has taken us this long to figure out what the problem was. Now we need to
figure out what to do about it.
Now I'm totally confused. How on earth does your program manage to send
an actual yen character instead of a code that is interpreted as a
pathname delimeter by the recipient system?
LOL. That's what we're trying to figure out! All we can see is that the
Japanese version is failing in many ways, whereas the other languages are
operating well. This is the only difference we can see.
I'm not sure I've made this clear: just because different operating
systems display that ASCII code differently to human eyes, doesn't mean
that the machines themselves are interpreting it differently. To a
computer, the string of electrical pulses is exactly the same, so it
ought to be interpreting it as a pathname, regardless of how it appears
to the user.
Post by Jeffrey Ellis
Are there any others?
Other purely cosmetic differences? The "|" character has a break in it,
there are no "high-ASCII" characters such as curly quotes, Japanese
fonts don't have the "hinting" for English text that makes it easier to
read, and the positions of some of the keys on the keyboard are wildly
different. But none of that should affect the way a computer interprets
the character codes internally.

If it does, you have made a huge breakthrough in computing.

________________________________________________________________________
Louise Bremner (log at gol dot com)
If you want a reply by e-mail, don't write to my Yahoo address!
Jeffrey Ellis
2003-12-25 23:03:55 UTC
Permalink
Post by Louise Bremner
Post by Jeffrey Ellis
Post by Louise Bremner
Post by Jeffrey Ellis
Unfortunately, our program uses the paths of files to send to servers, so
we need to have that character be correct. What we are seeing is that the
communications with the servers breaks down in the Japanese version, and it
has taken us this long to figure out what the problem was. Now we need to
figure out what to do about it.
Now I'm totally confused. How on earth does your program manage to send
an actual yen character instead of a code that is interpreted as a
pathname delimeter by the recipient system?
LOL. That's what we're trying to figure out! All we can see is that the
Japanese version is failing in many ways, whereas the other languages are
operating well. This is the only difference we can see.
I'm not sure I've made this clear: just because different operating
systems display that ASCII code differently to human eyes, doesn't mean
that the machines themselves are interpreting it differently. To a
computer, the string of electrical pulses is exactly the same, so it
ought to be interpreting it as a pathname, regardless of how it appears
to the user.
Post by Jeffrey Ellis
Are there any others?
Other purely cosmetic differences? The "|" character has a break in it,
there are no "high-ASCII" characters such as curly quotes, Japanese
fonts don't have the "hinting" for English text that makes it easier to
read, and the positions of some of the keys on the keyboard are wildly
different. But none of that should affect the way a computer interprets
the character codes internally.
If it does, you have made a huge breakthrough in computing.
________________________________________________________________________
Louise Bremner (log at gol dot com)
If you want a reply by e-mail, don't write to my Yahoo address!
Hi, Louise--

Sorry. Of course that's reasonable. I just can't see any other differences
between the Japanese version and the eleven other languages we have our app
at the moment -- even Korean is behaving exactly as it should.

So I guess I'm just searching for a clue of some sort.

Thanks again.

All My Best,
Jeffrey
Brandon Berg
2003-12-29 18:27:12 UTC
Permalink
Post by Jeffrey Ellis
Sorry. Of course that's reasonable. I just can't see any other differences
between the Japanese version and the eleven other languages we have our app
at the moment -- even Korean is behaving exactly as it should.
Don't Korean fonts have the same issue? IIRC, the backslash is replaced with
a character that looks like a crossed-out W.
Louise Bremner
2003-12-30 02:34:51 UTC
Permalink
Post by Brandon Berg
Post by Jeffrey Ellis
Sorry. Of course that's reasonable. I just can't see any other differences
between the Japanese version and the eleven other languages we have our
app
Post by Jeffrey Ellis
at the moment -- even Korean is behaving exactly as it should.
Don't Korean fonts have the same issue? IIRC, the backslash is replaced with
a character that looks like a crossed-out W.
Maybe Korean computers are better designed, so they translate that
character back into a backslash when communicating with non-Korean
computers?

________________________________________________________________________
Louise Bremner (log at gol dot com)
If you want a reply by e-mail, don't write to my Yahoo address!
Michael Cash
2003-12-30 10:42:45 UTC
Permalink
Post by Brandon Berg
Post by Jeffrey Ellis
Sorry. Of course that's reasonable. I just can't see any other differences
between the Japanese version and the eleven other languages we have our
app
Post by Jeffrey Ellis
at the moment -- even Korean is behaving exactly as it should.
Don't Korean fonts have the same issue? IIRC, the backslash is replaced with
a character that looks like a crossed-out W.
Is there some reason Brandon's post would have *my* name attached to
it?





--

Michael Cash

"While we thank you for considering our firm, regretfully we have no openings
for a person of your educational background and are returning your resume.
Despite what your academic advisor may have told you, there are, to the best
of our knowledge, no openings in our industry for a person with a degree in
farm ecology."

Dr. Seymore Butts
Human Resources
Acme Pharmeceuticals, Inc.
Paul Blay
2003-12-30 11:10:57 UTC
Permalink
Post by Michael Cash
Post by Brandon Berg
Post by Jeffrey Ellis
Sorry. Of course that's reasonable. I just can't see any other differences
between the Japanese version and the eleven other languages we have our app
at the moment -- even Korean is behaving exactly as it should.
Don't Korean fonts have the same issue? IIRC, the backslash is replaced with
a character that looks like a crossed-out W.
Is there some reason Brandon's post would have *my* name attached to
it?
Your post (that I am replying to here) is shown in OE on my newsserver as a reply
to your 'kanji for unsui' post. :- news:***@4ax.com

The post of Brandon's that you think has your name attached to it is shown
in OE on my newsserver under his name in the 'Problem with "/" character and
Japanese" thread. :- news:%l_Hb.509$***@attbi_s53

Moreover the start of the post of your's that I'm replying to is
'On Mon, 29 Dec 2003 18:27:12 GMT, "Brandon Berg" <***@cesmail.net>
brought down from the Mount tablets inscribed:'

So either
a) You are making a more than usually obscure 'joke'
or
b) Your newsserver / newsreader is playing a more than usually obscure joke
on you.
Michael Cash
2003-12-30 15:31:43 UTC
Permalink
On Tue, 30 Dec 2003 11:10:57 -0000, "Paul Blay"
Post by Paul Blay
Post by Michael Cash
Post by Brandon Berg
Post by Jeffrey Ellis
Sorry. Of course that's reasonable. I just can't see any other differences
between the Japanese version and the eleven other languages we have our app
at the moment -- even Korean is behaving exactly as it should.
Don't Korean fonts have the same issue? IIRC, the backslash is replaced with
a character that looks like a crossed-out W.
Is there some reason Brandon's post would have *my* name attached to
it?
Your post (that I am replying to here) is shown in OE on my newsserver as a reply
The post of Brandon's that you think has your name attached to it is shown
in OE on my newsserver under his name in the 'Problem with "/" character and
Moreover the start of the post of your's that I'm replying to is
brought down from the Mount tablets inscribed:'
So either
a) You are making a more than usually obscure 'joke'
or
b) Your newsserver / newsreader is playing a more than usually obscure joke
on you.
The correct answer is "b". His post shows up with my name out to the
right of the subject line. I had to look at the headers to see who
actually posted it.




--

Michael Cash

"While we thank you for considering our firm, regretfully we have no openings
for a person of your educational background and are returning your resume.
Despite what your academic advisor may have told you, there are, to the best
of our knowledge, no openings in our industry for a person with a degree in
farm ecology."

Dr. Seymore Butts
Human Resources
Acme Pharmeceuticals, Inc.
Paul Blay
2003-12-30 16:13:28 UTC
Permalink
Post by Michael Cash
On Tue, 30 Dec 2003 11:10:57 -0000, "Paul Blay"
Post by Paul Blay
b) Your newsserver / newsreader is playing a more than usually obscure joke
on you.
The correct answer is "b". His post shows up with my name out to the
right of the subject line. I had to look at the headers to see who
actually posted it.
Sounds like you should move to a quality newsreader like Outlook Express.


















... Can't believe I actually managed to type that with a straight face.
Louise Bremner
2003-12-25 23:00:50 UTC
Permalink
All we can see is that the Japanese version is failing in many ways,
whereas the other languages are operating well.
BTW, there is a mailing list that might reach someone who can help you
Nihongo Computing (a.k.a. "nihongo-computing") is a publicly accessible
mailing list (an email-based discussion group) for anyone who is
interested in designing, implementing, studying, evaluating,
criticizing, teaching, learning, simply using, and/or discussing
Japanese-language capability on computers, from personal to enterprise
systems for a variety of uses. Its purpose is to serve as a
platform-independent discussion list for people of any nationality who
need to use the Japanese language with computers. Topics can also
include (but are not necessarily limited to) computing in Japan and
L10N. The list is un-moderated, un-sponsored, automated, and supported
by people who think a list like this would be useful.
<http://www.msdi.co.jp/public/mailing-lists/listinfo/nihongo-computing>

Warning: It might be best to avoid mentioning that you think the problem
is related to the way Japanese operating systems display ASCII codes
differently.

________________________________________________________________________
Louise Bremner (log at gol dot com)
If you want a reply by e-mail, don't write to my Yahoo address!
Jeffrey Ellis
2003-12-26 00:56:35 UTC
Permalink
Post by Louise Bremner
All we can see is that the Japanese version is failing in many ways,
whereas the other languages are operating well.
BTW, there is a mailing list that might reach someone who can help you
Nihongo Computing (a.k.a. "nihongo-computing") is a publicly accessible
mailing list (an email-based discussion group) for anyone who is
interested in designing, implementing, studying, evaluating,
criticizing, teaching, learning, simply using, and/or discussing
Japanese-language capability on computers, from personal to enterprise
systems for a variety of uses. Its purpose is to serve as a
platform-independent discussion list for people of any nationality who
need to use the Japanese language with computers. Topics can also
include (but are not necessarily limited to) computing in Japan and
L10N. The list is un-moderated, un-sponsored, automated, and supported
by people who think a list like this would be useful.
<http://www.msdi.co.jp/public/mailing-lists/listinfo/nihongo-computing>
Warning: It might be best to avoid mentioning that you think the problem
is related to the way Japanese operating systems display ASCII codes
differently.
________________________________________________________________________
Louise Bremner (log at gol dot com)
If you want a reply by e-mail, don't write to my Yahoo address!
Hi, Louise--

Thank you. This sounds great :)

Will do :)

All My Best,
Jeffrey
Dale Walker
2003-12-29 18:16:25 UTC
Permalink
Post by Louise Bremner
Post by Jeffrey Ellis
Obviously, you're correct. That was an error on my part, sorry. It is the
"\" that I meant. Do you know what we need to do to correct input and output
so that everyone is getting/seeing the right character?
You can't, if the user is using a Japanese font--there is no backslash
character in the one-byte sections of those fonts. You could fake it
with the two-byte equivalent, or by forcing the text to appear in an
English font, I suppose.
But why do you want the backslash character to appear as a backslash,
might I ask?
In my (admittedly limited) experience, I've only
encountered its use in pathnames, but I assure you that the few Japanese
users nowadays who view pathnames do expect to see a yen-sign there, not
a backslash.
If you work with pathnames a lot you'll realise how obfuscating the
yen character really is. It considerably reduces my productivity when
working on pathnames.

I'm learning Japanese so I like to have the full functionality (some
apps I have only work with the system font set to Japanese) of
Japanese on my PC but my job involves lots of programming so I work
with pathnames all the time. Fortunately one of my clients has the MUI
disk which means I can now have different system fonts for different
login users. It still means logging in to either UK or JP accounts to
get the right characters so it's not the ideal solution but at least I
don't have to go through the rigmarole of changing system fonts each
time I want to switch languages.

I also have a couple of clients who would prefer to have a Japanese
system font but also work a lot with pathnames. One of my clients is a
married couple. The husband is a web designer and the wife is a
Japanese interpreter. They only have one PC and divorce has been
mentioned on at least one occasion.

-------------------------------------------------------------
Dale Walker London Techno Events Saiko!
***@sorted.org ***@sorted.org ***@sorted.org
London, UK london.sorted.org saiko.sorted.org
Brandon Berg
2003-12-29 18:26:00 UTC
Permalink
Post by Dale Walker
If you work with pathnames a lot you'll realise how obfuscating the
yen character really is. It considerably reduces my productivity when
working on pathnames.
You do realize that you can use the yen sign as a path delimiter, don't you?
Paul Blay
2003-12-29 19:36:59 UTC
Permalink
"Brandon Berg" wrote ...
Post by Brandon Berg
Post by Dale Walker
If you work with pathnames a lot you'll realise how obfuscating the
yen character really is. It considerably reduces my productivity when
working on pathnames.
You do realize that you can use the yen sign as a path delimiter, don't you?
You do realize* that using the yen sign as a path delimiter is exactly why
he realises* how obfuscating the yen character is when working with pathnames.
(Although it doesn't bother me)

* Darn it. Now they _both_ look wrong.
Drew Hamilton
2003-12-29 20:47:12 UTC
Permalink
Post by Dale Walker
If you work with pathnames a lot you'll realise how obfuscating the
yen character really is. It considerably reduces my productivity when
working on pathnames.
I "work with pathnames" as much as any developer, and I've just
stopped noticing the difference between backslash and yen sign --
to my brain, either is just as good as a path separator. Just
give it time and you'll get used to it.
--
- awh
http://www.awh.org/
Dale Walker
2003-12-29 21:28:05 UTC
Permalink
Post by Drew Hamilton
Post by Dale Walker
If you work with pathnames a lot you'll realise how obfuscating the
yen character really is. It considerably reduces my productivity when
working on pathnames.
I "work with pathnames" as much as any developer, and I've just
stopped noticing the difference between backslash and yen sign --
to my brain, either is just as good as a path separator. Just
give it time and you'll get used to it.
Been 'giving it time' for two years now. Still slows me down enough to
notice.

-------------------------------------------------------------
Dale Walker London Techno Events Saiko!
***@sorted.org ***@sorted.org ***@sorted.org
London, UK london.sorted.org saiko.sorted.org
Ken Yasumoto-Nicolson
2003-12-30 14:55:11 UTC
Permalink
Post by Drew Hamilton
Post by Dale Walker
If you work with pathnames a lot you'll realise how obfuscating the
yen character really is. It considerably reduces my productivity when
working on pathnames.
I "work with pathnames" as much as any developer, and I've just
stopped noticing the difference between backslash and yen sign --
to my brain, either is just as good as a path separator. Just
give it time and you'll get used to it.
I agree - I've never noticed a single problem regarding this issue - a
yen sign just *is* a backslash. Trying to do anything special is
probably the root of Dale's problems? Or has he got a funny keyboard
with both a yen key and a backslash key?

Ken
Dale Walker
2003-12-30 18:20:16 UTC
Permalink
On Tue, 30 Dec 2003 23:55:11 +0900, Ken Yasumoto-Nicolson
Post by Ken Yasumoto-Nicolson
Post by Drew Hamilton
Post by Dale Walker
If you work with pathnames a lot you'll realise how obfuscating the
yen character really is. It considerably reduces my productivity when
working on pathnames.
I "work with pathnames" as much as any developer, and I've just
stopped noticing the difference between backslash and yen sign --
to my brain, either is just as good as a path separator. Just
give it time and you'll get used to it.
I agree - I've never noticed a single problem regarding this issue - a
yen sign just *is* a backslash.
Yes, Paul has already explained I know that but it doesn't *look* like
one. It's my dislike of the appearance of the Yen sign when I'm
working with pathnames. Maybe I'm slightly dyslexic (never come up in
any tests mind) but I have noticeably more difficulty reading a
pathname dressed up in yens than backslashes. Simple as that.
Post by Ken Yasumoto-Nicolson
Trying to do anything special is
probably the root of Dale's problems? Or has he got a funny keyboard
with both a yen key and a backslash key?
Nope, I just don't like the look of the yen sign when I work with
pathnames and it scares other people that use my computer. Also, I
often have non Japanese clients over to check the progress of anything
I'm developing for them. I always have to spend ages explaining it's
not going to transfer over when it gets installed on their system.

-------------------------------------------------------------
Dale Walker London Techno Events Saiko!
***@sorted.org ***@sorted.org ***@sorted.org
London, UK london.sorted.org saiko.sorted.org
Louise Bremner
2003-12-30 23:45:10 UTC
Permalink
Post by Dale Walker
Post by Ken Yasumoto-Nicolson
Trying to do anything special is
probably the root of Dale's problems? Or has he got a funny keyboard
with both a yen key and a backslash key?
Nope, I just don't like the look of the yen sign when I work with
pathnames and it scares other people that use my computer. Also, I
often have non Japanese clients over to check the progress of anything
I'm developing for them. I always have to spend ages explaining it's
not going to transfer over when it gets installed on their system.
I used to rewrite the documentation for Japanese software that had been
localized for gaikoku (horrible job, not least because I never got to
see the programs, let alone play with them--after all, my job was just
to make sure that the "English" was correct). The documentation was
printed out from Japanese systems, so of course the pathnames had yen
signs in them. I kept commenting that they would have to make sure that
those yen signs were printed as backslashes in the English-language
versiqon, and they kept replying that that was how the pathnames appear
on screen, so the documentation is correct. If I persisted, they'd send
screenshots to prove it....

Eventually, I gave up, reckoning that the sight of yen-signs in the
documentation would warn the user that this developers lacked a couple
of hints towards a full clue, and there might be more problems lurking.

________________________________________________________________________
Louise Bremner (log at gol dot com)
If you want a reply by e-mail, don't write to my Yahoo address!
Drew Hamilton
2004-01-01 20:19:13 UTC
Permalink
Post by Dale Walker
I also have a couple of clients who would prefer to have a Japanese
system font but also work a lot with pathnames. One of my clients is a
married couple. The husband is a web designer and the wife is a
Japanese interpreter. They only have one PC and divorce has been
mentioned on at least one occasion.
I wonder if there are TrueType Font Editors. Then you could just
edit the system font and replace the \ with a backslash.

- awh
--
- awh
http://www.awh.org/
Dale Walker
2004-01-01 23:22:19 UTC
Permalink
Post by Drew Hamilton
Post by Dale Walker
I also have a couple of clients who would prefer to have a Japanese
system font but also work a lot with pathnames. One of my clients is a
married couple. The husband is a web designer and the wife is a
Japanese interpreter. They only have one PC and divorce has been
mentioned on at least one occasion.
I wonder if there are TrueType Font Editors. Then you could just
edit the system font and replace the \ with a backslash.
Actually, I've got a million year old copy of fontographer buried deep
inside my 'hide the mess' cupboard. It does TTF but not sure it'll
cope with unicode. I'll dig it out and have a play.


-------------------------------------------------------------
Dale Walker London Techno Events Saiko!
***@sorted.org ***@sorted.org ***@sorted.org
London, UK london.sorted.org saiko.sorted.org
Ross Klatte
2004-01-01 23:43:45 UTC
Permalink
Date: 2004-01-01 15:19 Eastern Standard Time
Post by Dale Walker
I also have a couple of clients who would prefer to have a Japanese
system font but also work a lot with pathnames. One of my clients is a
married couple. The husband is a web designer and the wife is a
Japanese interpreter. They only have one PC and divorce has been
mentioned on at least one occasion.
I wonder if there are TrueType Font Editors. Then you could just
edit the system font and replace the \ with a backslash.
I don't think the Font is a problem.
The problem is, in my opinion:
1. What does five charlie mean to a computer program?
2. How do we display five charlie on our human readable screens?

If you are dealing heavily with Microsoft-style "path" names, and you
are not dealing with yen symbols, typical of 99% of U.S. client
server programs, then there is no problem.
If you are dealing heavily with yen symbols, but do not give a hoot
about the back slash (as, for example, a Unisys or IBM mainframe
program running Japanese applications), then there is no problem.
The problem comes when you are running on a PC platform that
may be running a lot of C programs, for instance, and the display of
the 5C really ought to be done in two different ways, depending on
the particular usage at the time. Since the yen symbol is a human
symbol, the obvious solution is to have all operating systems and
all computer programs stop using 5C as a meaningful symbol. After
all, neither an OS not a program will care whether it is given a
"yen symbol" or a "backslash" or a "smiley face" or the "seven
stigmata of the crucified Christ" so long as the functionality of
the code is preserved.



Ross
Roebuck, South Carolina
http://community.webshots.com/user/ross_klatte
http://www.geocities.com/sparklecitysue/
http://www.savethescv.org/NC%20Division.htm
Norman Diamond
2004-01-02 02:09:54 UTC
Permalink
Post by Ross Klatte
I don't think the Font is a problem.
Yes you do.
Post by Ross Klatte
1. What does five charlie mean to a computer program?
That might or might not be the original poster's programming problem; I
still haven't figured out if it is or not. But Japanese isn't an issue in
this problem 1, so we're not done yet.
Post by Ross Klatte
2. How do we display five charlie on our human readable screens?
The font. Whether or not this is part of the original poster's problem,
this problem 2 says you do think the font is a problem.
Post by Ross Klatte
If you are dealing heavily with Microsoft-style "path" names, and you
are not dealing with yen symbols,
If you're using Japanese systems and if you're dealing heavily with
Microsoft-style path names then you ARE dealing with yen symbols.
Post by Ross Klatte
typical of 99% of U.S. client server programs,
Maybe the only part of the original poster's problem that is clear is that a
99% US solution is kind of meaningless. He needs to deal with Japanese.
Post by Ross Klatte
If you are dealing heavily with yen symbols, but do not give a hoot
about the back slash (as, for example, a Unisys or IBM mainframe
program running Japanese applications), then there is no problem.
True, but you forgot a more relevant example: a PC platform with
Microsoft-style pathnames running Japanese applications. In this setting
you do not usually give a hoot about the backslash and there is no problem.
(The occasions when you do give a hoot about the backslash, if you're
writing documentation for export to ASCII-speaking countries, then you
either choose a font that has a single-byte backslash character or you use a
wide character to display it.)
Post by Ross Klatte
The problem comes when you are running on a PC platform that
may be running a lot of C programs, for instance, and the display of
the 5C really ought to be done in two different ways, depending on
the particular usage at the time.
Still not an issue. If it's being displayed in a Japanese environment then
it displays as a yen sign.
Post by Ross Klatte
the obvious solution is to have all operating systems and
all computer programs stop using 5C as a meaningful symbol.
Yes, there exist a few programming languages that were designed to be usable
in most national environments, without using code points that map onto
different national symbols in different character sets. By coincidence,
some of these programming languages also make it possible to inform
programmers about invalid pointer dereferences instead of leaving it to
crash customers' systems after delivery. So, though possibly for different
reasons, you and I both wish that certain popular programming languages
would stop being used. But meanwhile, i.e. until hell freezes over, we're
going to have to live with popular programming languages.
Anonymous Gaijin
2004-01-02 06:31:29 UTC
Permalink
Post by Dale Walker
I also have a couple of clients who would prefer to have a Japanese
system font but also work a lot with pathnames. One of my clients is a
married couple. The husband is a web designer and the wife is a
Japanese interpreter. They only have one PC and divorce has been
mentioned on at least one occasion.
If the husband doesn't use any Japanese in his web design work (which
is the impression given by the last sentence) then he should just be
able to set the font of the program he uses to edit his code to a font
that doesn't support Japanese, e.g. Courier, to stop the backslashes
appearing as yen signs. If he can't change the font of the editor, get
one that can, e.g. a second hand copy of Visual C++ 4.0 for 1000 yen:

<http://page.auctions.yahoo.co.jp/jp/auction/63408918>

If the coding includes Japanese, e.g. in comments or string values,
then you are stuck with yen signs as backslashes, but in a lot of
cases a forward slash will also work as a path delimiter on Windows so
you can use this instead of a backslash.
Dale Walker
2004-01-02 08:20:34 UTC
Permalink
Post by Anonymous Gaijin
Post by Dale Walker
I also have a couple of clients who would prefer to have a Japanese
system font but also work a lot with pathnames. One of my clients is a
married couple. The husband is a web designer and the wife is a
Japanese interpreter. They only have one PC and divorce has been
mentioned on at least one occasion.
If the husband doesn't use any Japanese in his web design work (which
is the impression given by the last sentence) then he should just be
able to set the font of the program he uses to edit his code to a font
that doesn't support Japanese, e.g. Courier, to stop the backslashes
appearing as yen signs.
Whilst I think Courier is possibly one of the most unreadable fixed
font I know of, yes, that deals with any 'in application' coding but
what about any dialog boxes, toolbars & property panels? With most
programming done with IDE's these days (especially web design) those
panels and extras also need to show paths correctly.
Post by Anonymous Gaijin
If he can't change the font of the editor, get
Erm, Visual C++ might work for programming but not really much use for
web design. Besides, Visual C++ is M$ bodgeware and I reckon the
Borland C++ Builder runs rings around it.
Post by Anonymous Gaijin
If the coding includes Japanese, e.g. in comments or string values,
then you are stuck with yen signs as backslashes, but in a lot of
cases a forward slash will also work as a path delimiter on Windows so
you can use this instead of a backslash.
As it happens, Multi User Interface (MUI) CD sorts the problem out by
allowing Windows to switch system fonts on logging in. This means all
I'd have to do is set up one account for English and one for Japanese.
Trouble is, the MUI is only available in multiple licences and not
individuals. They would also have to upgrade to XP Pro.

-------------------------------------------------------------
Dale Walker London Techno Events Saiko!
***@sorted.org ***@sorted.org ***@sorted.org
London, UK london.sorted.org saiko.sorted.org
Norman Diamond
2004-01-02 10:13:12 UTC
Permalink
Post by Dale Walker
Post by Dale Walker
I also have a couple of clients who would prefer to have a Japanese
system font but also work a lot with pathnames. One of my clients is a
married couple. The husband is a web designer and the wife is a
Japanese interpreter. They only have one PC and divorce has been
mentioned on at least one occasion.
As it happens, Multi User Interface (MUI) CD sorts the problem out by
allowing Windows to switch system fonts on logging in. This means all
I'd have to do is set up one account for English and one for Japanese.
Trouble is, the MUI is only available in multiple licences and not
individuals. They would also have to upgrade to XP Pro.
They could buy a second computer, for not much more than the price of XP
Pro, though it would only include XP Home preinstalled on it. XP Home would
be enough for the person who wants to use Japanese. I'd also guess a second
computer would cost less than Microsoft probably charges for the MUI CD, but
I don't know what they really charge so this is just a guess.

On the other hand, if they both need to use a PC for their businesses and
they're too cheap to buy a second PC then maybe they had the right idea in
the first place. On the third hand, if both of them are too cheap to buy a
second PC for the purpose, then they were made for each other.
Dale Walker
2004-01-02 17:06:24 UTC
Permalink
On Fri, 2 Jan 2004 19:13:12 +0900, "Norman Diamond"
Post by Norman Diamond
Post by Dale Walker
Post by Dale Walker
I also have a couple of clients who would prefer to have a Japanese
system font but also work a lot with pathnames. One of my clients is a
married couple. The husband is a web designer and the wife is a
Japanese interpreter. They only have one PC and divorce has been
mentioned on at least one occasion.
As it happens, Multi User Interface (MUI) CD sorts the problem out by
allowing Windows to switch system fonts on logging in. This means all
I'd have to do is set up one account for English and one for Japanese.
Trouble is, the MUI is only available in multiple licences and not
individuals. They would also have to upgrade to XP Pro.
They could buy a second computer, for not much more than the price of XP
Pro, though it would only include XP Home preinstalled on it. XP Home would
be enough for the person who wants to use Japanese. I'd also guess a second
computer would cost less than Microsoft probably charges for the MUI CD, but
I don't know what they really charge so this is just a guess.
On the other hand, if they both need to use a PC for their businesses and
they're too cheap to buy a second PC then maybe they had the right idea in
the first place. On the third hand, if both of them are too cheap to buy a
second PC for the purpose, then they were made for each other.
So now your having a prod at the status of one my clients. Very
productive to the argument! Even if they were 'too cheap' as you put
it, the issue still remains there is a problem that deserves to be be
discussed.

As a matter of fact, it's not a matter of cost but space. Their little
box room they use as an office is rammed as it is. He uses the PC
during the day, she at night. Apart from one little font issue, one PC
is all they need. Not everyone wants computers all over their house
(Although mine has about 5 at the moment).

-------------------------------------------------------------
Dale Walker London Techno Events Saiko!
***@sorted.org ***@sorted.org ***@sorted.org
London, UK london.sorted.org saiko.sorted.org
Norman Diamond
2004-01-03 04:00:19 UTC
Permalink
Post by Dale Walker
So now your having a prod at the status of one my clients. Very
productive to the argument! Even if they were 'too cheap' as you put
But you said this was for their BUSINESSES, for Pete's sake. And you didn't
Post by Dale Walker
As a matter of fact, it's not a matter of cost but space.
OK, this time you gave a pretty good explanation. But it's still not that
difficult. If they have only one partition on one hard disk and they don't
want to take a chance on shrinking the partition to make a second one, then
add a second hard disk. With either two partitions or two disks, they could
add a Windows XP Home installation and dual boot. One installation could
have all Japanese defaults even if it's not the Japanese version of the OS
(though if it's not the Japanese version of the OS then it will probably
have fewer bugs than the real Japanese version, which will complicate
matters if they have to help their clients debug something on a real
Japanese version).

Or get a second computer anyway, and use a KVM switch so they still only
need one monitor, one keyboard, and one mouse. Actually a laptop would take
less space than this.

Though, a second hard disk would probably suit them best. By the way, if
they already have Windows XP Home on this computer then they can add a
second installation using the exact same CD, they won't even have to buy a
second one. They'll have to activate it again but so what, it is the same
computer. They can set all Japanese defaults in one installation of Windows
XP and all whatever their other defaults are in the other installation.
j***@i-never-read-hotmail.com
2003-12-25 10:20:07 UTC
Permalink
Post by Jeffrey Ellis
I seem to having a pretty significant problem with the way the Japanese
localized version of our application is behaving as regards the "/"
character.
I'll cut to the chase, since you are talking about \ not /.

The Japanese equivalent of extended ASCII is JIS X 0201. In the code-point
where ASCII has \, JIS 201 has a Yen symbol. If you are not aware of
this when you are localizing software in or out of a Japanese locale, you
can screw up badly on (Windows) directory/file paths, etc.
--
Jim Breen
http://www.csse.monash.edu.au/~jwb/
Computer Science & Software Engineering,
Monash University, VIC 3800, Australia
ジム・ブリーン@モナシュ大学
Jeffrey Ellis
2003-12-25 19:29:53 UTC
Permalink
Post by j***@i-never-read-hotmail.com
Post by Jeffrey Ellis
I seem to having a pretty significant problem with the way the Japanese
localized version of our application is behaving as regards the "/"
character.
I'll cut to the chase, since you are talking about \ not /.
The Japanese equivalent of extended ASCII is JIS X 0201. In the code-point
where ASCII has \, JIS 201 has a Yen symbol. If you are not aware of
this when you are localizing software in or out of a Japanese locale, you
can screw up badly on (Windows) directory/file paths, etc.
Hi, Jim--

I guess I've now become very aware of this. What is the usual method for
dealing with it? I.e, how do we make the paths correct again?

All My Best,
Jeffrey
j***@i-never-read-hotmail.com
2004-01-02 08:55:49 UTC
Permalink
Post by Jeffrey Ellis
Post by j***@i-never-read-hotmail.com
The Japanese equivalent of extended ASCII is JIS X 0201. In the code-point
where ASCII has \, JIS 201 has a Yen symbol. If you are not aware of
this when you are localizing software in or out of a Japanese locale, you
can screw up badly on (Windows) directory/file paths, etc.
I guess I've now become very aware of this. What is the usual method for
dealing with it? I.e, how do we make the paths correct again?
Erm. It doesn't matter whether you are seeing a back-slash or a Yen
symbol, it will still be treated as a directory marker in Windows
paths. Unless, as another person commented, you are mucking up the
usage of \ for escaping individual characters, your problem will be
largely visual.
--
Jim Breen
http://www.csse.monash.edu.au/~jwb/
Computer Science & Software Engineering,
Monash University, VIC 3800, Australia
ジム・ブリーン@モナシュ大学
Anonymous Gaijin
2003-12-26 06:40:19 UTC
Permalink
Post by j***@i-never-read-hotmail.com
Post by Jeffrey Ellis
I seem to having a pretty significant problem with the way the Japanese
localized version of our application is behaving as regards the "/"
character.
I'll cut to the chase, since you are talking about \ not /.
The Japanese equivalent of extended ASCII is JIS X 0201. In the code-point
where ASCII has \, JIS 201 has a Yen symbol. If you are not aware of
this when you are localizing software in or out of a Japanese locale, you
can screw up badly on (Windows) directory/file paths, etc.
Not sure if this is another example of Breen humour, but as Louise
pointed out a computer has no concept of a yen sign and as far as it
is concerned, the character is a path delimiter on Windows whatever it
may look like to the user. There is a common mistake made by novice
programmers on Windows due to the fact that this character is also an
escape character in strings, e.g.

Wrong:
fopen("C:\temp\test.txt","r");

Correct:
fopen("C:\\temp\\test.txt","r");

But this is true of any language version of Windows.
Jeffrey Ellis
2003-12-27 09:41:40 UTC
Permalink
Post by Anonymous Gaijin
Post by j***@i-never-read-hotmail.com
Post by Jeffrey Ellis
I seem to having a pretty significant problem with the way the Japanese
localized version of our application is behaving as regards the "/"
character.
I'll cut to the chase, since you are talking about \ not /.
The Japanese equivalent of extended ASCII is JIS X 0201. In the code-point
where ASCII has \, JIS 201 has a Yen symbol. If you are not aware of
this when you are localizing software in or out of a Japanese locale, you
can screw up badly on (Windows) directory/file paths, etc.
Not sure if this is another example of Breen humour, but as Louise
pointed out a computer has no concept of a yen sign and as far as it
is concerned, the character is a path delimiter on Windows whatever it
may look like to the user. There is a common mistake made by novice
programmers on Windows due to the fact that this character is also an
escape character in strings, e.g.
fopen("C:\temp\test.txt","r");
fopen("C:\\temp\\test.txt","r");
But this is true of any language version of Windows.
Ah...That may be the issue indeed. We aren't doing the second method at all.

Thanks!

All My Best,
Jeffrey
Louise Bremner
2003-12-30 02:34:52 UTC
Permalink
Post by Jeffrey Ellis
Post by Anonymous Gaijin
There is a common mistake made by novice
programmers on Windows due to the fact that this character is also an
escape character in strings, e.g.
fopen("C:\temp\test.txt","r");
fopen("C:\\temp\\test.txt","r");
But this is true of any language version of Windows.
Ah...That may be the issue indeed. We aren't doing the second method at all.
If so, it's interesting that you haven't noticed it occurring with the
non-Japanese versions....

________________________________________________________________________
Louise Bremner (log at gol dot com)
If you want a reply by e-mail, don't write to my Yahoo address!
Brandon Berg
2003-12-30 07:56:48 UTC
Permalink
Post by Louise Bremner
Post by Jeffrey Ellis
Post by Anonymous Gaijin
There is a common mistake made by novice
programmers on Windows due to the fact that this character is also an
escape character in strings, e.g.
fopen("C:\temp\test.txt","r");
fopen("C:\\temp\\test.txt","r");
But this is true of any language version of Windows.
Ah...That may be the issue indeed. We aren't doing the second method at all.
If so, it's interesting that you haven't noticed it occurring with the
non-Japanese versions....
Perhaps they simply didn't understand that the yen sign would have to be
escaped just like a backslash. This still raises the question of why this
problem wasn't encountered with the Korean version, unless the Korean
translator just knew better.
Continue reading on narkive:
Loading...