Re: backspace policy (Re: kernel tty patches)

[Ienup Sung <ienup.sung@xxxxxxxxxxx>]

Hi,

I think we might have a bit different definitions being used in the emails
of you and me. And so, as you also pointed out, before I start any further
discussion on this issue, I would like to clear one thing first regarding
this backspace because in my understanding, after seeing your email, we are
probably talking about samething in different view points and/or sticking to
one part only:

When there is a backspace character code that is given to kernel's line
discipline module as an input character to indicate that there has been
a backspace key press, I think we are in synch in terms of how this,
a backspace key press (or, in other words, a backspace code), should trigger
what kind of edit operation, i.e., erasing a character (which could occupying
one or two display screen columns) not just a column.

Also, as I wrote in previous emails to Bruno, I think we are in synch too on
the actual use of backspace codes/characters in actual erase operation, i.e.,
when the line discipline module does an "erase" in terms of output,
a character that occupies a single column  will only require a single
"backspace" (0x8) and then  a single "space" (0x20) to rubout the charcter.
For a character that occupies two screen columns, the "erase" will require
two "backspace" (0x08) codes and then two "space" (0x20) codes to move
the cursor to "the_current_position - 2" and then "  " so that the double
width character will be rubbed out.

Please let me know whether we are talking about samething or not.

With regards,

Ienup

PS. Regarding the multiple modes, as I implied in the emails before, if
there is one of those modes supported, pretty much everyone would be happy.
If two or more modes can be supported, we can make more people happy but
I also don't think that's something that has higher priority.


] Date: Sat, 27 Jan 2001 11:27:46 +0900
] From: Tomohiro KUBOTA <tkubota@xxxxxxxxxxx>
] Subject: backspace policy (Re: kernel tty patches)
] To: ienup.sung@xxxxxxxxxxx
] Cc: linux-utf8@xxxxxxxxxxxxxxxxxxxx
] MIME-version: 1.0
] Mail-Reply-To: tkubota@xxxxxxxxxxx
] 
] Hi,
] 
] At Fri, 26 Jan 2001 13:01:38 -0800 (PST),
] Ienup Sung <ienup.sung@xxxxxxxxxxx> wrote:
] 
] > All commercial CJK terminals and terminal emulators in my understanding
] > work that way, i.e., a backspace will erase a character and depending on
] > the what is the character being erase, it could be two screen columns or
] > one screen column, and also some public domain terminal emulators like
] > hanterm and so on if I remember correctly.
] 
] Really?   We are discussing about 'backspace' code (0x08 in ASCII), not
] about pressing 'backspace' key.  (I agree pressing 'backspace' key should
] erase one _character_.  Thus, for example, tcsh-i18n issues two 'backspace'
] code when 'backspace' key is pressed after double-width character.)
] 
] I checked hanterm and found that 'backspace' code erases one _column_
] even after one doublewidth character.  It is just as I expected.  I also
] checked the following terminals:
]  - Cxterm (big5, gb, ks, and jis mode)
]  - Rxvt (EUC mode)
]  - Eterm, Wterm and so on which are derive from Rxvt (EUC mode)
]  - Kterm
]  - Hanterm
]  - Kon2 (Kanji console, not X terminal emulator)
]  - Jfbterm (ISO-2022 console on Linux framebuffer)
]  - recent Xterm
]  - Japanese version of MS-DOS
]  - MS-DOS mode and MS-DOS prompt in Japanese MS-Windows
]  - Tera Term (telnet software for Windows)
]  - Telnet which is included in MS-Windows
]  - Japanese-enabled NCSA telnet (Macintosh version)
]  - N88-BASIC(86) (shipped with very old Japanese computer)
] and found that my strong opinion (so strong that I wrote CJK people
] would fall into panic if compatibility is abandoned) was confirmed.  All
] softwares which I used on these terminals in my life must be based on
] the same policy because they worked well on the terminals.
] 
] If one 'backspace' code (0x08) erases one character (two columns) for
] doublewidth character, many softwares (all doublewidth-enabled softwares)
] such as Emacs will fail to work well.
] 
] Though your opinion (implementing two 'modes') might be useful, I
] think it can bring confusion.  Some softwares will be developed based
] on one mode and others will be based on another mode.  Users will have
] to remember whether mode to be used for each software.  
] 
] Please check again your 'commercial' softwares.  Doesn't your shell
] interpret one 'backspace'key pressing into two 'backspace' codes?
] Doesn't your OS's line editor do such a thing?  Both of them are
] likely.  Please check with C code printf("X\bY") where X is a
] doublewidth character.
] 
] ---
] Tomohiro KUBOTA <kubota@xxxxxxxxxx>
] http://surfchem0.riken.go.jp/~kubota/
] "Introduction to I18N"
] http://www.debian.org/doc/manuals/intro-i18n/
-
Linux-UTF8:   i18n of Linux on all levels
Archive:      http://mail.nl.linux.org/lists/