I am converting an application from code-page 437 to UTF-8.
Previously Putty was set to interpret data as CP437, I changed that to
UTF-8 (Settings, Window, Translation)

Putty displays correctly the UTF-8 emitted by the app but it now ignores
the DEC graphics characters used for drawing lines and boxes as borders
around the applications text-mode windows

Here's the problem, Cut & paste from Putty window:

    1 ────┼─────────┼─────────┼─────────┼─────────┤
    2 ────┼─────────┼─────────┼─────────┼─────────┤
    3 ────┼─────────┼─────────┼─────────┼─────────┤
    4 ────┼─────────┼─────────┼─────────┼─────────┤
    5 ────┼lqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqk──────┤
    6 ────┼x                               x──────┤
    7 ────┼x                               x──────┤
    8 ────┼xFoo {         } ©RGB     я  ...x──────┤
    9 ────┼xBar {         } ®Acme    ئ  .  x──────┤
   10 ────┼x                               x──────┤
   11 ────┼mqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqj──────┤
   12 ────┼─────────┼─────────┼─────────┼─────────┤
   13 ────┼─────────┼─────────┼─────────┼─────────┤
   14 ────┼─────────┼─────────┼─────────┼─────────┤
   15 ────┼─────────┼─────────┼─────────┼─────────┤
   16 ────┼─────────┼─────────┼─────────┼─────────┤

The background grid is displayed using Unicode characters U+2500 and
U+253c sent in UTF-8 encoding (e2 94 80 and e2 94 bc). That works.

The border "lqqq...qqqk" is sent using the DEC graphics feature of
VT100/ANSI terminals. Esc ( 0 is sent to switch the G0 set to the
line-drawing character set. This is being ignored.

Here's part of a `hexdump -C` of data captured using Linux's `script`
command:

80 e2 94 80 e2 94 80 e2  94 80 e2 94 80 e2 94 a4  |................|
1b 5b 32 34 3b 35 36 48  1b 5b 35 3b 31 31 48 1b  |.[24;56H.[5;11H.|
28 30 1b 5b 33 34 6d 1b  5b 34 30 6d 6c 71 71 71  |(0.[34m.[40mlqqq|
71 71 71 71 71 71 71 71  71 71 71 71 71 71 71 71  |qqqqqqqqqqqqqqqq|
71 71 71 71 71 71 71 71  71 71 71 71 6b 1b 28 42  |qqqqqqqqqqqqk.(B|
1b 5b 30 6d 1b 5b 33 37  6d 1b 5b 34 30 6d 1b 5b  |.[0m.[37m.[40m.[|
36 3b 31 31 48 1b 28 30  1b 5b 33 34 6d 1b 5b 34  |6;11H.(0.[34m.[4|
30 6d 78 1b 28 42 1b 5b  30 6d 1b 5b 33 37 6d 1b  |0mx.(B.[0m.[37m.|
5b 34 30 6d 20 20 20 20  20 20 20 20 20 20 20 20  |[40m            |
20 20 20 20 20 20 20 20  20 20 20 20 20 20 20 20  |                |

You can see the UTF-8 triplets for background grid.
You can see Esc ( 0 to select DEC graphics (line draw) and
later Esc ( B to reselect normal characters. There is a colour selection
inside that sequence but the following test suggests that is OK.

I updated Putty to 0.62 - same issue.

As a further confirmation, I set Putty Window Translation to CP437 and
re-ran the program ...

ΓöÇ3öΓöÇΓöÇΓöÇΓöÇΓö╝ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö╝ΓöÇ
ΓöÇ4öΓöÇΓöÇΓöÇΓöÇΓö╝ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö╝ΓöÇ
ΓöÇ5öΓöÇΓö┌───────────────────────────────┐ÇΓöÇΓö╝ΓöÇ
ΓöÇ6öΓöÇΓö│                               │ÇΓöÇΓö╝ΓöÇ
ΓöÇ7öΓöÇΓö│                               │ÇΓöÇΓö╝ΓöÇ
ΓöÇ8öΓöÇΓö│Foo {         } ┬⌐RGB     ╤Å  ...ΓöÇΓö╝ΓöÇ
ΓöÇ9öΓöÇΓö│Bar {         } ┬«Acme    ╪ª  .│ÇΓöÇΓö╝ΓöÇ
Γö10öΓöÇΓö│                               │ÇΓöÇΓö╝ΓöÇ
Γö11öΓöÇΓö└───────────────────────────────┘ÇΓöÇΓö╝ΓöÇ
Γö12öΓöÇΓöÇΓöÇΓöÇΓö╝ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö╝ΓöÇ
Γö13öΓöÇΓöÇΓöÇΓöÇΓö╝ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö╝ΓöÇ
Γö14öΓöÇΓöÇΓöÇΓöÇΓö╝ΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓöÇΓö╝ΓöÇ

As expected the background grid is now interpreted as triplets of CP437
characters rather than as UTF-8 3-byte characters.

But interestingly, the window border is now displayed correctly. This
suggests that the relevant Putty settings for "handling of line-drawing
characters" are OK.

What is wrong?

http://www.chiark.greenend.org.uk/~sgtatham/putty/wishlist/utf8-plus-...

SecureCRT displays the output as I expected/wanted but that obviously
doesn't affect the factors that determined the priority of the Putty
enhancement.

Does anyone have ideas for workarounds? The app is developed using a
tool that has it's own variant of termcap. I have at my disposal

   gs=\E(0      start graphics mode
   ge=\E(B      end graphics mode
   gb=lmkjqx    a block of 6 ASCII bytes representing ┌ ┐ └ ┘ ─ │

I can't put UTF-8 sequences in gb.

I could leave gs and ge empty and put in something like ++++-| or ..`'-|
for gb but the resulting borders would be a bit sad.

Any ideas?