Connecting orthogonal and diagonal box-drawing glyphs

[Logo121]

Continuation of #2197.

This problem exists before the introduction of the Unicode 16.0 box-drawing diagonals, where the "diamond" diagonals do not connect with the orthogonals ones 🮮┼

To my understanding the "cell-masking" done in those diagonals are to remove the gaps that appear when 2 (non-parallel) lines join, which I followed for the diagonal lines in 16.0. However, in the schematic symbols section we have these: Which kinda assumes connection with diagonals and orthogonal lines (at least for these 4 characters. I can't really find examples of circuit diagrams with those microcomputers so I can't tell if diagonal/orthogonal connections are needed in general). Right now connecting them results in something like this:

The potential solutions are:

1. Make all lines unclipped, and only deal with cases where gaps appear in a single character (like 🯘🯙🯚🯛)
   • For the other cases like 🮠🮡 🮢🮣, we just ignore the gap problem, or try to reduce it somehow like using a rounded cap?
   • Essentially what FairfaxHD does.
2. Modify the width of the diagonal lines so that they connect to orthogonal lines directly.
   • This is cleaner but would probably result in lines varying quite a bit in thickness. Also requires some calculations for every kind of slopes.
3. Deal with the schematic symbols directly and assume no diagonal-orthogonal connection should happen otherwise.
   • This means making 𜰉𜰊 unclipped unlike the other diagonal lines, or making the vertical connection part of 𜰐𜰑 a spline like in FairfaxHD:
   • Note that the spline form is not used in the Fullwidth equivalent (KreativeSquare), which is the Unicode reference glyph.

... or a mixture of the above.

While we're at it we might also want to disable HVContrast for all box-drawing glyphs, if that makes the situation better.

Affected glyphs:

• The "diamond" diagonals 🮠🮡🮢🮣🮤🮥🮦🮧🮨🮩🮪🮫🮬🮭🮮 (and their derivatives)
• The "corner-to-edge" diagonals 🯐🯑🯒🯓🯔🯕🯖🯗🯜🯝🯞🯟
• Maybe these 2 double diagonals 𜰟𜰠
• Some schematic symbols 𜰉𜰊𜰐𜰑

[be5invis]

Well these symbols were included in Sharp MZ computers and their original fonts are bitmap fonts, so I do not think they ever considered this problem seriously... I think for the schematics we should not clip them at the cell border -- and only for them because they are designed to be connected to H/V lines. For the others? keep the current design.

[be5invis]

And ps. maybe you should consider make the logic gate symbols shorter vertically under NWID -- match the height of the horizontal resister symbol.

[Logo121]

I think for the schematics we should not clip them at the cell border -- and only for them because they are designed to be connected to H/V lines. For the others? keep the current design.

ok

And ps. maybe you should consider make the logic gate symbols shorter vertically under NWID -- match the height of the horizontal resister symbol.

I did consider this, but I wasn't sure what could function as a "non-inverted" version of LOGIC GATE INVERTED INPUTS if I make them not stretch to cell boundary.
...Well, not like the original charset has an equivalent anyway, so idk

[github-actions[bot]]

This issue is stale because it has been open 60 days with no activity. Remove stale label or comment or this will be closed in 15 days.