You may use it as a FILLER TEXT =
• ~ correct proportions & weight;
• strict 1 x-h (1-px asc./desc.);
+ clonable (please keep "dpla"!)…

US-ASCII compatible with a lot of duplicates:

!"#$%&'()*+,-./0123456789:;<=>?@
ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`
abcdefghijklmnopqrstuvwxyz{|}~

Cf. Yamaw's "nt1.00" lowercase experiment (on sept. 2010)…

@Greenstar967: Hi! So, I'm going to resume commenting there.

---

Well, I'll post a few interesting fontstructions that may be valuable for the coder or the cloner (like the forthcoming voxel or/and minimalist designs that would require a few skilled fontstructers, just because I cannot multiply my related hours on this tool and elsewhere).

@tsafontstruct: hi! Since you seem already quite aware of the limitations of non outline-based fonts (like your Braille-derived fontstruction), I needn't write a verbose/scientific answer to your deep question, just these few remarks/reminders in return, thank you.

• Latin (UC & LC) and Arabic (Num) characters often share the same matrix positions when one constrains their line-based shapes to discrete dot values (e.g. “B” and “8”).

• In order to get unique letters with this size-geared limitation, one needs to increase the amount of per-glyph data, which is usually done from a larger output grid, a deeper output palette, or/and even via extra hacks (like time- or/and angle-based added informations, shiftings, textures…).

• Because of this 2×3 design in the output, the maximum number of unique 1-bit glyphs is 64 (i.e. 2^(2×3) dots) at most (i.e. counting the doubles that could move horizontally in a non-fixed layout). A good proportion of duplicates arise, thus and indeed in a context of (= out of) 94+ writing characters (which is my [or should be our?] minimum expectation for a real, usable [computer] font, i.e. the visible US-ASCII, to begin with).

• For less demanding environments, e.g. in a single-word signature or title, your issue may not appear at first, e.g. one can write « LO'FI » with minor difficulty in a 2×2 subgrid (just copy-and-paste this 5-character citation to the “User input” field of our FS viewer).

• Many derivations could start/benefit from this 2×3 grid, although they may prefer to widen it at places (for the sake of legibility in the largest proportional characters, like “M”, by extending the grid width to 4 or 5 pixels in all), which would not be efficient any longer as a whole font (minimalism versus legibility, here).

• 2×3 parts of my “2x4-2x3-lc-hex-num” and “dpla334ld” pixel creations might be useful to the ones that need alternatives to simple projections/mappings (since the untrivial difficulty of the related designs/artworks lies in the puzzle/choice of deformations, with the maximum count of pixel differences between similar glyphs, and without losing too much readibility at the prioritized [single or blocks of] chararacters).

• My mentioned (and to-be-released [I only need to edit its old doc .txt]) “duplicate-free” 2x3 includes missing or shifted US-ASCII characters, to be still useful at this excessive width. You'll see a.s.a.p. (on my personal repository, at least – my uploads look very discontinued, but I think in years, or decades, really/sorry).

• In the “dpla's 2x3 signature” fontstruction, visually and manually plotted character after character (and not converted from a script), I guess that the extreme 1 x-height is one of the main causes of the impossible legibility at times. Even if you tweak my choices, you won't be able to get a less difficult legibility, in general.

• Please bear in mind that the readibility (even the legibility, statistically) is not set characterwise, but defined after words or phrases, in a broader/trivial context. (Which cannot apply at all to my project of the smallest computer pixel fonts, which expects to get unique glyph mappings, i.e. again: no duplicate at all in US-ASCII where spatially feasible.)

• Would you have expected a common blank glyph (or none as a fallback) instead of any duplication? (I don't think so, since this would imply a drastic prioritization of your letters, i.e. you would have to define the current font after a selection of words of yours, which would be even less multi-purpose.) Discarding the repeated characters would prevent the mentioned use as a “FILLER TEXT”, which was the main idea/goal of this fontstruction, don't you?

Hope this was helpful. See you!