Turn a name — or a whole guest list — into 3D printable script name tags for wedding tables, parties and desks. Type it, see it, download STL or 3MF.
The hard part is not drawing the letters, it is making them hold together. Joined script is full of gaps: a capital that never quite reaches the next letter, an i-tittle floating free, two lines that only touch if you nudge them into each other. Left alone that prints as a heap of loose pieces. NamR closes those gaps the way a signwriter would — tightening the spacing, sliding the surname to where the two lines interlock, stemming each accent to its own letter — and only bridges what is genuinely too far apart. Every tag comes out as one connected solid: no supports, no glue, no assembly.
- Batch the whole list. Paste names or drop in a CSV, and every tag is packed onto printer beds and zipped up with a manifest, ready to slice.
- Reads face-down on the glass, so the visible side comes off smooth.
- Polish and Latin Extended throughout — ą ć ę ł ń ó ś ź ż keep their accents, joined to the letter they belong to.
- Eight script faces bundled, Savoye LET and Brush Script from your own computer, or load any font file; it is parsed in the browser.
- Nothing is uploaded. No account, no server. It is a static site, so guest names never leave your machine.
| Stage | State |
|---|---|
| Font → outlines, Polish diacritics | done |
| Mark detection (tittles, accents) | done |
| Line overlap solve | done |
| Welding + bridging → one piece | done |
| Profile decimation | done |
| 3D mesh (60° revolve) | done |
| STL / 3MF export | done |
| Web UI + 3D preview | done |
| CSV bulk import | done |
| Bed packing / batches | done |
| Bridge editor | not started |
Leaving the surname empty gives a one-line tag.
npm run check builds 16 randomly generated Polish names and three one-line
tags in every bundled face, plus Savoye LET and Brush Script where the
machine has them. All 190 come out as a single watertight, correctly oriented
piece, and every two-line tag is held together on at least two different pairs
of letters, at the default 30mm font height. Short-link counts tags whose every added link is 2mm or less,
the length that reads as part of a stroke.
face pass short-link longest link
AlexBrush 19/19 4/16 4.2mm
Damion 19/19 4/16 8.0mm
GreatVibes 19/19 3/16 13.3mm
Lobster 19/19 5/16 6.8mm
Norican 19/19 10/16 4.3mm
Pacifico 19/19 4/16 4.7mm
Sacramento 19/19 4/16 6.7mm
Yellowtail 19/19 7/16 4.7mm
SavoyeLET 19/19 2/16 5.0mm
BrushScript 19/19 7/16 3.4mm
190/190 pass · 50/160 two-line tags with every link ≤2mm · 233ms/tag
Whether the lines have been pushed so far into each other that the name stops
reading is not scored: shared ink area misses a thin swash cutting through a
bowl, and so does the share of each letter's footprint the other line covers.
npm run check -- --sheets draws every tag flat into out/check/ to be judged
by eye.
A STEP export of the original CAD model settles it. Parsing that B-rep's 462 circles:
- every one shares a single axis, direction
(1,0,0)— parallel to the baseline - all centred on one line at
Y = -11.906, Z = 0 - radii run
5.000to46.534; the minimum is exactly the 5mm offset max Z / max radius = 40.30 / 46.534 = sin 60°exactly
So the tag is the merged profile revolved 60° about an axis parallel to the
baseline, 5mm past the lowest ink. The R50.00 in angle.png is
construction geometry for the sweep path and never reaches the solid, which is
why it looked inconsistent with the 5mm offset: a sweep along an arc whose
centre lies in the profile plane is a revolve, so the path radius drops out.
The name reads off the alpha = 0 face. mode: 'extrude' is kept as a flat
plate variant.
Pure client-side. Vite + TypeScript + React, Three.js preview, Web Worker pool for the geometry so a large CSV never blocks the UI.
src/geom/
types.ts Pt / Ring / Poly, shoelace area (CCW positive)
clipper.ts WASM Clipper wrapper: union, offset, closing, erosion
text.ts font → tagged contours, Polish fallback, mark detection
connect.ts line overlap, mark stems, MST bridging
simplify.ts Douglas-Peucker decimation
sweep.ts profile → watertight mesh (60° revolve)
export.ts binary STL, 3MF, manifold check
tag.ts the whole pipeline for one tag
csv.ts guest list parsing, delimiter sniffing
pack.ts plate packing: maximal rectangles, best of many strategies
batch.ts many names -> plates -> zip + manifest
-
Outlines. opentype.js, with pair kerning. Missing Polish glyphs fall back to the base letter and are reported.
-
Mark detection. A glyph whose rings form more than one island —
iand its tittle,ńand its acute — yields marks for every island but the largest. Provenance is kept per contour. -
Stems. Each mark is tied to its own letter. Proximity alone would graft an
itittle onto whichever letter happens to be nearest, which on a tight script is often the wrong one. -
Each line, joined on its own. A script is meant to join up, so a gap between letters is closed by pulling them together rather than bridging across it — the result reads as handwriting instead of two letters wired together. Each letter may travel
letterTighten.It closes the gap between the stroke ends meant to meet, not between whichever points happen to be nearest. That is the nearest approach at the font's own spacing, preferring one near the baseline if it is not much further off, and the pull stops while the rest of both letters is still clear of the weld. Pulled together by nearest points, Savoye LET's r met the d before it at the shoulder and fused there, and the letters trapped specks that printed as blobs. It also stops at contact and leaves the join its width from the weld: pulling further makes strokes meeting at a shallow angle cross, and the lens between the crossings prints as a slit. If some other part already sits within welding distance as set — Savoye LET's r keeps its top knob half a millimetre from the letter before — the letter is eased just clear, or it would hang off that knob.
Letters that only touch, or sit within welding distance, meet at a point the weld fills no wider than the gap, and that snaps however wide the connectors are set. Every such join gets a connector of that width laid along the two strokes where they meet; a join already overlapping by a connector's width is left alone.
Accents are stemmed, near-misses welded, and any letter still apart is linked at those same stroke ends, low down so the link runs along the baseline like the exit stroke it continues. Only then is the line placed. Solving both lines at once let a letter be held on only through the other line, so either name on its own fell apart.
-
Line placement. Sliding the surname straight up is the wrong single degree of freedom: two lines of script interlock at particular horizontal offsets, where a descender drops into the gap between two ascenders. Depth has to be searched too — the shallowest overlap that welds is often not the one that reads best, and pushing the lines further into each other frequently removes a strut altogether.
Readability comes first. A placement is costed mainly by the ink the two lines share, as a fraction of the smaller line, rising steeply past 2%; welds earn credit only up to the two a tag needs. Weighted the other way, a third weld bought several percent more overlap and the lines were pushed until a first name's letters sat inside the surname's, which bold strokes then fused into solid fills. A short strut is cheap next to that.
The two finished lines must be tied on two different letter pairs, because one contact is a hinge that snaps. A shortlist — the five best placements and the best at each depth — is re-ranked by the links that step would really add, each costing more steeply past 2.5mm, and a second pair out of reach costs as much as a broken tag. Without that the search settled on a single weld and left a long diagonal strut to some distant letter.
A coarse sweep of both axes on heavily decimated outlines, then a local refinement around each of the leaders at finer resolution.
-
Closing. Morphological closing (dilate then erode by
weldRadius) welds gaps up to2 × weldRadiuswithout fattening the letterforms. -
Bridging. Anything that survives is joined by a minimum spanning tree over inter-island distance: n islands need exactly n−1 bridges, each placed where the letters already almost touch. A second pass runs after filleting, because two strokes meeting at a single point come back from the union as one self-touching ring and only fall apart once the pinch is resolved — a contact with no width was never a connection worth counting.
-
Fillet and tidy. A small closing rounds where connectors meet strokes, then trapped slivers are filled. A hole has to fail two tests before it goes. Provenance: a counter is enclosed by one glyph on its own, so an open bowl that welding seals still counts — Yellowtail's R is one, and judging by the raw outline alone filled it into a blob. Size: the gap between two adjacent letters belongs to neither of them, but it is the eye of the script, and filling it turns the word solid. Only a hole that is both foreign and tiny is an artifact.
-
Decimate. Douglas-Peucker at 0.02mm. Cuts points ~3× for 0.07% area error, and clears the slivers that make ear-clipping drop a triangle.
-
Mesh. Earcut caps plus a quad band per boundary edge. No 3D booleans. Checked watertight before export.
Every step above is editable by hand afterwards: any link can be removed or dragged, and the surname can be repositioned directly, with the solver respecting those choices on the next rebuild.
The preview orbits in the readable face's own frame, not the world's. That face is tilted by the sweep angle, so orbiting about world-up merely rolls the name diagonally across the screen instead of walking around it — the writing never sits level and there is no way to get a side view. Building the basis from the face itself (baseline right, its own up, its normal out) makes a horizontal drag mean "look from the side" and keeps the name level throughout. Home is that face, a fraction off-axis so the depth reads.
The face you read is the one at the far end of the sweep, whose normal is
(0, -sin a, cos a). Exports rotate the tag by 180 - a about X so that
normal becomes -Z and the readable face beds against the glass, coming off
with the smooth finish. The preview keeps the as-built pose, which reads
better on screen.
A guest list goes in as CSV or pasted text. The delimiter is sniffed rather than assumed — Polish Excel writes semicolons — the BOM is stripped, a header row is detected if present, and a single-column file is split on the last space so multi-part given names survive.
Tags are packed onto as few plates as possible. Every free rectangle on a plate is tracked, and a tag may be turned a quarter about Z, which keeps the reading face on the glass; that is what fills the strip left at the end of a row, too narrow for another tag lying flat but deep enough across two rows for a turned one. No single ordering or placement rule wins on every list, so all twenty combinations are tried, then a few hundred seeded shuffles, and the result with the fewest plates and the fullest early ones is kept. The same list always packs the same way.
On random 40–150 name lists that averages 4.75 plates on a 256mm bed, against 5.17 for one fixed rule and a floor of 4.33 set by the tags' area alone; 82 tags usually land on exactly the number of plates their area requires. It takes well under a second, next to the seconds spent building the tags.
The download is a zip of either one file per plate or one per tag, plus a
manifest.csv naming every tag, its plate and any warning.
Morphological closing is extensive — the result always contains the input — so
it can never split a shape. The polygonal approximation of its round joins can,
though: the erosion cuts marginally deeper than the dilation grew, and on a
weld only as wide as the radius that severs the piece. Geom.close unions the
input back in to enforce the guarantee the maths already promised.
clipper2-js is not usable. Its negative offsets return garbage and miter
joins drop an edge. Erosion is load-bearing here for both welding and the
minimum-feature check. js-angusj-clipper (WASM Clipper 6.4.2) is exact on
every case tested; its wasm is a base64 data URI, so it needs no asset
plumbing in a worker or on Pages.
Angular is a phantom dependency. clipper2-js declares @angular/* as
peers purely because it is packaged with ng-packagr — 42MB and three
advisories for code that imports neither. .npmrc sets legacy-peer-deps.
Eight open-licensed connected scripts are bundled — Yellowtail, Pacifico, Lobster, Damion, Norican, Sacramento, Alex Brush and Great Vibes. All have complete Polish coverage and all solve to a single piece across the test names. They are fetched on demand, so only the chosen one is downloaded.
Savoye LET and Brush Script belong to Monotype and cannot be served from here,
but nearly every Mac already has both. The picker reads the visitor's own
installed copy through the Local Font Access API, which needs a Chromium
browser (Chrome or Edge) and the visitor's permission when it asks. macOS
keeps Savoye LET in a .ttc collection and the browser hands over the whole
file, so src/geom/sfnt.ts cuts out the one face by its PostScript name;
opentype.js reads single fonts only.
Fine faces print poorly as drawn: at a 20mm font height 40% of Savoye LET's ink is under 0.8mm wide, two nozzle widths. Stroke weight thickens every stroke with a round offset before anything is joined, so the welds and line placement work on the bolder letters; 0.8mm leaves none of Savoye that thin and every test tag still one piece.
Anything else can be loaded from disk, .ttc collections included. It is
parsed in the browser, kept in IndexedDB so it survives a reload, and never
transmitted — which is also how to use a licensed face you already own, and
how Savoye LET and Brush Script work in Safari and Firefox. Using either
yourself is fine; serving the file from the site would be redistributing
Monotype's font to every visitor, which no licence here covers.
npm install
npm run dev # the app, at localhost:5173/NamR/
npm run build # production build into dist/
npm run spike -- Bożena Dąbrowa # headless: one tag -> out/tag.stl, .3mf
FONT=/path/to/font.ttf npm run spike # try another face
npm run check # every face against the test names
npm run check -- --sheets # ...and draw them into out/check/Pushing to main deploys to GitHub Pages. Enable it once under
Settings -> Pages -> Source: GitHub Actions.
scripts/render_stl.py software-renders an STL with a z-buffer for eyeballing
geometry without a browser.