2D drafting

2D drafting follows AutoCAD: ribbon, command line, dynamic input, drafting-aid toggles in the status bar. If you know AutoCAD you know this. Enter it from the home page with "Create › 2D drafting".

Screen

• Ribbon: the "Home" tab has Draw, Modify, Annotation, Layers, Properties and Utilities panels. The "Insert" tab has the Block, Block Definition, Reference and 3D Views panels (see Blocks and attributes, External references and Import views from 3D). The "Annotate" tab has the text, dimension (including a drop-down for the current dimension style) and table panels. The "Manage" tab has the Cleanup (Purge PURGE), Drawing Settings (Units UNITS, Linetype LINETYPE, Lineweight LWEIGHT, Drafting Settings DSETTINGS), Styles (text, dimension, table and multileader styles) and Reference (External References palette XREF, Attach XATTACH) panels. The buttons start the same commands as typing, and the tooltip shows the command name and alias. The View tab has the Viewport Tools (UCS Icon, Navigation Bar), Model Viewports, Named Views and Visual Styles panels; the Output tab has the Plot (Plot, Batch Plot, Page Setup Manager, Preview) and Export (PDF, DXF) panels. Layout tabs add a Layout tab: New, Page Setup, Layout Viewports (Rectangular, Polygonal, Model Space / Paper Space) and Create View (Insert 3D View, Update View). See Named views, viewports and display and Layouts, viewports and plotting.
• Panel slideouts: a ▾ next to a panel title opens the less-used tools, for example Point, Construction Line, Ray, Boundary, Region, Divide, Measure and Donut in the Draw panel; Extend, Chamfer, Break, Break at Point, Join, Lengthen, Align, Edit Polyline, Polar Array, Path Array, Edit Array, Union, Subtract, Intersect, Group, Ungroup, Group Edit and the Group Selection toggle in the Modify panel; Measure, Distance, Area, ID Point, List, Mass Properties, Point Style, Units and Purge in the Utilities panel. The slideout closes when you pick a tool or move the mouse away.
• Canvas: dark model space, crosshair cursor, viewport controls "[-][Top][2D Wireframe]" at the top left, UCS icon at the bottom left. The status bar can switch between a dark and a light background.
• Properties palette: Ctrl+1. With a selection you can change layer, color and other properties directly.The right-hand properties column can collapse: click ‹/› in its title bar to hide it and the handle on the window edge to bring it back; drag the inner divider to resize (release when very narrow to collapse, double-click for the default width, arrow keys on the focused divider for fine steps). Open state and width are remembered per workspace. Tab is kept for dynamic input in CAD and does not hide panels.
• Command line: under the canvas. F2 expands the history. Ctrl+9 shows or hides it.
• Status bar: Model / Layout tabs, coordinates, GRID, SNAP, ORTHO, POLAR, OTRACK, OSNAP, DYN, full-screen drafting, annotation scale, units, save status. Left-click toggles, right-click opens the settings.
• Workspace switch: at the top right and in the status bar you can switch between "CAD drafting" and "Vector design". Same drawing, two interfaces.

Command line

• Typing at any time goes into the command line; no need to click it first.
• Enter or Space confirms. An empty Enter repeats the last command. Esc cancels and keeps what was already done.
• Options in the prompt can be clicked or typed by their letter, for example "Close" by typing C. The value in angle brackets is the default; Enter accepts it.
• Right-click opens the shortcut menu: during a command the first item is "Enter", otherwise "Repeat last command".
• Aliases and full names both work, for example L or LINE. There is an autocomplete list. Typing a command that is not done yet shows its name and "coming soon".
• Prefixes: _ English name, - command-line version (no dialog, for example -OSNAP), ' transparent command (runs inside another command, for example 'Z, 'P).
• One command is one undo step. Ctrl+Z during a command rolls the whole command back.
• Full-width punctuation and digits typed through an IME are treated as half-width.

Coordinate input

• x,y absolute. @dx,dy relative. d<a polar. @d<a relative polar.
• A single number is direct distance entry: that distance along the cursor direction.
• With dynamic input (F12) on, x,y after the first point is relative; prefix # for absolute. Tab locks the length and jumps to the angle field.
• At a point prompt you can type a snap override: END, MID, CEN, INT, PER, TAN, NEA and so on, or NON to turn snapping off. Modifiers: FRO (from), M2P (mid between two points), TT (temporary tracking point). Each applies to the next point only.

Commands and aliases

The alias table is a subset of acad.pgp. Commands marked "Available" work now; "Coming soon" commands show a message when typed and are never mistaken for another command. The table below is generated from the source code.

Object snaps and drafting aids

The status bar toggles and function keys match AutoCAD. In full screen the browser hands F11 and F12 to the drafting page; otherwise use Shift+F11, Shift+F12 or the status bar buttons. The table below is generated from the source code.

• There are 14 object snaps: endpoint, midpoint, center, geometric center, node, quadrant, intersection, extension, insertion, perpendicular, tangent, nearest, apparent intersection, parallel. Endpoint, center, intersection and extension are on by default.
• Right-click OSNAP in the status bar to open Drafting Settings (DSETTINGS, DS) and tick the kinds. Shift+right-click on the canvas opens the snap override menu.
• Polar tracking increments are set in Drafting Settings: 90, 45, 30, 22.5, 18, 15, 10, 5 degrees. Ortho and polar are mutually exclusive.
• Object snap tracking: pause over an endpoint or edge to wake the tracking line.
• These settings are stored on this computer, not in the drawing, as in AutoCAD.
• Isometric drafting: ISODRAFT chooses isoplane Left (L), Top (T), Right (R) or Orthographic (O). In isometric drafting Ortho (F8) follows the two axes of the current isoplane: Left = 90° and 150°, Top = 30° and 150°, Right = 30° and 90°. Ctrl+E or ISOPLANE cycles Left → Top → Right and echoes "<Isoplane Top>". The grid is not drawn isometrically.

Selection

• With no command active, click to select. Left→right window (fully contained), right→left crossing (anything touched). Press and drag for a lasso.
• Shift+click removes. Ctrl+A selects all. Esc clears. Objects highlight on hover.
• Grips appear on the selection. Drag a grip to stretch or move; multi-function grips offer more options.
• Shift+Ctrl+A toggles group selection (PICKSTYLE, see Groups); while it is on, Ctrl-click selects just that one object of a group. Ctrl+W selection cycling, which the browser only passes through in full-screen drafting.
• Pickfirst: select first, then type M; the command uses that selection.
• Delete erases the selection (same as ERASE).

Modify

The flow is AutoCAD's: select objects, Enter to end selection, then follow the prompts to click a base point or second point, or type a distance, angle or scale factor. The prompt text follows AutoCAD too. A command is one undo step; "Undo (U)" inside a prompt only goes back one action within that command. Objects to be modified are refused when they are on a locked layer ("Object is on a locked layer."), but the cutting edges and boundaries used by trim and extend may be on locked layers.

• Basic: MOVE, COPY, ROTATE, SCALE, MIRROR, ERASE, OOPS, STRETCH, MATCHPROP, DRAWORDER, PROPERTIES.
• MATCHPROP (MA): pick the source object, then one or more destination objects. It copies the basic properties (color, layer, linetype, linetype scale, lineweight, transparency) and the special ones: dimension style; text style, height and width factor of text; hatch pattern, scale, angle, spacing and so on; polyline width; table style. External references, images and tables can be destinations too. Settings (S) opens the Property Settings dialog to choose what is copied (kept in your user settings, not in the drawing). The whole command is one undo step.
• Trim TRIM (TR), Extend EXTEND (EX): the default is Quick mode: every visible object is a cutting edge or boundary, and you click the segment to trim. Click an empty spot and then a second point for a fence; hold and drag for a freehand fence. Shift-click switches between trim and extend. "Mode (O)" switches to Standard: first select the cutting edges or boundaries (Enter = all), then click objects; "Edge (E)" sets whether to extend to implied intersections. Other options are Crossing (C), Fence (F), Remove (R) and Undo (U). In Quick mode an object with no intersections at all is deleted entirely; a closed object needs two intersections to be trimmed. Circles, closed objects and construction lines cannot be extended.
• Offset OFFSET (O): type a distance (or click two points to measure it), or use "Through (T)" and click a through point; then select the object and click the side to offset to. "Multiple (M)" offsets repeatedly from the newest copy, "Erase (E)" deletes the source object, and "Layer (L)" puts the result on the current layer or the source layer. There is a live preview. An ellipse or spline offsets to a spline.
• Fillet FILLET (F): the default radius is 0. Options are Radius (R), Trim (T), Multiple (M) and Polyline (P) (every corner of a whole polyline). Hold Shift while selecting the second object to join with a sharp corner. Between two parallel lines it makes a semicircle. When both objects are on the same layer the new arc goes on that layer, otherwise on the current layer (the same rule applies to color, linetype and lineweight, and to chamfers).
• Chamfer CHAMFER (CHA): Distance (D) (two distances), Angle (A) (length and angle), Method (E) switches between the two, plus Trim, Multiple and Polyline. It works on straight segments only; parallel lines give a message.
• Break BREAK (BR), Break at Point BREAKATPOINT: with BREAK the point where you select the object is the first break point; click the second point and the part between is removed; use "First point (F)" to specify the first point separately. A circle is broken counterclockwise from the first point to the second. BREAKATPOINT splits an object in two at one point with no gap (a circle cannot be split this way).
• Join JOIN (J): lines, arcs and open polylines that meet end to end become one polyline (a closed loop if the ends meet); collinear lines become one line; arcs on the same circle become one arc or a full circle. Circles, closed polylines and text are skipped, and the command line says how many were skipped. When the only source object is an arc (or elliptical arc), "cLose (L)" turns it into a full circle (or ellipse).
• Lengthen LENGTHEN (LEN): click an object to see its current length. Options are DElta (DE), Percent (P), Total (T) and DYnamic (DY). It works on lines, arcs, open polylines, elliptical arcs and open splines (splines can only be shortened).
• Align ALIGN (AL): select objects, then click 1 to 3 pairs of "source point, destination point". One pair = move; two pairs = move and rotate, and it asks whether to scale objects based on the alignment points; with three pairs, a third point on the other side mirrors. There is a live preview.
• Explode EXPLODE (X): a polyline explodes into lines and arcs (width is lost); a block reference explodes into its contents (attribute values are lost, and attribute definitions become text showing the tag); a hatch explodes into pattern lines (a solid hatch does not explode); a dimension explodes into lines, arcs and text; multiline text becomes single-line text; a group is dissolved. Blocks that do not allow exploding, and MINSERT, cannot be exploded.
• Edit Polyline PEDIT (PE): if you pick a line or arc it asks whether to turn it into a polyline. Options are Close or Open, Join, Width (one width for the whole polyline), Decurve, Reverse and Undo. "Multiple (M)" handles several at once, and for Join you can set a fuzz distance and join type (Extend, Add, Both). If you pick a spline it also asks for a precision.
• PEDIT curves and vertices: "Fit (F)" joins each pair of vertices with tangent arcs, so the curve passes through every vertex; "Spline (S)" makes a smooth curve that uses the vertices as control points; "Decurve (D)" restores the original vertices if the polyline was not changed after Fit or Spline, and otherwise straightens the arcs. If there are too few vertices, a message says so. "Edit vertex (E)" marks the current vertex with an X; its options are Next (N), Previous (P), Break (B) (mark two vertices, then "Go (G)" removes the segments between them and splits the polyline), Insert (I), Move (M), Regen (R), Straighten (S), Width (W) (start and end width of the next segment) and eXit (X). Ltype gen is Coming soon.
• System variables: FILLETRAD (fillet radius), CHAMFERA, CHAMFERB (chamfer distances) and TRIMEXTENDMODE (0 = Standard, 1 = Quick) can be typed directly. They are saved in the drawing and undone together with the command. The trim, offset, chamfer and hatch settings are saved in the drawing the same way.

Arrays

Associative arrays work as in AutoCAD: the array is one object, and you can still change its columns, rows, spacing or item count later. The source objects are replaced by the array, and the new array goes on the current layer. Moving, rotating, scaling or mirroring the array moves it as a whole. Objects on locked layers are not picked as source objects.

• ARRAY (AR): select objects, then choose the type: Rectangular (R), PAth (PA) or POlar (PO); the default is Rectangular. ARRAYRECT, ARRAYPOLAR and ARRAYPATH go straight to that type. The Array button in "Home › Modify" makes a rectangular array; Polar Array, Path Array and Edit Array are in the Modify panel slideout.
• Rectangular: 4 columns × 3 rows by default, spaced at 1.5 times the source width (height). Options are ASsociative (AS), Base point (B), COUnt (COU), Spacing (S) (or "Unit cell (U)": click two corners to set both spacings), COLumns (COL), Rows (R) and Levels (L). For columns and rows, "Total (T)" lets you enter the distance between the first and last column (row) instead.
• Polar: click the center point first. By default there are 6 items filling 360°, and items are rotated. Options are Items (I), Angle between (A), Fill angle (F) (positive = counterclockwise, negative = clockwise; between -360 and 360 and not 0), ROWs (ROW) (rings going outward) and ROTate items (ROT). When items are not rotated they are only moved.
• Path: select a path curve (it cannot be one of the source objects; the path is copied into the array, so later changes to the original curve do not carry over). Method (M) is Divide (D) (items spread evenly; on an open path there is one at each end) or Measure (M) (by spacing; by default as many as fit on the path). There are also Items (I), Rows (R) and Align items (A) (items turn with the path tangent).
• Press Enter or choose eXit (X) when you are done with the options. The whole command is one undo step. Answering No to ASsociative (AS) explodes the array into separate objects when the command ends.
• Ribbon "Array Creation" tab: appears while the command runs, with Columns (or Items), Rows, Levels, Properties and Close panels. A number field is applied when you press Enter or leave the field; Esc restores it. With no command running, selecting one associative array turns the same tab into the "Array" tab, and each change is one undo step.
• ARRAYEDIT: select an associative array and change its parameters (the options are the same as when creating it). If other copies share the same array, only the one you selected changes.
• EXPLODE breaks an array into separate objects that are no longer associative. In DXF an array is written as an INSERT of an anonymous block with its parameters in XDATA; AutoCAD sees an ordinary block reference, and Synhedra restores the array when the file is read back.
• Limits: items × source objects cannot exceed 20,000 (at most 100,000 items); a message tells you when the array is too large. A 2D drawing has a single level, so levels do not change the result.
• Grip editing: select an associative array to get array grips besides the insertion point (which moves the whole array). Rectangular: 2nd column = column spacing, last column = column count, 2nd row = row spacing, last row = row count, far corner = both counts. Polar: 2nd item = angle between items, last item = item count (not when filling 360°), row spacing and count. Path: Measure method 2nd item = spacing, last item = count; Divide method 2nd item = count; row spacing and count. Click a grip (it turns red), move to preview live, click again to apply; each edit is one undo step.
• Key point, Tangent direction, expressions, and Edit Source, Replace Item and Reset Array are Coming soon.

Regions and Boolean operations

• REGION (REG): turns closed polylines, circles, ellipses, splines, vector paths, and lines and arcs that meet end to end in a closed loop, into regions. The source objects are deleted; a region takes the layer and properties of its first source object. Loops with zero area or that intersect themselves are rejected, and the command line says how many regions were created and how many objects were rejected. Objects on locked layers are skipped.
• Union UNION (UNI), Intersect INTERSECT (IN): select at least 2 regions; selected objects that are not regions are filtered out. Subtract SUBTRACT (SU): select the regions to subtract from, press Enter, then select the regions to subtract. The result keeps the properties of the first region; if the result is empty the region is deleted and a message says so.
• MASSPROP: lists a region's area, perimeter, bounding box, centroid, moments of inertia, product of inertia, radii of gyration, and principal moments and directions about the centroid. At the end it asks whether to write the analysis to a file; Yes downloads massprop.mpr.
• Ribbon: Region is in the Draw panel slideout, Union, Subtract and Intersect in the Modify panel slideout, and Mass Properties in the Utilities panel slideout. LIST shows a region's area and perimeter. In DXF a region is written as ordinary path geometry with a marker in XDATA, and Synhedra restores it as a region when the file is read back; in AutoCAD it is not a REGION object.

Divide, measure and point style

• Divide DIVIDE (DIV): select a line, arc, circle, polyline, ellipse or spline and enter the number of segments (2 to 32767). An open object gets one point fewer than the number of segments (none at the ends); a closed object starts at its start point.
• Measure MEASURE (ME): select an object and enter the segment length (or click two points to measure it). An open object is measured from the end nearer to where you picked it; a closed object starts at its start point. If there would be more than 32767 items, use a larger length.
• Both have a "Block (B)" option to place a block instead: enter the block name, then choose whether to align the block with the object (aligned = each block is rotated to the tangent). The points or blocks go on the current layer and the object itself is not changed, so objects on locked layers work too. The new points become the Previous selection set, so you can pick them again with P. The whole command is one undo step.
• Point style DDPTYPE (PTYPE): a dialog with 20 point styles and the point size (relative to the screen or in absolute units). PDMODE and PDSIZE take a value directly. The settings are saved in the drawing and can be undone. Ribbon: Divide and Measure are in the Draw panel slideout, Point Style in the Utilities panel slideout.

Groups

• GROUP (G): select objects first and type G to create an unnamed group straight away (named *A1, *A2, …). Without a selection you can use "Name (N)" and "Description (D)" to name and describe the group, then select the objects. A group name cannot start with *, cannot contain < > / \ " : ; ? * | , = `, and must be unique.
• -GROUP (-G): the command-line version, with the options ? (list groups), Add (A), Remove (R), Explode (E), REName (REN), Selectable (S) and Create (C). Removing every member deletes the group. Order (O) is Coming soon.
• UNGROUP dissolves groups (vector workspace groups too); GROUPEDIT has Add objects (A), Remove objects (R) and REName (REN). Members on locked layers are not changed.
• Group selection PICKSTYLE: when it is on, clicking a member of a group selects the whole group (unless the group is not selectable), and Ctrl-click selects only that object. Toggle it with Shift+Ctrl+A or the Group Selection button in the Modify panel. It is a personal setting, not saved in the drawing. Group, Ungroup and Group Edit are in the Modify panel slideout. DXF keeps a named group's description and selectable setting.

Hatch

• HATCH (H, BH, BHATCH): click inside a closed area (Pick Points), or use "Select objects (S)" to choose the boundary. Once a first boundary is found it shows live; you can click several areas, and they combine into one hatch. Enter or "Close" finishes, Esc cancels. If the boundary has a gap, the gap is marked with a red circle; close it or increase the gap tolerance. Only visible objects are used as boundaries.
• "Hatch Creation" tab: while the command runs, the ribbon switches to this contextual tab: Boundaries (Pick Points, Select), Pattern (thumbnail gallery), Properties (type: Solid, Pattern, User defined; color, background color, transparency, angle, scale; User defined uses spacing and Double), Origin, Options (Associative, island detection: Normal, Outer, Ignore; gap tolerance) and Close. With no command running, selecting a single hatch turns the same tab into the "Hatch Editor", and each change is one undo step.
• CN-* patterns: 38 self-drawn CN-* patterns plus SOLID are built in, covering section hatching, metal, glass, wood, concrete, soil, stone, brick, tile, flooring, roofing and insulation. Section-type patterns are defined in paper millimeters, so set the scale to the plot scale (100 for 1:100); brick, tile, flooring and roofing patterns are defined in real sizes, so use a scale of 1. When you type a pattern name at the command line, common names such as ANSI31 and AR-CONC map to a similar CN-* pattern; type ? to list them all.
• Associative: on by default. When you move or stretch a boundary object, or edit it with grips, the hatch updates with it, in the same undo step. Moving only the hatch itself, deleting a boundary object or pulling the boundary apart removes the association. Associative hatches on locked layers are not recalculated.
• HATCHEDIT (HE): edit an existing hatch: remove association, island style, properties, origin, re-associate. -HATCH (-H) and -HATCHEDIT are the command-line versions and do not open the contextual tab.
• BOUNDARY (BO): click inside an area to create a closed boundary polyline (including islands). -BOUNDARY is the same.
• Match Properties: the Options panel of the "Hatch Creation" and "Hatch Editor" tabs has Match Properties (type MA at the command line). Click another hatch to take its pattern, type, scale, angle, color, background color, transparency, island detection and layer; the origin, associativity and gap tolerance keep their current settings.
• Blocks as boundaries: block references also act as boundaries (their contents are expanded, up to 8 nesting levels). An associative hatch is recomputed when the block is moved, rotated or scaled. Hatches, dimensions, groups, points, hidden objects and objects on layers that are off or frozen are not used as boundaries.
• Copying associative hatches: when you copy, array, mirror while keeping the source, grip-copy or paste, and the boundary objects are copied too, the copy is bound to the copied boundaries; otherwise the copy is not associative. Seed points and the pattern origin follow move, rotate, scale and mirror.
• Boundaries (editor): in the "Hatch Editor", "Pick Points" adds boundaries and "Remove" removes boundaries from the hatch (an island, for example; the outermost boundary cannot be removed). The slideout has "Recreate" (creates a polyline or region boundary along the hatch, optionally re-associating the hatch with it) and "Display Boundary Objects" (associative hatches only).
• Advanced options: the Options panel has "Annotative"; its slideout has Gap Tolerance (0–5000), Island Detection and Draw Order (do not assign, send to back, bring to front, send behind boundary [default], bring in front of boundary; HPDRAWORDER), plus "Create Separate Hatches" while creating (several areas picked at once each become their own hatch; HPSEPARATE) or a "Separate Hatches" button while editing. The Properties panel slideout has "Hatch Layer Override" (HPLAYER; "Use Current" while creating).
• HATCHEDIT command-line options: Disassociate (DI), Style (S), Properties (P), DRaw order (DR), ADd boundaries (AD), Remove boundaries (R), recreate Boundary (B), ASsociative (AS), separate Hatches (H), Origin (O), ANnotative (AN), hatch COlor (CO), LAyer (LA), Transparency (T). -HATCH also gains DRaw order, ANnotative and Transparency.
• Gradient fill is not available yet.

Blocks and attributes

• Ribbon "Insert" tab: the Block panel has Insert, Create Block and Write Block; the Block Definition panel has Define Attributes and Edit Attributes. The Block panel also has the Block Editor.
• BLOCK (B): opens the Block Definition dialog: Name, Base point (Pick point), Objects (Select objects, with Retain, Convert to block or Delete), Scale uniformly, Allow exploding, Description. If the name already exists it asks whether to redefine. -BLOCK (-B) is the command-line version, and the source objects are deleted.
• INSERT (I): opens the Insert dialog: pick a block from the current drawing (with preview); the insertion point, scale and rotation can be specified on screen or typed, and you can check "Explode". A block with attributes asks for the attribute values in order. When the drawing has no blocks yet, it asks you to create one with BLOCK first. On a layout tab, INSERT inserts a title block (see Layouts, viewports and plotting). -INSERT (-I) is the command-line version.
• WBLOCK (W): writes a block, the whole drawing or the selected objects to a new DXF file (AutoCAD 2000 format) and downloads it. The base point becomes the origin of the new drawing.
• ATTDEF (ATT): opens the Attribute Definition dialog: Mode (Invisible, Constant, Verify, Preset, Lock position), Tag (no spaces, stored in uppercase), Prompt, Default, Insertion point, text height and rotation. Select the attribute definitions together with the objects when you create the block. -ATTDEF is the command-line version.
• ATTEDIT (ATE): pick a block reference and change its attribute values in a dialog. Constant attributes cannot be changed.
• Block Editor BEDIT (BE): opens the "Edit Block Definition" dialog; pick a block in the list (with preview and description) and click OK. If you select a block reference first and then type BE, that block is already picked in the list. You can also use the Block Editor button on the "Insert" tab, or double-click an ordinary block reference in the CAD workspace (not blocks with attribute values, arrays or anonymous blocks). -BEDIT is the command-line version (? lists the blocks you can edit). The Block Editor opens in model space only; anonymous blocks, dynamic blocks and vector workspace symbols cannot be edited in it.
• While editing: the block's contents are placed in model space with the base point at 0,0; all other objects are hidden for the time being and cannot be selected or snapped to. The ribbon switches to the "Block Editor" contextual tab: Save Block, Define Attribute, Base Point, common draw and modify tools, and Close Block Editor. BASE sets a new base point (only inside the Block Editor). While the editor is open you cannot plot, publish, import or export, switch layouts or workspaces, or open the AI panel, and WBLOCK, VIEWBASE and VIEWUPDATE are refused; close the Block Editor first.
• BSAVE: redefines the block from the current contents, and all references update. The description, Allow exploding, Scale uniformly and other settings are kept. The contents cannot be empty.
• BCLOSE: closes the Block Editor. If there are unsaved changes it asks whether to save the changes, discard them and close, or cancel (stay in the editor). After closing, every step from opening to closing becomes one undo step; discarded edits cannot be redone. The temporary contents are never autosaved or exported. If you switch to another drawing, the Block Editor is closed as if you discarded the changes.

External references

Place another DXF drawing in this one as a reference, following AutoCAD's external references. The ribbon "Insert › Reference" panel has "Attach" (XATTACH, XA), "Clip" (XCLIP, XC) and "External References" (XREF, XR).

• The source is a DXF file or a 2D design in this project: the browser has no access to file paths on your computer, so attaching reads the DXF content and stores it in this drawing, keeping only the file name as the path. When the source file changes, press "Reload…" in the External References palette and choose the file again. Save DWG files as DXF in AutoCAD first; .lsm files cannot be attached. Nested xrefs (references inside the reference) are not supported. Blocks, attributes and other programs' dimensions inside the reference are flattened into plain geometry; objects other than lines, polylines, arcs, circles, points, ellipses, splines, construction lines, rays, text, multiline text, hatches and vector paths are skipped and listed. Files are limited to 192 MB and are parsed in the background without freezing the screen. At most 100,000 objects are read, and they are scaled by INSUNITS to this drawing's units. When the content stored in the drawing exceeds about 4 MB you are warned: it counts toward every save, the undo history and the cloud design size limit, so consider simplifying it in AutoCAD first.
• XATTACH: choose a DXF file, then the "Attach External Reference" dialog opens: Name (default: the file name), Reference Type (Attachment or Overlay), Insertion point, Scale (with Uniform Scale) and Rotation (all three can be specified on-screen), plus the read-only Block Unit and object count. The name cannot match a block in the drawing or an xref from another file; the same file under the same name = another reference. Scale, rotation and type carry over to the next attach. Because nesting is not supported, Attachment and Overlay currently differ only in the type shown.
• A 2D design in the same project as the source: below the file dialog, "Or: a 2D design in this project" lists your other 2D designs in the same project (or equally unfiled), plus 2D designs shared with you that sit in this project. Pick one and press "Use this design" ("☁ From cloud" first fetches the latest version from the cloud, one connection), then the same "Attach External Reference" dialog opens, with the design name as the default name. The drawing stores the design ID (for a shared design it stores an irreversible ID, not the share link), and the palette's path column shows "Design: name". After the source design changes, the palette status gains "(source changed)" when the palette is opened; select that row and press "Reload" to re-read that design directly (no file picking), or "☁ Reload from cloud" to pull the latest cloud version first. External references inside the source drawing are not brought in (one level only, so two drawings referencing each other never loop); views the source imported from 3D become ordinary geometry. A drawing cannot reference itself, and a shared drawing cannot read your other designs.
• External References palette XREF: it does not block the view and can be opened in the middle of another command. Columns are Reference Name, Status (Loaded, Unloaded, Not Found), Size, Type, Date, References and Saved Path. The toolbar has "Attach DXF…"; with a row selected (or on right-click) there are "Reload…" (choose the file again), "Unload", "Detach" (deletes all references and the definition) and "Bind…". "Open" is not available. Double-click the "Type" column to switch between Attach and Overlay. For an unloaded xref whose content is still stored in the drawing, the menu also has "Reload (stored copy)", which does not ask for the file again.
• Binding: "Bind…" in the palette offers "Bind" (layers are renamed to "name$0$layer", or $1$, $2$… if taken) or "Insert" (layers keep their names, and layers with the same name merge into this drawing's layers). The result is an ordinary block with the same name, which can be exploded. Only loaded xrefs can be bound.
• XBIND (XB), -XBIND: only LAyer (LA) is supported. Enter the dependent layer names ("xrefname|layer"; several with commas, and * and ? work); those layers are bound into this drawing and renamed to "name$0$layer" (or $1$, $2$… if taken), while the xref stays an xref. 0 and Defpoints cannot be bound, and only loaded xrefs can be bound. Blocks, dimension styles and text styles are already flattened when the xref is attached, and linetypes are already merged then, so choosing those types only shows the reason.
• -XREF: the command-line version, with the options ? (list), Bind (B) (always the "Bind" method), Detach (D), Path (P), Unload (U), Reload (R), Overlay (O) and Attach (A). Several names can be given with commas, and the wildcards * and ? work. Reload here only re-shows an unloaded xref whose content is still stored in the drawing; when the drawing holds no content, the palette opens with that row selected, so press "Reload…" and choose the file again. To read a new file, use the palette. An xref with a reference on a locked layer cannot be detached.
• XCLIP: clips an xref or block reference (not a MINSERT array). The options are ON, OFF, Delete (D), generate Polyline (P) and New boundary (N). A new boundary can be Rectangular (R) (default), Polygonal (P) (at least 3 points) or Select polyline (S) (closed, straight segments only); choose "Invert clip (I)" first to switch to inside mode: objects inside the boundary are hidden and those outside are shown. Lines and curves are cut at the boundary, hatches are intersected with it (subtracted when inverted), text is kept when its layout box touches the boundary (when inverted: unless it lies completely inside), and points and nested blocks are kept or hidden by their insertion point; clipped-away parts cannot be selected or snapped to. XCLIPFRAME sets the clipping frame: 0 = not shown, 1 = shown and plotted, 2 = shown but not plotted (default 2, as in AutoCAD). The setting is saved in the drawing.
• Layers: xref layers are listed in the Layer Properties Manager as "xrefname|layer", and the filter can be set to "Xref". These layers cannot be renamed, deleted or made current, but On, Freeze, Lock, Plot, Color, Linetype, Lineweight and Transparency can be changed; the changes are stored on the xref and kept on reload. ByLayer properties take the xref layer's settings. Freezing the layer a reference sits on hides the whole xref; unloaded or not-found xrefs draw nothing. The layer drop-down on the ribbon "Home" tab lists only this drawing's own layers.
• PURGE never purges xrefs (detach them with XREF), and the INSERT and BEDIT lists do not show xrefs. AI cannot create or change xrefs. Each action is one undo step.

Import views from 3D

Projects plans, elevations, sections and axonometrics of a 3D model into this drawing as 2D linework you can keep drafting on, and updates them when the 3D model changes. The "Insert › 3D Views" ribbon panel has "From 3D" (VIEWBASE; you can also type VR3, VIEWREF or 3DIMPORT) and "Update from 3D" (VIEWUPDATE). It is available in the CAD workspace only. The projection is the same computation as Make2D.

• Source: a 3D design in the same project (folder), or a .json file downloaded from the Make2D dialog in 3D (up to 60 MB). If the project has no 3D design, only the file option is available.
• Views: a plan for each level (the cut height can be changed and defaults to the level's view range; with no levels there is one plan cut 1.2 m above the bottom of the model), North, South, East and West elevations, four isometric axonometrics, and the section planes in the 3D design (sloped section planes are not supported). By default the lowest level's plan and the active section are ticked. At most 16 views at a time.
• Options: hidden lines (CN-HIDDEN linetype, off by default), join connected lines into polylines, one block per view (unticked = loose lines), view label (name and scale), scale (1:100 by default, or custom), and set the annotation and linetype scale to it. Model space is always 1:1 real size; the scale only affects the label height, annotation scale, linetype scale and layout viewports.
• Place in: model space, or model space plus a new layout for each view (a locked viewport on the smallest landscape A4–A0 sheet that fits). Insertion point: specify on screen, align to the 3D origin (plans sit at world coordinates), or type X and Y. Several views are laid out in a row from left to right.
• Layers: by category (A-WALL-CUT, A-DOOR, A-GLAZ, A-PROJ, A-HIDDEN and so on) or by Make2D layer (you can map each one to a CAD layer). By default layer names get a "view name$" prefix so you can switch views on and off separately. Block references and view labels go on layer A-VIEW.
• An import is one undo step. Each view remembers its source and settings (stored in XDATA, so they survive a DXF round trip). One import creates at most 150,000 objects; if it would be more, pick fewer views or turn hidden lines off.
• VIEWUPDATE: choose All (A), Select (S) or Force (F). Views whose 3D model has not changed are reported as up to date; choose Force to recompute them anyway. If the views contain your edits (modified, erased or locked objects), they are listed first and you are asked whether to update anyway (the default is No); objects you added inside the blocks are kept. Views whose source is not in this drawing's project, was deleted or cannot be read are not updated; views imported from a .json file must be imported again. Press Esc to cancel while it computes. An update is one undo step.
• Out-of-date views and auto-update: each imported view remembers a fingerprint of the 3D model it was made from. After the 3D design changes (only changes that affect the projection count, such as walls, levels or parameters; moving nodes in the node graph, scenes or cameras do not), a non-blocking bar appears at the top right: "N views from 3D are out of date". The Model tab and any layout tab whose viewports show such a view get an orange dot. Update runs VIEWUPDATE on the out-of-date views only (it still asks first if they contain your edits). Tick Auto-update, or set VIEWUPDATEAUTO to 1 (saved with the drawing; default 0), and out-of-date views without your edits or locks are updated in a background worker after the 3D model changes (applied when no command is running, as one undo step); edited views stay in the bar for you to confirm. 2D edits reach the 3D model only through VIEWWRITEBACK.
• Fields (FIELD, UPDATEFIELD): FIELD inserts text that updates itself: project information (project name, number, address, issue date… of a 3D design in the same project), a global parameter (lengths, areas and angles with units), the sheet number or name (a layout named “A-101 - Ground Floor Plan” gives number A-101 and name Ground Floor Plan), or the date (2026-10-20, 2026/10/20 or ROC 115.10.20). Title blocks placed with Insert Title Block also bind the project name, drawn, checked, date, number and title attributes you did not change. Fields are re-read from the 3D design when the drawing opens, after VIEWUPDATE, and with UPDATEFIELD (selected or all); only changed values are written (one undo step), and global parameters driven by the graph are evaluated in a background worker. If the 3D design cannot be read (on another device, a shared 2D drawing, or deleted), the last text is kept and a bar at the top right says “N fields cannot read their 3D source (may be out of date)”; Update fields retries. Fields are stored in the object's XDATA (CN_FIELD) and survive DXF round trips; the resolved text is stored in the object too, so it shows offline and in the viewer. Up to 5,000 fields per drawing, 500 characters each.
• Writing 2D edits back to 3D (VIEWWRITEBACK): lines, door and window symbols and tags imported from 3D remember which 3D element they come from. Edits that can be written back: a rigid move of wall lines (the 3D wall moves, joined walls stretch and its doors and windows go with it), of door and window symbols (they slide along the host wall) and of family component lines; changing a door or window tag (= its mark); changing a room name or number (level plans carry room tags); overriding the text of a linear or aligned dimension associated with two 3D objects with a number (for example 5000) = moving the element of the second reference so that the distance becomes that value (like clicking a dimension value in the 3D plan; DIMLFAC is applied). After such an edit a blue bar appears: "N elements imported from 3D were edited in 2D and can be written back". Click "Write back to 3D" or type VIEWWRITEBACK (Apply (A) / Detach (D) / Cancel (C)). The write-back checks the source first (it must be your own 3D design in this project, not open in another tab, and the view must not be out of date), changes the 3D design with the same write-back rules as the 3D editor, then updates the affected views from the new 3D. In 2D it is one undo step, and Ctrl+Z also restores the 3D (if the 3D changed again afterwards, only the 2D is undone and you are told). Values driven by a global parameter (for example a door position bound to one) are not written back; the message names the global parameter, so change it in 3D. Edits that cannot be written back (stretching, rotating, trimming, only moving a tag, a segment of a polyline wall, a move whose original position is unknown because the drawing was reopened) are listed with the reason; choose Detach to keep them as plain 2D objects, and VIEWUPDATE no longer generates the same object (up to 100 per view). If a view still has edits that cannot be written back, you are asked first whether to detach them or let the update replace them. Limits: axonometric views and sloped sections cannot be written back; elevations and sections support horizontal moves only; dimension values work only for associative dimensions in loose (non-block) views; not supported while real-time collaboration is on or the design is shared for editing (changes must follow the collaboration borrowing rules, so change it in 3D); up to 500 edits at a time.
• Convert a 2D plan to BIM (TOBIM): ribbon “Insert › 3D Views › To BIM” or type TOBIM, select the plan objects (preselected, or ALL), and the 2D plan is recognized as walls, doors/windows and rooms of a 3D design in the same project (the “2D → 3D” half of two-way work inside the modeler, not a general DWG-to-BIM converter). Walls = pairs of parallel lines (lines and straight polyline segments) 8–60 cm apart on layers whose name contains wall (for example A-WALL; or set the layers first with the Wall layers option, * allowed); the gap is the wall thickness, the centerline runs along the overlap of the two lines, and L/T corners join at the neighbouring centerline. Doors/windows = blocks whose name contains door, window, 門 or 窗 (placed on the nearest wall, width = the block’s width along the wall), or a 0.5–3 m gap in one wall (a door by default; a door/window block inside the gap decides the kind; the walls on both sides become one wall when created). Rooms = text inside a closed polyline, or text that reaches recognized walls in all four directions (name = the text, one per area). The results are previewed over the drawing (cyan = walls, yellow = doors/windows, green = rooms, grey = skipped; temporary, nothing stays in the drawing); in the panel on the right tick items, switch door/window, choose the 3D design, level and wall type (Automatic = closest thickness), then press Create to write them into the 3D design. On the 2D side it is one undo step (the source objects remember they were converted, and running TOBIM again skips them by default); Ctrl+Z also restores the 3D design (if the 3D changed afterwards only the 2D is undone, with a message). The same safeguards as VIEWWRITEBACK: this 2D drawing must be a design in your workspace (shared drawings are not converted), the target must be your own 3D design in the same project, not open in another tab and not in real-time collaboration, and the 3D design must not have changed in the meantime; external references are only read from your own 2D designs in the same project (file sources are skipped); objects imported from 3D are not converted. The 2D origin is the 3D origin. Limits: 2,000 walls, 2,000 doors/windows and 500 rooms at once; 20,000 segments on wall layers (select one area at a time if there are more). Base point: in the "Base point (2D → 3D)" box of the panel, set the 2D base point (type the coordinates, or press "Pick" and click in the drawing; command option BA) and the 3D point it lands on (m); by default the 2D origin is the 3D origin. For storeys drawn side by side in one drawing, give each storey its own base point (for example the same outer wall corner) instead of stacking them with MOVE first. Selecting with ALL then asks for an area: draw a window to take only the objects inside it that are not converted yet (to any 3D design); Enter takes every unconverted object in the drawing. Unconverted text of another storey (such as a storey label) is no longer carried along.
• Office standards: layer rules, phases, component blocks, annotations, auto-dimensions: the "Office standards" section of the wizard has five options, and they affect only newly imported views; views imported earlier stay as they were on update (layers don't suddenly change and no dimensions appear).
  • Apply layer mapping rules (on by default, only for "By category"): type LAYERRULES (or DRAFTRULES) to open the "Layer and dimension rules" dialog, and set "category, cut state, phase, tag, level → layer" together with color (ACI) and lineweight, for example putting the cut lines of walls on "Wall-Cut". Rules are matched in order and the first match wins; when nothing matches, the default A-WALL-CUT, A-DOOR… apply. The dialog has two sample sets (Taiwan office layers in Chinese, NCS layers by phase). The plan rules of the 3D design in the same project (project level; the 3D plan cells use the same set) are applied too, and this drawing's rules take priority; "Copy to this drawing" brings the whole set in for editing. Write categories as English codes (wall, door, window, floor, column, beam, roof, stair, component, grid, tag, keynote, room…); tags and levels accept the * wildcard. After changing the rules, regenerate already imported views with VIEWUPDATE and choose "Force".
  • Layers by phase: elements whose phase is "Existing" in 3D go to the -EXST layer (gray), those demolished in this phase go to -DEMO (dashed), and new construction gets no suffix; when a layer rule has a "phase" condition, the rule decides.
  • Doors, windows and family components as blocks: doors, windows and components of the same type and shape (including rotations by multiples of 90° and left/right mirroring: a mirrored insert has an X scale of −1) share one block definition (named 3D-type-code), with one insert per instance, and the attributes are the schedule columns (type, tag, level, width, height, sill and project parameters; invisible, so the drawing looks the same). Type ATTEXT (EATTEXT, DATAEXTRACTION) to extract the attributes to CSV: when the header is set to "Schedule columns", the column names match the 3D schedule CSV. Moving a whole block can be written back to 3D with VIEWWRITEBACK (doors and windows move along the host wall); attribute positions you have dragged are kept on update.
  • Tags and keynotes drawn as multileaders, plus structural grids: tags, keynote tags, multi-category tags and material tags on the plan become multiline text + leader + arrowhead (remembering the 3D element being tagged), and structural grids and grid bubbles go on S-GRID and S-GRID-TEXT (centerlines). Tags you have dragged stay where you put them on update (the leader follows). No grids are added when the view's Visibility/Graphics turns the "Grids" category off.
  • Auto-dimensions from dimension rules (on by default; generated only when dimension rules exist): the dialog's "Auto-dimensions" sets the kind (overall, continuous, grid spacing, width of each element, along the longest side), the categories, the distance from the object (mm on paper) and the layer (default A-ANNO-DIMS). Dimensions are generated and updated together with the view and are not written back to 3D. Up to 2,000 rule-based dimensions per view.
  • Layouts and plot styles: when importing into a new layout, the layout rules decide the paper size and the title block (the title block must be a block in this drawing); the 3D view template (Visibility/Graphics) applied to a plan view becomes viewport layer overrides: hidden categories are frozen in the viewport, and colors and lineweights follow the template. "View template → Plot style (STB)" in the dialog generates the template as a named plot style table; choose it when plotting to print according to the template.
  • Scale from the layout rule: when you choose a 3D design, if the layout rule (this drawing’s first, otherwise the 3D design’s plan rules) has a scale, the wizard preselects that scale and says so; you can still change it.
  • Level and material tags in sections/elevations; rules updated: the automatic dimensions of LAYERRULES have two more kinds: level lines and labels (one horizontal line per level in sections and elevations, with a level symbol and “1F +0.00” at the right end) and material tags (one leader plus the material name per element, from the BIM material: walls and floors = the core material of the type’s layers, others = the object material). They apply to sections and elevations only (not plans), go to layer A-ANNO-DIMS by default and are regenerated as a batch by VIEWUPDATE; distances are paper mm (for example 6 mm at 1:100 = 0.6 m). Each view remembers this drawing’s rules (layer mapping and dimension rules) when it was made; after you change the rules, the bar shows “Rules updated: N views were made with the old drawing rules” (even when the 3D did not change, so they are not out of date), and “Update with new rules” runs VIEWUPDATE on those views (one undo step). Views imported before this feature are not flagged.

Named views, viewports and display

• VIEW (V): the View Manager. It lists the current view and the model views; you can set one current, create a new one (current display or a defined window), rename and delete. OK applies everything at once (one undo step). The drop-down in View › Named Views restores a view directly. -VIEW (-V) is the command-line version with the AutoCAD options ?/Delete/Orthographic/Restore/Save/sEttings/Window. Named views exist in model space only; restoring zooms so the saved area fits, in the current viewport when viewports are tiled.
• VPORTS: tiled viewports in model space. The dialog offers the standard configurations (Single; Two: Vertical, Horizontal; Three: Right, Left, Above, Below, Vertical, Horizontal; Four: Equal) and named configurations. -VPORTS: Save/Restore/Delete/Join/SIngle/?/2/3/4. 2, 3 and 4 split the current viewport, SIngle returns to one viewport, Join merges the current viewport with an adjacent one. Click a viewport to make it current; each one pans and zooms on its own (not undoable). On a layout tab use MVIEW.
• UCSICON turns the UCS icon at the lower left on or off, and NAVBAR the navigation bar. Both are your own preferences, kept in this browser and not saved in the drawing. 2D drafting only has the World Coordinate System, so the other UCSICON options have no effect.
• VSCURRENT (VS): 2D drafting only has the 2D Wireframe visual style; choosing another shows a message.

Raster images

• IMAGEATTACH (IAT): choose a PNG or JPEG file, set the name, insertion point, scale and rotation in the Attach Image dialog (specify on screen or type values), then place it in the drawing. The scale is the image width in drawing units. One attach is one undo step.
• Limits: 30 MB per file, 16,384 pixels per side and 40 million pixels in total, checked before decoding. The drawing keeps a reduced snapshot (2048 pixels on the long side); you are warned when a snapshot is large, because the whole design, the undo history and saves carry it. Up to 1000 image definitions.
• IMAGE (IM): the image palette with name, status, size, references and path; attach, reload (choose the file again), unload and detach. -IMAGE (-IM): ?/Detach/Path/Reload/Unload/Attach.
• IMAGECLIP (ICL): ON/OFF/Delete/New boundary. A new boundary can be a selected polyline, a polygon or a rectangle, and the clip can be inverted.
• IMAGEADJUST (IAD): brightness, contrast and fade in a dialog (OK = one undo step); -IMAGEADJUST (-IAD) is the command-line version. TRANSPARENCY: shows the transparent pixels of a PNG as transparent (ON/OFF).
• IMAGEFRAME: 0 = frame hidden, 1 = frame shown and plotted (default), 2 = shown but not plotted.
• Images can be moved, rotated, scaled, mirrored, copied and arrayed; grips and snaps work as for blocks. Images cannot be exploded.
• DXF: export writes IMAGE entities with only the original file path and pixel size (like AutoCAD external images); the snapshot is not written to the DXF. Images imported from DXF have only their frame and path and show as "not found"; Reload in the image palette and choose the file to display them.

Inquiry

Inquiry commands do not change the drawing, and they also work on a read-only drawing. Results appear in the command line history (F2). DIST, ID and AREA can run transparently, for example 'DIST.

• DIST (DI): the distance between two points, the angle in the XY plane and the X and Y deltas. "Multiple points (M)" accumulates a total length.
• ID: click once to show the coordinates of that point.
• AREA (AA): click the corners of a polygon in turn, or use "Object (O)" to select a circle, polyline, ellipse, spline, hatch and so on; it shows the area and perimeter. "Add area (A)" and "Subtract area (S)" keep a running total. While clicking corners you can switch to "Arc (A)" for tangent arc segments ("Second pt (S)" for a three-point arc, "CLose (CL)" to close back to the first point with an arc), or use "Length (L)" to enter a length in the direction of the last segment.
• LIST (LI): lists the type, layer, space and handle of the selected objects, plus data for each type (for example a hatch's pattern and area, a block's name and attributes).
• MEASUREGEOM (MEA): Distance (D), Radius (R) (an arc, circle or polyline arc segment; the diameter is shown too), Angle (A) and ARea (AR). Angle: pick an arc for its included angle; pick a circle and then a second point; pick two lines for the angle between them (which angle depends on where you click); or press Enter and click the vertex and two end points. A 2D drawing has no volume to measure; "Quick" measurement is Coming soon.
• Hyperlinks HYPERLINK (Ctrl+K): select objects, then enter the address and a description; the hyperlink is stored in the object's XDATA (PE_URL, as in AutoCAD, kept through DXF). -HYPERLINK Remove (R) removes it. HYPERLINKOPEN opens the selected object's hyperlink in a new tab; only http, https and mailto links are opened. Object hyperlinks only (no AutoCAD area hyperlinks).
• Revision cloud REVCLOUD: Rectangular (R, the default): pick two opposite corners; Polygonal (P): pick points one by one, Undo (U), Enter to finish (at least 3 points); Object (O): turns the selected polylines or circles into revision clouds (layer and properties kept). Arc length (A) sets the length of each arc, 0 = auto (8% of the outline’s diagonal). Arcs always bulge outwards; a cloud has at most 5,000 vertices (arcs are lengthened automatically when too dense). The result is an ordinary polyline with bulges, so it round-trips through DXF and opens in AutoCAD; the arc length and mode are remembered for this session.
• Revision tags and revision table REVINFO, REVTABLE: REVINFO selects revision clouds (or press Enter straight away), then asks for the revision number, date (today by default), description and drawn-by; the selected clouds get the revision tag, and it becomes the “current revision”, so clouds drawn later with REVCLOUD carry it automatically. The tag is stored in the cloud’s (polyline’s) XDATA (application CN_REV), survives DXF round trips and does not bother AutoCAD. REVTABLE merges the tagged clouds by revision number (case-insensitive, natural order); pick the upper-left corner and a text height to draw the revision table (lines + single-line text: Rev, Date, Description, By; text that is too long is cut with “…”). After revisions change, run REVTABLE again: it no longer asks for the location and updates the table in place, at the same position with the same text height; each run is one undo step. A table on a locked layer is not updated. Limits: 200 rows, revision number 16 characters, date 24, description 120, drawn-by 40; control and bidirectional characters are removed. Only clouds in model space are read; hyperlinks to issues (lsm://issue) are not provided. Also, REVCLOUD refuses (with a message) an outline or object with more than 20,000 points instead of silently cutting it into a different shape.

Layers

• Layer Properties Manager LAYER (LA): the "Layer Properties" button in the ribbon "Home › Layers" panel, or type LA. The palette does not block the view, can be dragged, and can be opened in the middle of another command. Columns are Status (✓ = current layer, double-click to make current), Name (double-click to rename), On, Freeze, Lock, Plot, Color (ACI 1–255 or true color), Linetype (loaded linetypes), Lineweight, Transparency (0–90) and Description. At the top are New, Delete, Set Current, filters (All, All Used Layers) and a search box. Each change is one undo step.
• Limits: layers 0 and Defpoints cannot be deleted or renamed; the current layer cannot be deleted or frozen; a layer with objects on it (including objects in layouts and block definitions) cannot be deleted; a frozen layer cannot be made current. A layer name cannot be empty, has at most 255 characters, cannot contain < > / \ " : ; ? * | = ` ,, and must be unique regardless of case.
• -LAYER (-LA): the command-line version. Options are ?, Make (M), Set (S), New (N), Rename (R), ON, OFF, Color (C), LType (L), LWeight (LW), Plot (P), Freeze (F), Thaw (T), LOck (LO), Unlock (U). Layer names can be listed with commas, and the wildcards * and ? work.
• Locked layer fading LAYLOCKFADECTL: objects on locked layers are faded on screen, 50 by default. Enter an integer from -90 to 90; 0 or a negative value turns fading off (a negative value keeps the number). It only affects the screen in the CAD workspace, not plotting or export, and it is saved in the drawing.
• The drop-down in the ribbon Layers panel changes the current layer directly. For the Properties panel's Linetype and Lineweight drop-downs see Linetypes, lineweights, units and purge.
• Xref layers are listed as "xrefname|layer"; see External references.

Linetypes, lineweights, units and purge

• Properties panel drop-downs: the Linetype drop-down on the ribbon "Home › Properties" panel has ByLayer, ByBlock, Continuous, the loaded linetypes and "Other..." (opens LINETYPE); the Lineweight drop-down has ByLayer, ByBlock, Default, the 24 standard lineweights and "Lineweight Settings..." (opens LWEIGHT). With a selection the choice changes the selected objects (one undo step; objects on locked layers are skipped); with nothing selected it changes the current linetype or lineweight. The Color drop-down has ByLayer, ByBlock, the 7 standard colors and "Select Colors..." (index colors 1–255 and true color); it works the same way (selection or the current color CECOLOR).
• Linetype Manager LINETYPE (LT): linetype filters (all, used, unused, with Invert filter), Load..., Delete, Current (or double-click a row), Show details (Name, Description, Global scale factor LTSCALE, Current object scale CELTSCALE). ByLayer, ByBlock, Continuous, the current linetype and linetypes in use cannot be deleted. "Load..." picks linetypes from the built-in CN-* library or from a .lin file you choose; already loaded ones are reloaded. A .lin file is limited to 4 MB; a faulty definition (invalid name, invalid dash length, more than 64 dash specifications, reserved name, duplicate name) is skipped on its own, the dialog lists its name and the reason, and the other definitions can still be loaded. -LINETYPE is the command-line version: ? (list), Load (L) (built-in library only, wildcards allowed) and Set (S).
• Custom linetypes (-LINETYPE Create (C)): enter a name, a description and the .lin "A," pattern, e.g. 12.7,-6.35,0,-6.35: positive = dash, negative = space, 0 = dot, in paper mm (scaled by LTSCALE). Dashes, spaces and dots only (no text or shapes), at most 64 elements. An existing name asks before overwriting. The definition is stored in the drawing (AutoCAD writes a .lin file) and exported to DXF.
• Lineweight Settings LWEIGHT (LW): current lineweight, Units for Listing (millimeters or inches), Display Lineweight, Default (LWDEFAULT, 0.25 mm unless changed) and Adjust Display Scale.
• Drawing Units UNITS (UN): length type (Architectural, Decimal, Engineering, Fractional, Scientific) and precision; angle type (Decimal Degrees, Deg/Min/Sec, Grads, Radians, Surveyor's Units), precision and Clockwise; insertion scale (INSUNITS); "Direction..." sets the base angle (East, North, West, South or Other). The settings are used for lengths and angles shown on the command line, by the inquiry commands (DIST, ID, LIST, AREA) and in the status bar. You can type decimal numbers (drawing units) or a length with a unit: 12'-6 1/2", 10', 6", 500mm, 3m, converted to drawing units; with the Architectural or Engineering length type the drawing unit is the inch (as in AutoCAD), and Architectural and Fractional also accept 6 1/2. In a feet or inch design whose 3D “Project information › Units” uses the feet-inches format, new CAD drawings default to inches, Architectural length units (1/16 in) and the ARCH-IMPERIAL dimension style (Architectural dimension units); existing drawings are unchanged. The system variables LUNITS, LUPREC, AUNITS, AUPREC, ANGBASE, ANGDIR, INSUNITS, LTSCALE, CELTSCALE, LWDISPLAY and LWUNITS can also be typed directly.
• Purge PURGE (PU): the "Purge" dialog has two tabs, "Purgeable Items" and "Find Non-Purgeable Items". Tick categories (Blocks, Dimension Styles, Groups, Layers, Linetypes, Table Styles, Text Styles) and the preview below lists what will be purged; options are Purge nested items (on by default), Purge zero-length geometry and Purge empty text objects. Press "Purge Checked Items" or "Purge All"; the dialog stays open, and everything purged in it is one undo step. Items that cannot be purged show the reason (in use, current item, default item, external reference, symbol, anonymous block). Anonymous table and array blocks (*T, *U) with no references are listed as purgeable. Objects inside xrefs do not count as use in this drawing; layers, linetypes and text styles used by attribute formatting kept from a DXF import count as in use. "Refresh" re-analyzes the drawing. -PURGE is the command-line version: choose a type (Blocks B, Dimstyles D, Groups G, LAyers LA, LTypes LT, textSTyles ST, Tablestyles T, Zero-length geometry Z, Empty text objects E, All A), the names, and whether to verify each one.
• Units and Purge are in the slideout of the ribbon "Home › Utilities" panel and on the "Manage" tab (see Screen).

Layouts, viewports and plotting

• Layout tabs: the left of the status bar has a "Model" tab and one tab per layout. A layout is one sheet of paper. Right-click a tab to add, rename, delete, move or copy a layout, or to open the Page Setup Manager, Plot and Publish. The command is -LAYOUT (LO).
• Paper: standard sizes include ISO A/B, ANSI and ARCH, in landscape or portrait.
• Viewports MVIEW (MV): creates a viewport in a layout. Drag a rectangle, or choose "Polygonal(P)" for a polygon viewport; there are also "Fit(F)", "Lock(L)", "ON" and "OFF". A new viewport shows the whole model extents. Typing the command on the Model tab gives "Command not allowed in Model Tab".
• Viewport scale: select a viewport and pick a scale in the viewport scale dropdown of the status bar. Standard scales (for example 1:100) are listed; any other value shows "Custom…", where you can type a scale such as 1:75.
• Locking a viewport: the viewport lock button in the status bar, or "Lock" in MVIEW. When locked, zooming and panning do not change the viewport scale or view.
• Model space and paper space: MSPACE (MS) enters the model space inside a viewport and PSPACE (PS) returns to paper space. You can also double-click inside a viewport or use the "PAPER"/"MODEL" button in the status bar. Drawing directly in paper space is not available yet; enter a viewport first.
• Page setup PAGESETUP: opens the Page Setup Manager. Set the paper size, plot area (layout, extents, window, limits), plot scale (including fit to paper), plot style table (5 built-in custom CN-* styles; you can also import a .ctb), drawing orientation and plot centering. A page setup can be made current or applied to a layout.
• Plot PLOT (PRINT): opens the plot dialog. Press "Preview…" to see the result first, then output a PDF or PNG file (a download). Plotting runs in a background worker, so the screen does not freeze.
• Publish PUBLISH: publishes several layouts into one multi-page PDF. Check the layouts to publish in the list; each row has a sheet name and a page setup.
• Title block: INSERT or "Insert Title Block" opens a dialog where you choose a title block and fill in its attributes (for example the drawing name).
• In the Vector design workspace, "Object › Convert Artboard to Layout" turns an artboard into one sheet with a locked viewport.
• Preview PREVIEW (PRE): opens the Plot dialog and generates the preview right away. EXPORTPDF: export to PDF (ribbon Output › Export; on a layout tab it follows the page setup). DXFOUT: export DXF.
• Insert a 3D view on a layout: Layout › Create View › Insert 3D View (VIEWBASE) on a layout tab. The view's lines go into model space to the right of the existing content, and the layout gets one locked viewport per view at 1:scale. After the 3D model changes, use Update View (VIEWUPDATE); the viewports follow. For the other options see Import views from 3D.

Text

• Single-line and multiline text: TEXT (DTEXT, DT) draws single-line text and MTEXT (MT, T) draws multiline text. The command line asks for the start point, justification, style, height and rotation (multiline text asks for two corners of the frame, with options such as width and columns). After the parameters you edit the text in place on the drawing, and a "Text Editor" tab appears in the ribbon. Typing the text is one undo step.
• Editing DDEDIT (ED), TEXTEDIT: edit existing text. Editing the words does not lose multiline format codes.
• Style STYLE (ST): opens the text style dialog. A text style that is in use cannot be deleted.
• Find FIND: opens "Find and Replace" with find, replace and replace all, a search scope, and whole words only. Before replacing, it checks that the match has not changed since you found it; if it has, find again. There is a limit on how many replacements one run can make.
• Chinese typesetting: Traditional Chinese has line-break prohibition rules (kinsoku). The centered punctuation used in Taiwan is not squeezed; it keeps its full-width size.

Tables and donuts

• Donut DONUT (DO): enter the inside diameter (default 0.5; 0 = a solid dot) and the outside diameter (default 1), then keep clicking centres to place donuts; Enter ends. Diameters can be typed or picked with two points. Each donut is a closed polyline with width. The diameters are remembered for this session only, not saved in the drawing. On the ribbon it is in the Draw panel slideout.
• Table TABLE (TB): ribbon "Annotate › Tables". The "Insert Table" dialog has Table style, Insertion behavior (Specify insertion point = the table's top-left corner; Specify window = column width and number of data rows worked out from the window), Columns (1–100, default 5), Column width (default 63.5), Data rows (1–998), Row height in lines (1–100), and the cell styles of the first and second rows (Title, Header, Data; all other rows are Data). A table has at most 10,000 cells. After placing, editing starts in cell A1.
• Editing cells TABLEDIT: double-click a cell, press "Edit Text" on the table tab, or type TABLEDIT and pick a cell. Type straight into the cell: Enter finishes, Alt+Enter or Shift+Enter adds a line break, Tab and Shift+Tab move to the next or previous cell (Tab in the last cell adds a row), and Esc ends.
• "Table" contextual tab: appears when one table is selected and no command is running. Click a cell to make it current; Shift+click extends the range. Rows: Insert Above, Insert Below, Delete Row(s); Columns: Insert Left, Insert Right, Delete Column(s); Merge: Merge Cells (Merge All, Merge By Row, Merge By Column) and Unmerge Cells; Cell Format: Alignment (by style or one of 9 positions); plus Edit Text and Close. Each action is one undo step.
• Table style TABLESTYLE (TS): the "Table Style" dialog has the style list, Set Current, Delete (Standard, styles in use and the current style cannot be deleted) and New (copies the selected style). Modify: text style, horizontal and vertical margins, grid lines (borders), and the text height and alignment of Title, Header and Data. Standard defaults to title text 6, header and data 4.5, margins 1.5, centred. On OK, every table using the style is rebuilt in the same undo step. CTABLESTYLE sets the current table style directly.
• Tables on locked layers cannot be edited.

Dimensions

Dimension commands follow AutoCAD, and so do the prompts. The ribbon "Annotate" tab has the same buttons. One dimension command is one undo step.

• Linear DIMLINEAR (DLI), Aligned DIMALIGNED (DAL): pick the origins of two extension lines, or press Enter and pick an object, then pick the dimension line position. Before placing you can choose "Mtext", "Text", "Angle", and for linear dimensions also "Horizontal", "Vertical" and "Rotated".
• Angular DIMANGULAR (DAN): select two lines, an arc or a circle, or press Enter to use three points (the vertex and two endpoints).
• Radius DIMRADIUS (DRA), Diameter DIMDIAMETER (DDI), Arc length DIMARC (DAR): select an arc or circle (for arc length, an arc or a polyline arc segment), then pick the dimension line position.
• Ordinate DIMORDINATE (DOR): pick the feature point, then the leader endpoint. Use "Xdatum" or "Ydatum" to set the direction.
• Continue DIMCONTINUE (DCO), Baseline DIMBASELINE (DBA): carry on from the last dimension you created, or use "Select" to choose an existing dimension as the start. Baseline spacing follows the dimension style.
• Quick dimension QDIM: select a set of geometry and place a whole row of dimensions at once. The modes are continuous, staggered, baseline, ordinate, radius and diameter, and "datumPoint" changes the datum.
• Adjust spacing DIMSPACE: select the base dimension, then the linear or aligned dimensions to space out, and enter the spacing or choose "Auto". A spacing of 0 aligns them on one line. Dimensions that are not parallel are skipped with a message.
• Smart dimension DIM: hover over an object and click. A line gives a linear or aligned dimension, an arc or circle gives a radius or diameter dimension; you can also just pick two points. The command repeats until you press Enter or Esc. Options are angular, baseline, continue, ordinate, "Layer" (put new dimensions on a given layer) and "Undo".
• Edit DIMEDIT (DED): "Home" puts the text back at its default position, "New" changes the text, "Rotate" turns the text. DIMTEDIT (DIMTED): drag the dimension text to a new position, or choose "Left", "Right", "Center", "Home" or "Angle". Both can be run after selecting the dimension first.
• Multileader MLEADER (MLD): pick the arrowhead position and the landing position, then type the text. "Options" can switch to leader-landing first or content first, and set the maximum number of points. Styles are set in the manager of MLEADERSTYLE (MLS). What is drawn is a group of lines, an arrowhead and text, not a leader object whose grips you can adjust later.
• Properties palette (Ctrl+1): with a dimension selected you can change the dimension style, text override (<> stands for the measured value), text height, text rotation, overall scale and precision; the measurement and whether it is associative are view-only.
• Dimensions on locked layers are not edited, and the command tells you how many it skipped.
• Jogged dimension DIMJOGGED (DJO, JOG): "Jogged" on the ribbon "Annotate › Dimensions" panel. Select an arc or circle, pick the center location override, pick the dimension line location (Mtext, Text and Angle are available), then pick the jog location. It is meant for large-radius arcs whose centre is off the sheet. The jog angle is DIMJOGANG (default 45°); change it on the "Symbols and Arrows" page of the dimension style (5–90°). The dimension is associated with the arc or circle, and DIMREASSOCIATE works on it.
• Dimension break DIMBREAK: "Break" on the ribbon "Annotate › Dimensions" panel. Select a dimension, then choose "Auto (A)" (break at every object that crosses the dimension; default), pick the objects to break the dimension with (Enter ends), "Manual (M)" (pick two break points) or "Remove (R)" (clear all breaks). "Multiple (M)" selects several dimensions first and then offers Break (B) or Remove (R). Only dimension lines, extension lines and leaders are broken, not arrowheads or text; text, points, hatches, construction lines, rays and groups never cause breaks. The break size is the dimension style's DIMBRK times the overall scale (default 3.75); change it with "Dimension break: Break size" on the "Symbols and Arrows" page of the dimension style. Auto and picked breaks follow their objects when those move, and disappear when the objects are deleted; manual breaks move and scale with the dimension.

Associative dimensions

• When an extension line origin snaps to an object (endpoint, midpoint, center, quadrant, intersection, node, insertion point and so on), that point is bound to the object. Later, when you move, stretch or modify the object, the dimension updates with it, in the same undo step.
• If the bound object is deleted, that point becomes non-associative and the dimension stays where it was.
• If you move only the dimension, not the object, points that no longer match are disassociated, as in AutoCAD.
• Locked dimensions (object or layer) are not updated by association.
• A dimension placed on a view line derived from the 3D model also updates when the model changes. If the original line cannot be found, the dimension keeps its old position and value and is marked yellow; use DIMREASSOCIATE to assign it again.
• DIMREASSOCIATE (DRE) reassigns the association point by point; DIMDISASSOCIATE (DDA) removes the association of the selected dimensions. The "Associative" field in the Properties palette shows the current state.

Dimension styles and annotation scale

• The default dimension style of a new drawing is CN-METRIC: scale 1:100, pen width 0.25. These values are our own design. ISO-25 and Standard are only compatibility aliases that point to the same values.
• DIMSTYLE (D) opens the Dimension Style Manager: create a style (choose a starting style), set it current, modify arrowheads, text style, prefix and suffix and so on, then Apply. The "Symbols and Arrows" page also has the dimension break size (DIMBRK) and the radius jog angle (DIMJOGANG, 5–90°). CN-METRIC and the style currently in use cannot be deleted. -DIMSTYLE is the command-line version (Save, Restore, Status, Apply, List).
• The drop-down in the ribbon "Annotate" tab switches the current dimension style directly.
• Annotation scale: annotative dimensions are sized by the annotation scale in the status bar in model space, and by the viewport scale in a layout viewport. For example, text that is 2.5 on paper is 250 at 1:100.
• The Aligned Dimension DI in 3D plan mode uses CN-LINEAR, with the same values as CN-METRIC.

CN-* linetypes

• The built-in linetype library is our own CN-* series of 30 linetypes, each with a description in Chinese and English.
• Common AutoCAD linetype names map to the closest CN-* linetype, and old names in older drawings map across too.
• DXF linetypes and dimensions also go through CN-*. When a file is read, linetype definitions that come with it are used as they are.
• The dash panel in Vector design can be written out as DXF linetypes; see Vector design › Path tools.

DXF notes

• Grid detection: when 2D CAD reads a DXF, grid lines stay as lines, circles and text. To turn the drawing's grid lines (grid layers, or bubbles + labels + long lines) into a 3D structural grid, use File ▸ Import DXF… in the 3D modeler; see DXF grid lines → structural grid.
• Export notes: when you export DXF, content that cannot be carried over or is written differently (vector objects, xrefs written as path only, simplified clipping boundaries and so on) is listed in the command line under "Export notes:".
• Dimensions: written as DIMENSION, together with visible lines, arrowheads and text (an anonymous block), so they show in AutoCAD. For dimensions created in Synhedra, the association, style overrides and annotation scale are also stored in XDATA, and when the file is read back into Synhedra they are restored as editable associative dimensions. When a DXF is read, dimension styles follow the definitions in the file; when it is written, all dimension variables are kept.
• DIMENSION objects created by other software are kept as they are and written back unchanged, but they cannot be edited as dimensions in Synhedra.
• Arc length dimensions are written as ARC_DIMENSION, which only AutoCAD 2004 (AC1018) and later recognize; the file is written as AC1015, so a warning appears.
• A multileader is ordinary lines and text, and in DXF it is ordinary geometry too, not a MULTILEADER object.
• Only ASCII DXF is read and written, and the written version is AC1015 (AutoCAD 2000). DWG is not supported.
• Blocks: a block definition is written as a BLOCK and an instance of it as an INSERT, including scale, rotation and arrays (MINSERT), and ATTRIB is written when there are attributes. Files written by WBLOCK are also AC1015.
• Hatches: written as HATCH with the full pattern definition, so AutoCAD needs no separate .pat file. Association is not written, but the shape is kept. Synhedra also marks them in XDATA, so reading the file back into Synhedra restores editable hatches. HATCH objects created by other software are kept as they are.
• Symbols: Vector design symbols are also written as BLOCK plus INSERT. The symbol type and dynamic overrides are stored separately in XDATA. AutoCAD ignores XDATA, so it shows an ordinary block reference and the overrides do not appear; reading the file back into Synhedra restores them. When the XDATA would exceed AutoCAD's limit of 16 KB per object, only the type is written and the overrides stay in the Synhedra file.
• Vector paths: written as LWPOLYLINE, SPLINE or HATCH with Synhedra's own XDATA, and restored as vector paths when read back.
• Arrays, regions, groups and views imported from 3D: an array is written as an INSERT of an anonymous block and a region as ordinary geometry; their parameters or markers go in XDATA, and Synhedra restores them when the file is read back. If an array's parameters are too large for XDATA, it is written as a plain block reference and a warning is shown. Named groups keep their description and selectable setting. The records of views imported from 3D are stored in XDATA too.
• Jogged dimensions and dimension breaks: a jogged dimension is written as a radius DIMENSION, with the override and jog points in XDATA; breaks are drawn into the dimension's anonymous block, so AutoCAD shows them. Synhedra restores both when the file is read back.
• External references: written as an xref BLOCK (type, loaded flag and the file name as path); when the drawing has xrefs, the export first asks whether to write them as "Reference" or "Bind content". Reference writes only the path, not the content; the export notes remind you that the recipient needs the source file, and xref layers are written as "name|layer". Bind content writes loaded xrefs with their content as ordinary blocks (layers renamed "name$0$layer"; the drawing itself is unchanged); unloaded ones are still written as path only. The XCLIP boundary (including Invert) is stored in XDATA, so AutoCAD shows the reference unclipped; a boundary too large for XDATA is simplified first, and the export notes give the change in point count. When read back from DXF, xrefs have no content (status Unloaded or Not Found); press "Reload…" in the palette and choose each file again.
• Tables: written as an INSERT of an anonymous block (lines and multiline text) with the table data in XDATA; this is not ACAD_TABLE, so AutoCAD sees a block. Table style definitions are not written; each table carries its own copy of the style settings.
• INSERT with attributes: when a DXF is read, an INSERT followed by ATTRIBs becomes a block reference whose attribute values can be edited; each attribute keeps its layer, color, style, width factor, oblique angle, invisible flag and other formatting, and is written back in that format. INSERTs with duplicate tags or data that cannot be represented are kept and written back unchanged.
• Units: $LUNITS, $LUPREC, $AUNITS, $AUPREC, $ANGBASE, $ANGDIR, $INSUNITS and $LWDISPLAY are written and read back. LWUNITS, LWDEFAULT, the lineweight display scale and XCLIPFRAME are kept only in the Synhedra file.

Other shortcuts

This table is maintained by hand (source: keys.ts).

Import, export, saving

• Drawings save automatically; there is no Ctrl+S. The save status is at the right of the status bar. 2D drawings are stored in the browser's database (IndexedDB).
• Import DXF: application menu or DXFIN. ASCII DXF only. Save binary DXF as ASCII in AutoCAD. For DWG, save as DXF or convert with ODA File Converter.
• Export DXF: application menu. For content that cannot be carried over, see "Export notes" in DXF notes.
• The unit is chosen when the design is created; the display format of lengths and angles is set with UNITS (see Linetypes, lineweights, units and purge).
• When this site has configured cloud drives (not yet available on the live site), the "File" menu has "Open from cloud drive…", "Save to cloud drive…" and "Save a copy to cloud drive…", optionally with DXF. A read-only drawing cannot be saved to a drive. See Third-party cloud drives.
• To turn a photo, scanned plan or sketch into drawing objects, AIIMAGE, AISKETCH or "AI canvas (image / freehand)…" in the application menu open the AI canvas.

Coming soon

Gradient fill; Ltype gen in PEDIT; Quick measurement in MEASUREGEOM; Key point, and Edit Source, Replace Item and Reset Array for arrays; Order in -GROUP; and Save Block As in the Block Editor. Typing these commands shows a message in the command line.