Help

Core concepts#

A graph is made of nodes (entities) and the edges between them. Every node and every edge has a type; color, size and the fields to fill in come from the type.

A graph opens in two modes. In view mode anyone can explore, search and read cards. Edit mode (E) is open only to people with edit rights; adding, connecting, deleting and type settings live there.

The Legend at the top right lists the graph's types with their colors and lets you show, fade or hide each type. The Graph info panel (I) shows the description, the Node types and Edge types lists and the saved views.

Selecting a node shows the labels of its edges; hovering a row in its card's relations list highlights that neighbor in the graph. Edges between the same two nodes share one label (e.g. H2 · H3); a label that would overlap another label or a node label is not drawn and appears as you zoom in. Double-click a node whose description contains a graph link (e.g. a province node linking to that province's graph) to open that link. Double-click only goes down: links to the current graph or its parent graph are skipped (a district in a province graph does not take you back to the Türkiye graph; use the link on its card to go up). On a node with no such link, double-click zooms in.

The search box (/) finds nodes by name and by the label shown on screen, ignoring Turkish letters and case. A node's Former names property (key eski_adlar; add it to your own nodes as “Eski adlar”) is searched too: in the village and neighbourhood showcases an old place name (“Şağuç”) finds today's name (Aydınpınar) and the result shows a Former names: … line. Former names come from Wikipedia articles and from the Ministry of the Interior's Köylerimiz (1968) list; a name from the list is marked “(1968 list)” (the list does not give the year of the change). When a known former name also appears in the Ministry's Köylerimizin Adları (1928) list, the year 1928 is added (“Sinek (1928; 1968 list)”). A village that could be matched to that book has a Timeline on its card: dated former names in order, the 1928 entry (Turkish reading, the book's Ottoman and Latin-script spellings, that year's province, district and sub-district, and a link to the right page of the book's PDF), the former name in the 1968 list and today's name. The 1928 Ottoman spelling is searchable too: just type “قره بوجاق”; spelling variants such as ى/ي/ی and ك/ک/گ count as the same. The 1928 readings were made by machine and may contain errors. The card shows a source line naming only the sources that place's names rely on. Matches by former name come after matches by name. Search the site below the box searches every node on the site for the same word, former names included.

Your first graph#

Start from an empty graph: add a few nodes, pick their types and connect them; everything is saved automatically.

The New graph button on My graphs opens an empty graph. An empty graph reads “This graph is empty” and offers three ways in: Double-click to add a node, Enter from the table or Upload a file. Rename the graph in place by clicking the title in the middle of the top bar: Enter or clicking elsewhere saves, Esc cancels.

Enter edit mode with Edit (E) and double-click an empty spot; on a phone, long-press the empty canvas. Type a name in the New node box and pick a type; typing a type name that is not in the list creates a New type.

To connect two nodes, hold Alt and drag from one to the other; on a phone, tap Connect on the card, then the target node. Then pick an edge type or type a new one.

Every change is saved automatically; when the indicator in the top bar says Saved, your work is on the server. Press E again (Exit editing) to leave edit mode.

Graph list and groups#

My graphs and profile pages show only top-level (root) graphs as cards; on a card with subgraphs, the N subgraphs button opens the group.

A group page starts with a location path (My graphs › … › group); below it, subgraphs are listed under section headings (e.g. Black Sea Region (18)). Clicking a heading collapses the section; the section search box searches only that section. Subgraphs without a section are under Other; the arrow on a row with its own subgraphs goes one level down. Showcase graphs (provinces and regions, state organisation, parliament, local elections, geographical indications) come in this structure; in the Provinces and Regions of Türkiye graph the seven geographical regions sit below Türkiye with the provinces below them, and the province graphs are listed by region.

When you open a graph, the Subgraphs section of the Info panel (I), below Views, lists that graph's direct subgraphs (those you can see) under their section headings with node counts; click a row to open it (e.g. a dynasty's family graphs, province graphs). Subgraphs page below the section opens the same list as a separate, shareable page (address: the graph address followed by /alt-grafikler); on the home page, a card of a graph that has subgraphs shows an N subgraphs link to that page. The page shows each subgraph's node count and, if it has its own subgraphs, a link to their page; only subgraphs everyone can see (public and members) are counted.

The Search graphs box at the top searches all levels (ignoring Turkish letters and case) and shows each result with its path. Root cards are sorted by Recently edited or By name; within a section the order is the one the showcase gives (e.g. plate number), then name. Shared with me is a separate list in the same form. A subgraph whose parent you cannot see (e.g. private) counts as a root for you.

To put your own graph in a group, choose a Parent graph under Place in the graph list in the graph settings (only your own graphs; type in the box to filter) and optionally give a Section name; None (top level) moves the graph back to the top. Only the graph owner can do this; a graph cannot be placed under one of its own subgraphs.

Direction and naming#

Edge types are directed or undirected. A directed edge reads like a sentence from source to target: Kerem —FATHER→ Tarık means “Kerem's father is Tarık”.

Directed types draw an arrow on the canvas, and the node card lists edges in outgoing (↗), incoming (↙) and undirected (—) groups. If both ends play the same role (MARRIED TO, SIBLING OF, PARTNER OF), make the type undirected: the line has no arrow and both ends show the same name. The Directed box of the edge type in Cartography sets the direction; the import summary also lets you choose it for each edge type.

Do not enter the reverse direction as a second edge. Adding Tarık —CHILD→ Kerem next to Kerem —FATHER→ Tarık writes the same fact twice: nodes sized by connection count grow, the layout pulls the two nodes together twice as hard, and once one of them is deleted and the other forgotten the data contradicts itself. A single edge can be read from both ends.

For edges that read differently from the other end, give the type an inverse name: the Inverse name field of a directed edge type in Cartography. If the inverse name of FATHER is CHILD, Tarık's card shows this edge as Kerem in the “↙ CHILD” group, and the edge card shows a “← CHILD” line under the name. Only directed types have an inverse name; when a type is made undirected the field is hidden and the value is kept.

Reading rule: the name reads from the source, the inverse name from the target. Choose the name so it completes the source's sentence (Garanti —LENT TO→ Star Refinery) and write the inverse name from the target's point of view (Star Refinery's card: “BORROWED FROM: Garanti”). Canvas labels, the legend and the data table always show the name.

Naming: keep edge names short and in CAPITALS (FATHER, MEMBER OF, PARTNER OF). Do not write the same meaning under two names: every distinct name becomes a separate type and their edges are counted separately. Type names are unique within a graph.

Editing#

Enter edit mode with E or the pencil button in the left toolbar. Adding nodes, connecting, editing cards, deleting and undo all happen in this mode.

Add a node: double-click an empty spot (long-press on a phone) and pick a name and type. Connect: drag from one node to another with Alt, or press Connect on the card and click the target; Esc cancels.

Clicking a node or edge opens its card on the right. In edit mode you change the name, type, Description and image (Image URL or Upload) there; + Property adds a new field to the type. On an edge card, Weight sets the line thickness. http:// and https:// addresses in a description become clickable links on the card and in the Graph info panel: links to this site open in the same tab, external ones (and all links in an embedded view) in a new tab. A line consisting only of Label: address (e.g. “Municipalities: https://…”) shows its label instead of the address; consecutive such lines form one small footnote line separated by “·” (a repeated label gets superscript 1, 2, 3) (in “Parent: Some graph https://…” the name becomes the link); other addresses are shown with an icon and shortened (this site's graph addresses by graph and node name, e.g. “duzce-cografi-isaretleri › duzce-81”), and the full address appears on hover. In a card's Relations list, relation groups shown in the same direction with the same name merge under one heading (e.g. NATURAL SITES on a province card in a showcase root); each row shows a coloured dot, the name and below it the relation's label (e.g. 5 sites) or the node type. In edit mode, the × at the end of each row in a node card's Relations list removes that relation (always visible on touch screens, on hover on desktop); the card stays open and ⌘Z brings it back. The Edge type field at the top of a relation card changes the relation's type: click it to list every relation type in the graph, then pick one or type a new name. If you type the name of an existing type (case does not matter) that type is used and no duplicate is created; ⌘Z undoes the change. You do not have to hit a line exactly to open a relation: as the pointer nears a line it thickens, shows its name and the cursor becomes a hand; click to open its card (on touch screens tapping near the line is enough).

Right-click a node (long-press on a phone, or ⋯ More actions in the card header). The menu is grouped: Edit (in edit mode Connect, Duplicate, Change type), Select (Select with descendants, Select with neighbors, Add to selection) and View (Focus here and the view rows); in edit mode Delete sits at the bottom. On a phone the menu also has Arrange; on tablet and desktop the arranging tools live only in the Arrange toolbar. Right-click a relation — its line on the canvas or its row in a node card's Relations list — for a small menu: Open relation shows the edge card, and in edit mode Remove deletes it.

Multiple selection: hold Shift and click nodes one by one (Shift-clicking again removes a node; the card of the first node you clicked stays open) or drag on an empty area to draw a box; on a phone add nodes one by one with Add to selection in the menu. The selection feeds the align, distribute and view tools. With a selection, the arranging buttons in the arrange toolbar (Selected: around, below, above, left, right, arcs; Pull selected closer / Push selected apart) arrange the selected nodes, instead of the linked nodes, around the node they belong to (the parent most of them share, even if it is not selected); without a shared parent, the node whose card is open is the centre. Hold ⌘/Ctrl while dragging a node to move it with everything under it (its subtree); the node it belongs to stays put.

Delete: Delete or Backspace removes the selected item (or the whole selection). Undo is ⌘Z / Ctrl+Z, Redo ⌘⇧Z / Ctrl+Shift+Z; the toolbar has the same buttons. A bulk action is undone in one step; a manual label position set in the node card can be undone too.

Changes are saved automatically. If the connection drops, the indicator retries; if saving fails for good, use Download changes as JSON or Discard pending changes. If someone else edited at the same time, the graph reloads.

Cartography#

Cartography (C) sets how types look: color, size, label and direction. A setting applies to every node or edge of that type. Deleting nodes does not delete their types: types stay as the graph's template. Type pickers show how often each type is used; unused ones are dimmed at the bottom with an unused tag, and Remove unused types in Cartography deletes them all in one step (⌘Z brings them back).

Node types: Color; the Size by rule with minimum and maximum size; Label (a property shown on the canvas instead of the name). + Node type adds a type; Delete type removes the type and its nodes (after confirmation).

Edge types: color, the Directed box and, for directed types, the Inverse name (see Direction and naming). If the type has properties, Line label writes a property value on the line instead of the Type name (e.g. the amount on a CONTAINS edge); edges with an empty value keep the type name. + Edge type adds a type. If several relation types share a name (e.g. Add organization used to create a separate UNIT OF for every pair of node types) a warning appears above the list: Merge moves the extra types' relations to the first one and deletes the extras; no relation is lost and the change can be undone from the graph history. Add organization and Add place now reuse an existing type with the same name.

With Shared scale across types on, sizes follow the smallest and largest value in the whole graph; with it off, each type is scaled on its own.

Set as visitor default saves the current view settings (layout, size and label options) and the legend's type filters to the graph: visitors open the graph with them. While a saved view is open, your own type selection outside the view is stored, not the view's.

Size rules:

Connections

The number of all edges touching the node; well-connected nodes look bigger.

Neutral

Every node of the type has the same size: halfway between minimum and maximum.

Incoming

The number of incoming edges (undirected edges count too); nodes pointed at a lot grow.

Outgoing

The number of outgoing edges (undirected edges count too).

Incoming weighted

The sum of the weights of incoming edges.

Outgoing weighted

The sum of the weights of outgoing edges.

Data table#

The data table (T) shows the graph as rows, one type at a time; it is the fastest way to enter many nodes by hand.

Pick a node or edge type from the tabs at the top. Double-click a cell, or press Enter / F2 on a focused cell, to edit it; Esc cancels. For node types, typing a name in the New row field at the bottom and pressing Enter adds a node.

Search table filters the rows; Download CSV writes the rows of the open type to a file. Selecting a row also selects that node or edge in the graph.

Import#

Import (edit toolbar) adds or updates nodes and edges from a file. A summary comes first; nothing is written until you confirm.

The File tab accepts CSV, Excel (.xlsx), a wide.tr JSON backup and an e-Devlet lineage document (PDF). A JSON backup restores a graph with its types, positions and settings; spreadsheets are read by the headers below; for the lineage document see "e-Devlet lineage document".

Each sheet (or CSV file) is recognized by its header row: a node sheet needs at least a type and a name column, an edge sheet needs source type, source name, target type and target name columns. Every unrecognized column becomes a property; ones with an empty header, one that collides with another column, or one that is an alias of a recognized column are skipped, and the summary reports this as a warning. Case, Turkish characters and underscores in headers do not matter.

A node with the same type and name is not created again but updated; if there is an Id column, matching uses the identifier. That is why you can edit an exported Excel file and import it back.

On the Google Sheets tab, paste the spreadsheet address into Google Sheets link and press Fetch. The sheet in the link (gid) is fetched; the spreadsheet must be viewable by anyone with the link.

The summary shows how many nodes or edges each type brings. You can leave a type out with Exclude, choose the direction of edge types, and take the settings of a JSON backup with Apply graph settings too. Warnings are listed with row numbers; if there are errors, fix the file and upload it again. The last step is Import.

Recognized column headers (in each row the leftmost is the most specific; if a sheet has several, that one wins):

e-Devlet lineage document#

Save the result of the Alt-Üst Soy Bilgisi Sorgulama (lineage) service on turkiye.gov.tr as a PDF from your browser and drop it into the Import window. People, ANNESİ/BABASI (mother/father) relations — read as ÇOCUĞU (child) from the other end —, registry villages/neighborhoods and birth places are created, and the tree is laid out by generation.

Uploading the same document again creates no duplicates: people are recognised by registry place and volume-household-person number. Another family member's document (e.g. your spouse's) merges the shared people; a parent known only by name is updated in place when they arrive with their own record.

Parents without their own record are added from the child's father/mother name field as light-colored "yalnızca adı" (name only) nodes; exclude those two types on the summary screen if you don't want them.

Registry places are linked to the place layer; a village that cannot be found is linked to its district and listed as a warning. Birth places outside the place layer (abroad or ambiguous names) become "Diğer Yer" (other place) nodes placed at the top.

The document contains personal data. If the graph is public or shared by link, import asks for confirmation first; the uploaded file is deleted from the server once imported, or when you abandon the import (close the dialog or switch to another file).

Warnings on the summary: name or sex does not match the kinship, unreadable kinship, village not found, ambiguous or unknown birth place, a parent that exists as two separate nodes (delete one by hand). Corrupt PDFs and PDFs that are not lineage documents are rejected.

The Also create the family views box on the summary (on by default) builds the five standard family views after the import, rooted at the document's own person; views whose name already exists are skipped.

Export#

Export (edit toolbar; visible only to people with edit rights) downloads the graph as a file.

JSON backup: types, nodes with their positions, edges, graph settings, the graph's language and all translations; restores into the same or another graph (saved views and members are not included). Excel (.xlsx) and CSV (zip): one sheet per type, with the headers Import recognizes. GraphML and GEXF: for network analysis tools such as Gephi; these two formats cannot be used for very large graphs, and a warning is shown if the limit is exceeded.

To save the current picture of the graph, use Download PNG in the Share dialog.

Views#

A view saves a part of the graph you choose, in its own arrangement. Moving nodes in a view does not change the base graph; it is how you tell different stories with the same graph.

Views are listed in the Graph info panel (I). Select nodes and press ⇧V, or choose Create view from selection… in the right-click menu; with nothing selected, New view starts from the whole graph. With Also save camera checked, the view opens at the same zoom.

In an open view, the right-click menu offers Add to view, Remove from view and Add neighbors to view; dragging and running a layout affect only the view; in edit mode, moves, adding/removing nodes, running a layout (label direction included) and Back to the view's saved positions inside a view can be reverted with Undo (⌘Z) too. The view chip at the top has Save view, Save as… and Close view; Back to the view's saved positions in the Layout panel returns to the last save.

Use Rename or Delete on a view in the list. The ?view= link in the address bar opens the graph with that view; creating a view needs write access, opening one does not.

In family graphs, select a person who has an ANNESİ/BABASI (mother/father) link and choose Create family views to build five views from them: Family tree (the whole lineage, hourglass), Mother's side and Father's side (ancestor trees), Migration (people and birth places; abroad at the top, Turkey at the bottom) and Registry places (village → district → province chain). Places start hidden in the views where they don't belong. A view whose name already exists is skipped; the notice says how many were created and skipped.

The showcase's own views (such as Kompakt and Hilal in the administrative graphs, or Cumhurbaşkanlığı in the organization graph) are refreshed whenever the showcase is rebuilt. With write access you can edit them too and keep the result with Save in the chip, or rename them in the list: the layout you save survives showcase rebuilds (nodes new to the showcase are added to the view), and the showcase's own version is kept separately. Back to the showcase version, in the note under the chip and in the row menu of the list, restores what the showcase generated. Use Save as… for a separate copy; deleting a showcase view is allowed, it comes back with the next rebuild.

Legend and type filters#

The Legend at the top right lists node and edge types. Clicking a row's button cycles the type through visible, faded and hidden; Show all brings everything back.

Nodes of a hidden type and their relations are not drawn and cannot be clicked; a hidden edge type only removes those lines, the nodes stay. A faded type is drawn smaller in a light tone of its color; it is still highlighted when it neighbors the selected node. Edge types with the same name share one row; if their states differ (for example only one of them is hidden) the row shows mixed and clicking it makes them all visible.

Layouts (ForceAtlas2 and one-off layouts) and clusters only use visible types: hidden nodes neither move nor push others. Faded types are included. Hiding or showing a type clears computed clusters; compute them again if needed.

Search also finds nodes of a hidden type and marks them type hidden; choosing one makes that type visible. In a node card's neighbor list, neighbors of a hidden type are faded and marked too; clicking one makes its type visible.

The selection is yours and is not written to the address bar; reloading the page returns to the graph's default. The data table and exports are not affected by the filter.

When you save a view, the current selection is stored with it and applied whenever the view is opened. This holds even when no type is hidden or faded: that view then shows every type, whatever your own selection or the graph's default. Views saved before type filters existed leave your selection unchanged. Closing the view returns you to your selection from before. Changing the selection while a view is open marks the view as unsaved.

Set as visitor default in Cartography also stores your own selection as the graph's default: everyone opening the graph starts with it. While a view is open, it is your selection outside the view (the one you return to when you close it) that is stored, not the view's. A JSON backup carries this default; when the backup is opened in another graph the selection is mapped to that graph's types.

Arrange tools and layouts#

A layout positions nodes automatically; the arrange tools spread, align and distribute selected nodes by hand. Positions are saved in edit mode; changes in view mode stay on your screen only.

Pick a Layout in the Layout panel (V or the lowest button at the bottom right). ForceAtlas2 runs continuously: Space or the button at the bottom right toggles Run layout / Stop layout; in edit mode positions are saved when the layout stops. On large graphs Cancel computation appears while it runs. Save positions writes the current positions at once, Reset positions returns to how the page opened.

With ForceAtlas2 selected, Gravity pulls nodes to the center and Scaling spreads them apart.

Node and relation appearance is in the button column at the bottom right. The Node button (a dot and two lines): Node names — When (Auto — names that fit on screen, without overlaps; In focus — only the hovered or selected node and its connected nodes; Never — only the node in focus) and Name position (Auto, Right, Left, Top, Bottom; the default for the whole graph, tried first — a position you set on a single node comes first), Node size, Label size (shrinks or enlarges node and relation text; shrinking it in crowded graphs reduces overlapping labels), Shorten labels and Show images; clicking it switches node names between Auto and In focus. The Aa button: Relation names — When (Always; In focus — relations of the hovered or selected node; Selected — only the clicked node's relations; Never), Where (Above the line, On it — cutting through the line, Below), Direction (Parallel to the line or always Horizontal) and Show weights; clicking it switches relation names between Always and In focus (pressed means names always show; struck through means Never, dotted underline means Selected). On both buttons the settings palette opens on hover (on touch screens tap it); the name of the relation under the pointer or with its card open shows in every mode. These change only your screen; to show them to visitors use Set as visitor default in Cartography. Restore default settings at the bottom of the Layout panel returns these settings and the legend filter to the graph's saved default (like reloading the page; positions and manually placed labels are not changed).

Labels never overlap each other or nodes: the selected (or hovered) node and its neighbors come first, then larger nodes; labels that don't fit are hidden and appear as you zoom in. Each label tries right, left, above and below in turn and takes the first spot no edge crosses (or the one crossed least); a white outline cuts any edge passing behind the text.

In Flower and Crescent labels follow the rays: the first level points inward, deeper levels outward, the root's label sits below it. A label first looks for room along its ray, then shifted to either side of it, and last in the opposite direction. Running another layout turns labels horizontal again (manually chosen positions stay). Make the direction permanent for visitors with Set as visitor default in Cartography; saved views keep their own direction.

A label's position and direction are set by hand in the label group of the arrange bar (below). Phones have no arrange bar, so there the node card shows a Label row: Automatic, Right, Left, Top or Bottom; Apply to connected nodes too gives the same choice to the node's neighbors that are visible on screen (not those hidden by the view or the legend filter). A manually placed label is drawn even below the size threshold as long as the node is at least 3 px on screen and not faded by focus; it is drawn on top and horizontal (in Flower and Crescent too), and other labels avoid it. The label of a very small or faded node is not drawn. In edit mode the choice is saved at once (with the view's Save when a view is open) and can be reverted with Undo (⌘Z); in view mode and for visitors it stays on your screen only. The label group in the arrange bar (tablet and desktop) applies to the Shift-selected nodes, or to the node whose card is open — besides Right, Left, Top, Bottom and Auto it sets the text direction: bottom to top, top to bottom, at an angle (degrees) (∠ button; 0 horizontal, 90 bottom to top, negative goes downward; Apply) and radial: labels point OUTWARD from the node whose card is open (the selection if any, otherwise its connected nodes; with no center, click a node first). A direction removes the position and vice versa; all can be undone with Undo. At most 5,000 nodes per graph can have a manually placed label.

Arrange in the long-press menu on a phone: Items under it go Around, Below, Above, Left, Right of the node or on a half circle (Arc on the right, Arc on the left, Arc above, Arc below); Pull linked nodes closer / Push linked nodes apart scale the subtree, each moving together with its own sub-items; with Wrap into rows on, many items are laid out in several short rows (or columns for Left/Right) instead of one long line; Fan out neighbors opens the neighbours into an arc towards the free side; under Selection: Spread selection and Contract selection (⌘⇧] / ⌘⇧[; the whole graph when nothing is selected), Align (at least two selected nodes: Left, Right, Top, Bottom, Align vertical centers, Align horizontal centers) and Distribute (at least three: Even horizontal spacing, Even vertical spacing). On tablet and desktop the same tools are in the Arrange toolbar.

Snap to alignment: while dragging a node in edit mode, getting close to the same horizontal or vertical line as another node on screen shows a pink dashed guide and the node snaps to it (siblings under the same parent come first). The distance to the nearest linked node and the angle from horizontal are shown at the same time (e.g. 4.0 · 45°); the angle snaps to multiples of 15° (this wins over lining up with an unrelated nearby node); the measurement label sits in the middle of the line so it does not cover the node's name. Equal spacing: when lined up in a column or row, the node snaps to the exact middle between its two neighbors on that line, or to where it continues the neighboring gap; equal distance: when not locked to a line, the distance to the linked node snaps to the distance of another node linked to it (the direction is kept). In both cases the equal parts are drawn as purple lines with end ticks and the label starts with =. Circle guide: when the node's siblings sit on a circle around their parent, that circle is drawn dashed while dragging, with small rings where the siblings would be evenly spaced; the node snaps onto the circle, and onto a ring when close to one (the circle wins over the other snaps), and the measurement is taken from the parent. Dropping on the circle shows Placed on the circle with a Space siblings evenly button that spreads all siblings on that circle at equal angles (undo with ⌘Z). On tablet and desktop the magnet in the left toolbar turns snapping and the circle on and off (on phones snapping is always on); holding Alt while dragging turns it off temporarily. Direction lock: holding Shift once a drag has started keeps the node on a horizontal or vertical line from where it started (whichever way you pull more); the lock line is drawn in pink. Pressing Shift BEFORE dragging and clicking still adds the node to or removes it from the selection, as before. The arrow keys move the Shift-selected nodes, or the node whose card is open, by 1 pixel on screen, 10 with Shift (for fine adjustment; saved in edit mode and undone with ⌘Z, written to the view when a view is open).

Arrange toolbar (tablet and desktop): while a node's card is open or at least two nodes are selected, a short bar appears at the bottom of the screen; on these screens it is the only place for the arranging tools. The tools are grouped like a Photoshop toolbox: each button shows its group's last used action and repeats it in one click; hover it (on touch screens tap the small triangle in its corner) and all the group's options open above it, with the main button's action framed. Groups: Arrange linked nodes — linked nodes around / below / above / left / right, Arc on the right, Arc on the left, Arc above, Arc below (evenly on an arc on that side) and fan out neighbours; below them Arc angle (180°, 120°, 90°; each click moves to the next, remembered) tightens the arc and Wrap into rows splits many items into short rows. Closer, apart, rotate — Pull linked nodes closer / Push linked nodes apart scales the node's subtree without changing its shape, Rotate linked nodes turns it 15° per click (with no open card, Spread and Contract for the selection). Align and distribute (at least two selected nodes): Align and Distribute (three or more). Label position and Label direction (see below). With a selection the actions apply to the selected nodes; they work in every mode. On a phone the same actions are in the Arrange part of the menu that opens on a long press on a node.

Clusters (Graph info panel): Compute clusters finds tightly connected groups of nodes; then Color by cluster, Cluster labels, Separate clusters and Gather clusters become available. Recompute after the graph changes.

The Family tree layout reads ANNESİ/BABASI (mother/father) relations — and types whose inverse name is ÇOCUĞU (child) — and uses the SELECTED person as the root; if the selected node has no ANNESİ/BABASI link (nothing selected, or a place selected) the button is disabled, and if the link type is hidden in the legend a hint is shown and nothing moves. There are four types: Ancestors (root at the bottom, mother's side on the left, ancestors upwards), Descendants (root at the top, children downwards by age, partners beside them), Hourglass (both) and Fan chart (root in the centre, one ring per generation; mother's side on the left half, father's side on the right). The root stays in place; places and nodes outside the lineage do not move.

Camera: Fit to screen (⌘0), Zoom in and Zoom out (⌘+ / ⌘-), the mouse wheel or two fingers.

Layouts, in three groups:

Places#

Add place (edit toolbar) adds one of Turkey's provinces, districts, towns, neighborhoods, villages or streets to the graph as a real node, together with its parent units.

Type at least two letters in the search box (Search province, district, town, neighborhood, village or street); add the parent unit to narrow it down (“moda kadıköy”). Move through the results with the arrow keys and add with Enter.

Streets (avenues, streets, boulevards, squares, housing clusters) are left out of one-word searches, so that provinces, districts and neighborhoods are not buried under thousands of streets with the same name. To find a street, add its parent unit (“bağdat kadıköy”) or tick Search streets too; then a single word searches streets as well. With Add with parent units on, a street is added with the chain Street → Neighborhood → District → Province (a village street has Village instead of Neighborhood; in a town, Town comes in between). Street data dates from April 2021: newer or renamed streets may be missing.

With Add with parent units on, the place is added together with its district and province, linked by PART OF edges. If you open the dialog with a node selected, the selected node is linked to the place under the name you type in Relation (optional) (e.g. BIRTHPLACE).

A place is never added twice: a place and its parent units already in the graph are reused. A single ⌘Z undoes everything that was added.

On the card of a province, district or village (neighborhood, town) node (also while viewing, and for places added with "Add place"), the Communities section lists the communities of the "Communities of Türkiye" showcase in which that place appears: each row shows the community's colour and name, for a province or district the number of places (current and historical records separately), for a village current or historical record and its tribe if known. Clicking a row opens the place selected in that community's graph. The first 3 rows are shown and Show all expands the rest; if the place is in no community the section is hidden. The card goes Relations, Museums and sites, Natural sites, Communities, Geographical indications.

On the card of a district, village, neighbourhood or town node (in any graph) you can enter sourced information about that place: Community (current or historical, optional subgroup), Former name (year and language of origin optional), Local (community-language) name, Tribe , Correct a wrong record (removes a community record or former name that came from Wikipedia, with a reason) or 1928 reading correction (Ottoman) (when the machine reading of the 1928 book is wrong for that village: the correct Ottoman spelling, 1928 Latin spelling or Turkish reading; at least one; once approved it shows in the timeline and the 1928 graphs). Every member with a verified email can add information: a signed-out visitor sees the same buttons and is taken to sign in (and back to the same card); a member who has not verified their email is asked to Verify your email and sent to the settings. A member's entry is a suggestion: it is published once a site administrator approves it (at most 20 suggestions a day); a site administrator's entry is published immediately. Source is required (“Local knowledge”, a book, an article…); link and note are optional. No person's name or personal data is asked for — do not enter any: a contribution is only about the place. Once published, "Communities of Türkiye", "Language Origins of Former Place Names" and the administrative structure graph of the place's province are updated in the background within a few seconds; village cards show the contribution with a "Source: wide.tr contribution (@user, date) — source" line. The card's Contributions section (below Communities) lists the place's approved contributions with their sources and your own suggestions with their status: awaiting approval (only you and administrators see it; meanwhile you can change it with Edit or remove it with Withdraw), rejected (with the administrator's reason; only you see it) or only for the contributor (not published; only you see it on the card). An approved, rejected or only-for-you suggestion cannot be changed; open a new suggestion. You are emailed about every decision; all your suggestions are in the My contributions section of the My graphs page. A site administrator decides on a pending suggestion on the card or in the admin panel with Approve, Reject (reason required) or Contributor only, and may edit it before approving; an administrator can also change a contribution with Edit (the type stays the same; every change is kept as a version) or remove it from the graphs and from visitors with Hide; a hidden contribution appears faded to administrators and comes back with Restore. The Contributions page of the admin panel (the tab shows how many are pending) lists pending suggestions first and offers the same actions plus History (versions, “Revert to this version”) and, only for a personal-data (KVKK) request or a meaningless entry, Delete permanently (with a reason and a two-step confirmation; the versions are deleted too). Trusted contributor: a site administrator can make a member a trusted contributor on their member page in the admin panel; their entries are published without waiting for approval (the daily limit of 20 still applies; a trusted badge appears next to their name on the card), and correcting a published contribution is the administrator's job. Discussion: on the row of your own suggestion (and on every contribution for administrators) a Discussion (n) area opens; there you and the site administrators talk about the suggestion — other members and visitors do not see it. When you write, the other side is emailed (at most one email per suggestion every 10 minutes); do not write personal names or personal data. The discussion is also reachable from My contributions; you can Edit your own comment (it gets an edited label; hover to see when; editing sends no email) or Delete it; an administrator cannot edit someone else's comment, only delete it. You can turn discussion emails off with Settings → Contribution discussion emails; emails about decisions (approved, rejected, only for you) still arrive. Site administrators are notified of new suggestions by a summary email at most once an hour; decision reasons (including the optional approval note) are kept with every version in the panel's History.

On a province or district node's card (also while viewing), the Geographical indications section at the very end of the card lists that place's geographical indications registered with or applied for at TÜRKPATENT (a province card includes its districts'); on an institution node it lists what the institution registered (including joint applications with other applicants). If the product is in the showcase, clicking its name (or photo) opens it selected in its province's graph of the "Geographical Indications of Türkiye" showcase. Each row shows a photo, the kind (Indication of source, Designation of origin, Traditional product name), an application badge for pending ones, the Registration certificate (PDF) and a Source record link (links open in a new tab); when any applicant is an individual (even alongside an institution) the certificate is not archived (it contains their name), only the TÜRKPATENT source link is given; the first 3 rows are shown and Show all expands the rest. The section is hidden when the list is empty. Data and images come from the TÜRKPATENT Geographical Indications Portal (the source note is on the product's own page); names of individual applicants are not shown. All of them are also in the "Türkiye Coğrafi İşaretleri" showcase: the main graph shows provinces and product groups; the products are in the sub-graphs linked from the province and product group cards.

The "Museums and Archaeological Sites of Türkiye" showcase links museums, palaces and pavilions, and archaeological sites in Türkiye to the administrative structure: the main graph shows the three groups and the provinces; each group's own graph (Museums in Türkiye, Palaces and Pavilions, Archaeological Sites) shows its sites on a map with their provinces, districts and villages/neighbourhoods, coloured by type (each type is its own node type; general and other types are pale grey); views filter by type, operator (e.g. Ministry (muze.gov.tr)), UNESCO and sites open to visitors. A site card shows its type, operator, address, visiting hours and closing days from muze.gov.tr (as of the snapshot date; they may have changed), MüzeKart (only where muze.gov.tr states it) and links to Wikipedia, Wikidata, OpenStreetMap, the state organization and the administrative structure. On a province, district or village (neighbourhood, town) node's card (also while viewing, and on nodes added with "Add place") and on an institution node, the Museums and sites section lists the sites in that place (for an institution, its sites): each row shows the group colour, the name and type · operator; a site matched only as the nearest settlement of a village says nearby. UNESCO World Heritage comes first, then sites listed on muze.gov.tr or by the National Palaces, then by name; clicking a row opens the site selected in the showcase graph. The first 3 rows are shown and Show all expands the rest; the section is hidden when the list is empty. The card goes Relations, Museums and sites, Natural sites, Communities, Geographical indications.

The "Natural Sites of Türkiye" showcase links Türkiye's protected areas, waterfalls and lakes, highland pastures and mountains, islands and caves to the administrative structure: the main graph shows the five groups and the provinces (a protected area spanning several provinces counts in each); each group's own graph (Protected Areas, Waterfalls and Lakes, Highland Pastures and Mountains, Islands, Caves) shows its sites on a map with their provinces, districts and villages/neighbourhoods, coloured by type (national park, nature park, waterfall, highland pasture …; each type is its own node type). A protected area spanning several provinces sits in its main province's tree and is linked to its other provinces with a pale OTHER PROVINCE line. Views filter by type; Protected Areas also has Designated in the last 5 years and Highland Pastures and Mountains has Highland Pasture Inventory. A site card shows its type, province and district (Provinces when there are several), designation date, area, elevation, nearest administrative unit and links to the DKMP list page, Wikipedia, Wikidata, OpenStreetMap and the administrative structure; highland pastures from the Türkiye-wide Highland Pasture Inventory (Ministry of Interior, 2017) also show the distance to the district centre, area, road condition and period of use, with the volume and page in the source line and a link to the inventory PDF at that page (a copy of the source document is kept on wide.tr). Protected areas come from the official DKMP lists and are complete; highland pastures come mostly from the inventory; the other groups are limited to what is recorded in Wikidata, Wikipedia and OpenStreetMap. On a province, district or village (neighbourhood, town) node's card, the Natural sites section below Museums and sites lists the natural sites in that place the same way: protected areas on the official DKMP list first, then by name; each row shows the type and, where relevant, nearby (or, when the nearest settlement is more than 3 km away or the distance is unknown, e.g. nearest settlement (28 km)); a province card also lists protected areas extending into that province.

The "Former Place Names" showcase shows the name history of villages and neighbourhoods across three periods: 1928 (the Ministry of the Interior's book Köylerimizin Adları, with Ottoman and Latin spellings), 1968 (Ministry of the Interior, Köylerimiz) and today. The main graph shows the provinces on a map, coloured by the share of their villages with a known former name; each province's own graph shows province → district → place, red for a changed name, green for the same name since 1928 and grey for no known history (a former name can be added from the card), with a timeline on the card. The Journey of the Districts shows in three columns which province the 391 districts of 1928 are in today (1928 provinces on the left, districts in the middle, today's provinces on the right); an orange district is in another province today, and provinces created later (plates 68–81: Düzce was a district of Bolu in 1928) have their own colour. 1928 Administrative Divisions shows the 63 provinces and their districts; each province's own graph shows district → sub-district → village, with the village card giving the Ottoman and 1928 Latin spellings, the right PDF page of the book and the place today; the Uncertain readings view shows rows whose Ottoman and Latin machine readings disagree (the book page has the original). If you know the correct reading, propose it with Correct the reading on the village card — even when the village is not matched to a place today: the form shows the line as printed in the book (a three-line excerpt of the page, the line to correct marked with a red arrow) with an In the book, p. N link below; an approved correction appears in the 1928 graphs (and, for a matched village, in its timeline). The book was read by machine: readings may contain errors, and a village is linked to a place today only when its name is unambiguous in that area. The Name parts (interpretation) line on cards splits a name into its roots and suffixes with short meanings (Akpınar: ak “white” + pınar “spring”); it is an interpretation, not sourced information, and appears only when every part is known.

Contributing places: the Museums and archaeological sites and Natural sites sections are open to everyone's contributions (members with a verified e-mail; published after an administrator approves, while entries by administrators and trusted contributors go live directly). On a province, district or village card, Add a museum or site / Add a natural site below the section (or the Add information menu when the section is empty) adds a place missing there (a member's entry is a suggestion): group, type, name (operator for museums), an optional description and Wikidata ID. Below each row in the list are Correct (name, type, location — with a place search — or description) and Remove (with a reason); on a place's own card the same Correct and Remove buttons are in the contribution bar at the top of the card. Every suggestion needs a source. An approved contribution reaches the showcase within a few minutes: a new place joins its graph, a correction is applied, a removed place goes away; the place's card shows the contribution in a "Contribution:" line. Suggestions are reviewed in the same administration panel as community contributions. Where to contribute on a card: everything sits next to its own section. When a section has data, adding is below it (Add a former name under the timeline, Add a community under Communities, Add a museum or site under Museums); correcting is on the row itself: Wrong on a former-name row of the timeline (a removal suggestion with a reason), Correct the reading on the 1928 row (the form comes filled with the current reading), Correct / Remove on a place row. An empty section is not shown; the first entry goes through the Add information ▾ menu under the card header (Name history: former name, local name; Communities: community, tribe; Places: museum or site, natural site). Add its former name on the card of a place with no known former name (grey) opens the same form. When entering several items in a row, Save and add another clears the form and keeps the source; your recent sources are remembered in this browser and suggested in the source field. A place's own card has Correct and Remove in the bar, and every node of every showcase graph (institutions, communities, geographical indications, cuisines, museums, the Grand National Assembly…) has Report an error at the right of the bar. In an error report you choose where the error is (name, description, relations, properties, location or other) and write what is wrong and what is correct; a source is optional. Reports are not applied to the showcases automatically: a site administrator reads them on the Contributions page of the admin panel (type: Error report) and fixes the data; a trusted contributor's report also waits for review. Your report appears with its status in the card's Contributions section and under My contributions.

Place kinds and their colors:

• Province
• District
• Town
• Neighborhood
• Village
• Street

Organizations#

Add organization (edit toolbar, below Add place) adds a public organization or unit from the DETSIS registry to the graph as a real node: ministry, directorate, provincial office, school, hospital, municipality, court…

Type at least two letters or a DETSIS number in the search box (Search an organization or unit (name or DETSIS no)); several words find units containing all of them. Each row shows the unit type, its main organization and its province/district; move with the arrow keys and press Enter to add. The bottom line shows the data date (Source: DETSIS).

With Add with parent units on, the unit is added with its parents up to its main organization, linked with relations such as UNIT OF / AFFILIATED AGENCY OF. Also add its location adds the unit's district and province (as in Add place) and links them with LOCATED IN; it is disabled for units without location data. If a node is selected when you open the dialog, it is linked to the unit with the name typed in Relation (optional) (e.g. WORKS AT).

The same unit is never added twice: a unit already in the graph (including ones copied from the organization showcases) and its parents are reused. A single ⌘Z undoes everything that was added. Closed units do not appear in search.

Sharing and members#

Share in the top bar manages the graph's access, link, embed code and members; Graph settings in the edit toolbar opens the same dialog on its settings tab.

Access: Public (anyone can view, appears in search; requires a verified email), Members only (the card is visible to everyone; opening it requires a logged-in member with a verified email — others see the name, description and network shape; signing up from that page and verifying the email brings you back to the graph; node pages — name, type, description, properties and relations — are public and appear in search), With link (only people with the link) or Private (only you and members). If the graph has sub-graphs you manage, the Apply to sub-graphs box (checked by default) applies the chosen level to all of them as well.

Share the Link with Copy; Generate new link invalidates the old one. Embed gives code to show the graph read-only on another site (access must be public or with link). Copy link on a node card gives an address that opens that node. Addresses are readable: they are built from the graph's and the node's names (e.g. /graphs/duzce-cografi-isaretleri?node=duzce-81); a copied link does not carry the current zoom. When a graph is renamed its address changes too, and the old address redirects to the new one.

On the Members tab, Add member by email or username. Roles: owner, admin (manages access, members and settings), editor (edits) and viewer (sees a private graph). Addresses without an account receive an invitation; in Pending invitations you can Resend or cancel it. Remove takes a member out; Leave takes yourself out.

The Settings tab: name, description and Cover image. Without a cover image, the graph card shows the network's own shape (nodes in their type colours); an empty graph shows a coloured placeholder with its initial. In the danger zone, Delete graph content permanently deletes all nodes, edges and saved views but keeps types and settings; you must type the graph name to confirm (to only restore node positions, use Reset positions in the Layout panel). Delete graph cannot be undone.

Translations and graph language#

You write a graph in your own language; people who open the site in another language see the translation you enter. Any text without a translation appears in the graph's own language, so no field is ever empty. Switch the language with TR · EN in the top bar; the same page (with its open node) opens in the other language. Addresses follow the language too: with the site in English, graph and node addresses are built from the English names (/en/graphs/family-tree-of-the-prophets?node=adam); the Turkish address still works and redirects to the English one.

Graph language: a new graph starts in the interface language you create it in; change it under Share → Settings. Name, description, types and nodes are always written in this language.

The Share → Translations tab (owner, admin and editor) lists the graph's name, section and description; every node and edge type's name, the Inverse name of directed edges and property names, each in its own language on the left with a translation box on the right. An empty box shows visitors the automatic translation or the source text; Save writes the whole translation for that language.

In view mode everyone, including the people who can edit the graph, sees the chosen language; switching to edit mode shows names and types in the graph's own language, and that is what you edit (a translation is never written over the source text). The name and description boxes in Settings are always filled with the source text too. In the Nodes section you can search node names and translate them one by one (listed 50 at a time); in ready-made graphs most names come automatically from their sources.

In wide.tr's ready-made graphs, type, property and view names, descriptions and the names of people and institutions whose sources give an English name are translated automatically; sub-unit names in the state organization (directorates, schools, courts, mukhtar offices…) are rendered in English by translating their institutional words, while proper names such as villages, schools and people stay as they are; what you type in the box replaces it and is kept when the graph is regenerated.

Graph statistics and interest map#

See how much interest your own graph gets in the Share → Statistics tab. Only the graph's owner (and the site administrator) sees it; not even members who can edit the graph. Pick a period: Last 7, 30, 90, 365 days or All time. The cards on My graphs and on your own profile also show each graph's views in the last 30 days next to an eye icon (visible only to you).

Views are page loads of the graph (members and signed-out visitors shown separately); Average time is how long the page stayed on screen (the total below it). Counts of Node card opens, PNG downloads and Exports are given too. Hover the Daily views bars to read a day's count. Click a row in Most opened nodes (top 20) to select that node. Days turn over on Türkiye time; bots are not counted.

Show interest map colors nodes by how often their card was opened in the selected period: never opened in gray, few in light orange, many in dark orange (the scale is logarithmic, so a few popular nodes do not flatten the rest). While it is on, a chip at the bottom left of the canvas shows the period and the legend; click it to return to Statistics, × hides the map. The map is only on your screen, is not saved, and takes precedence over cluster colors.

All numbers are anonymous totals: who looked, IP addresses or cookies are neither shown nor stored (see Privacy).

History: who did what, and when#

The History tab in the graph dialog lists every change to the graph, newest first: who did what and when (adding, editing and deleting nodes and relations; settings, translation and view changes; imports, resets). People who can edit the graph (the owner, admin and editor members) and site administrators can see it.

Autosave sends each edit separately, so one person's edits to one graph within 30 minutes are grouped into one row; the row shows counts and examples of added, edited and deleted node names. Rows that only move nodes are marked m (minor edit) and can be removed with Hide minor edits; older records load with Older. History is kept from 7 October 2026.

Show diff on a row opens the details of the change: every added, deleted and changed node, relation and type; changed fields (name, description, type, image, properties, a relation's ends and weight) with the old value struck through and the new value next to it. Nodes that only moved are shown as a count. You can see all of your own changes under My graph changes on the My graphs page.

The graph's owner, graph admins and site administrators can use Undo on a row to undo that change (the whole grouped 30-minute session): added nodes and relations are removed, deleted and changed ones return to their previous state. If someone else touched the same records later, the undo is refused — undo the later change first. An undo is recorded in the history too and can itself be undone; imports, resets, settings and translation changes cannot be undone. Under Admin → Contributions, site administrators see Contributors (per person: edits, graphs edited, node counts and place/venue suggestions) and Recent changes (every graph change on the site, filterable by person, with undo), and on a member's page their Contributions. Deletion records of deleted graphs remain.

Account and privacy#

Your account consists of email, username, display name, short bio and profile picture, edited on the Settings page. What data is processed and how is explained on the Privacy page (/gizlilik), together with your rights under the KVKK and where to apply.

When you are logged in, the graphs and nodes you open, time spent on a graph, your exports and PNG downloads are recorded; site administrators can see these detailed records and they are deleted after 90 days. Your last login time is part of your account and is kept while the account is open. If you delete your account under Settings → Delete my account, your activity records are deleted too.

Every login to your account (sign-up and password reset included) records the IP address, browser, operating system and device type; Settings → Your recent logins lists your last 10 logins — if you don't recognise one, change your password. Login records are kept for 1 year and only site administrators can see them (the Logins section of a member's page in the admin panel, and the last login IP in the member list).

Views by visitors who are not logged in only go into anonymous daily counters; no IP address, cookie or fingerprint is stored for them. The only cookies are the strictly necessary session cookie created when you log in, the language preference cookie (NEXT_LOCALE) and Cloudflare's security cookie; interface preferences stay in the browser's local storage. For security the web server keeps access logs (IP address, time, requested address) for a limited period, and Cloudflare processes IP addresses while forwarding traffic.

Every member with a verified email can turn off emails about new comments in contribution discussions with Settings → Contribution discussion emails; emails about decisions on your suggestions arrive regardless. Site administrators can also turn the summary email of new contribution suggestions (at most once an hour) on or off with Contribution notifications; that checkbox is visible to administrators only.

A site administrator can open the site the way a member sees it with View as this member on that member's page in the admin panel, to review their graphs, private and shared content and settings. Viewing is read-only — nothing can be added, changed or deleted, and the administrator's views are not recorded as the member's activity; if a change is attempted it is not saved and the strip turns red with the change was not saved. While it lasts, a Viewing as @member strip stays at the bottom of the page; Stop viewing returns to the administrator's own view (it ends by itself after 2 hours at the latest). The start and end of every session are written to the server log. With End their sessions on the same page an administrator can sign the member out on all devices (the password does not change and they can sign in again); the box estimates from the last sign-in whether a session is open and shows when sessions were last closed.

On Admin → SEO, the View list links under the counters open the full lists of graphs and profiles that are or are not indexed (searchable by name, paged); every fixable issue has a full list too. The page also shows the graphs most viewed by signed-out visitors in the last 30 days and the non-indexed graphs that still get signed-out visits.

Copyright and terms of use#

The ready-made graphs on wide.tr are mostly compiled from publicly available sources (official records, open databases, encyclopedias). Data being public does not mean these graphs may be freely copied and reused: the graphs are licensed content, and careless reuse can raise copyright problems.

A graph is not the raw data itself. Which records were selected, how they are linked, how they are split into types and groups, the layouts and views, descriptions, translations and visual design are the product of work and creativity. This compilation is protected under Turkish Law No. 5846 on Intellectual and Artistic Works, which covers works and databases; unless stated otherwise, all rights are reserved.

The license of the source data also applies separately. Graph descriptions name their sources; for example rail systems come from OpenStreetMap (ODbL), street data in the Add place tool from an open compilation of NVİ records (GPL-3.0), and the Ottoman Dynasty from Wikidata (CC0). If you need raw data, take it from the original source under that source's license; the compiled version on wide.tr does not replace those licenses.

Allowed: linking to graphs and nodes, showing a graph on your own page with the embed code under Share, viewing for personal study and learning, and short quotations for criticism, news or teaching with attribution ("wide.tr" and the graph's address).

Needs permission: bulk copying of graphs or their contents (including scraping with automated tools), republishing them on another site, app or publication, using exported files and PNG images without attribution or for commercial purposes, and distributing data sets derived from the graphs. Doing these without written permission may infringe copyright.

Graphs created by users belong to the person who created them; ask the owner before reusing them. When you add data, text or images to your own graph, make sure you have the right to use them: whoever uploads someone else's content without permission is responsible for it.

For permission, licensing or to report an infringement, write to the contact address on the Privacy page (/gizlilik). This section is general information, not legal advice.

Shortcuts#

Shortcuts work on the graph screen; they are off while you type in a field and inside Cartography and the data table (except Esc). Macs use ⌘, other systems Ctrl; editing shortcuts need edit rights.

Mouse and touch: double-clicking empty space adds a node (long-press on a phone), Alt+drag connects two nodes, Shift+drag selects with a box, right-click (long-press a node on a phone) opens the menu; the wheel or two fingers zoom.

Frequently asked questions#

How do I enter two people with the same name? The editor lets you add two nodes with the same name, but import matches nodes by type and name: two rows with the same type and name in a file become one node. Give namesakes a distinguishing suffix (“Kerem Deneme (1950)”) or use an Id column in the file.

Is there a date type? No: properties are text, number or yes/no. Write dates as text in year-month-day form (1987-05-23) so they sort correctly; if you only know the year, use a number property for it.

What is weight for? Every edge has a weight (1 by default). Weight makes the line thicker, is shown in the label when Show weights is on, and is summed by the weighted size rules (Incoming weighted, Outgoing weighted).

What do visitors see? They open public graphs, and graphs whose link they have, in view mode; they search and read cards and views but cannot change anything. The layout and look they start with are the settings saved with Set as visitor default in Cartography.

Are graphs shown in other languages? Yes: every graph has a language, and people who open the site in another language see the names its editors entered in the Translations tab; text without a translation stays in the graph's own language. In wide.tr's ready-made graphs, types, property names, descriptions and views are in English; names of people and institutions come from the English of their sources (Wikidata, DETSİS, hand-compiled data), while place, party and line names stay as they are, being proper names.