@@ -24,8 +24,7 @@ a project.
 4. Click **OK**.
 
 The project appears in the tree. Names may contain letters, digits, `.`,
-`_` and `-`. [TO CHECK: that a new user's home folder offers **New
-Project...**; the resource types allow a project only inside a project.]
+`_` and `-`.
 
 ![The Organizer with the new project](img/organizer-project.png)
 
@@ -60,10 +59,10 @@ library:
 1. In the **Mode Toolbar**, click the button with the tooltip **Import from
    Structure Library**. You can also choose **Mode > Add Structure**
    (Ctrl+7).
-2. In the **Structure Library** panel, open the folder `SimpleStructures`,
-   then the folder `Miscellaneous2`. Open a folder by double-clicking it.
-   [TO CHECK: the name the library is listed under in the **Libraries**
-   drop-down, and that folders are shown by these names.]
+2. In the **Structure Library** panel, choose `SimpleStructures` in the
+   **Libraries** drop-down, then open the folder `Miscellaneous2` by
+   double-clicking it. The button next to the drop-down goes up one
+   level.
 3. Click `h2o` in the list. A preview appears in the panel.
 4. Click in the empty 3D view. The molecule is added there.
 
@@ -76,7 +75,9 @@ If you prefer to draw the molecule yourself:
 4. Choose **Build > Add Hydrogen** (Ctrl+F) to complete the valence with
    hydrogen atoms.
 5. Choose **Build > Clean** (Ctrl+K) to tidy the geometry.
-   [TO CHECK: that **Add Hydrogen** gives two hydrogens on a bare oxygen.]
+
+The default shape of oxygen is Bent, with two open bonds, and **Add
+Hydrogen** fills each open bond with a hydrogen atom.
 
 Then save the structure:
 
@@ -116,8 +117,12 @@ options for the chosen theory and run type. The defaults are right for
 this calculation.
 
 To see the input file ECCE will send to NWChem, click **Final Edit...**.
-[TO CHECK: what **Final Edit...** opens and whether closing it is safe
-without changes.]
+The button is enabled when the setup is complete (the run state is
+**Ready**). It opens the input file in your text editor: the editor set in
+**Edit > Preferences**, otherwise the one named by the `VISUAL` or `EDITOR`
+variable. If you close the editor without saving, nothing changes. If you
+save a change, ECCE uses your edited file, and later changes in the
+Electronic Structure Editor ask before they replace it.
 
 ![The Electronic Structure Editor](img/calced-water.png)
 
@@ -131,8 +136,9 @@ without changes.]
 2. In the Launcher, choose `localhost` in **Machine:**.
 3. In **Run Directory:**, enter a folder in your home directory where the
    job may run, for example `/home/<you>/ecce-runs`. The field is marked
-   `*`, which means it is required. [TO CHECK: whether the field is
-   pre-filled and whether the folder is created if it does not exist.]
+   `*`, which means it is required. The field starts with the directory
+   you used last for this machine. The path must start with `/` or `~`.
+   ECCE creates the folder, with its parents, if it does not exist.
 4. Leave **Username:** empty. An empty user name, or your own, runs the job
    on this computer.
 5. Click **Launch**.
@@ -203,8 +209,8 @@ calculation:
    Rerun**. Or create a new calculation as in step 2.
 2. In the Electronic Structure Editor, choose `GeoVib` in **Runtype:**.
    `GeoVib` optimises the geometry and then computes the frequencies.
-   [TO CHECK: that **Duplicate for Rerun** is under **Edit** and keeps the
-   structure and settings.]
+   **Duplicate for Rerun** copies the calculation with its structure and
+   settings and resets it, ready to launch.
 3. Save and launch as before.
 4. In the Viewer, choose **Properties > Vibrational Frequencies**.
 5. Use the **Table** and **Graph** tabs to see the three frequencies.
 
@@ -25,8 +25,8 @@ Choose one of these:
   source tree.
 
 ECCE does not include a computational chemistry code. To run the example
-in this chapter you also need NWChem. [TO CHECK: whether the NWChem package
-name is the same on Rocky and Fedora.]
+in this chapter you also need NWChem. Its package is called `nwchem` on Debian, Ubuntu,
+Rocky and Fedora.
 
 ## Install on Debian or Ubuntu
 
@@ -93,8 +93,8 @@ Always start ECCE with `ecce`. Do not start `ecce-builder` or the other
 `ecce-<name>` commands directly: `ecce` first starts the services the
 windows need. The Organizer opens first. Open the other windows from it.
 
-[TO CHECK: whether the package installs a menu entry (`ecce.desktop`) that
-starts the same command.]
+The `ecce-client` package also adds an **ECCE** entry to the applications
+menu. It starts the same `ecce` command.
 
 ## What you see the first time
 
@@ -125,8 +125,8 @@ account on it:
 6. Click **OK**.
 
 If you tick **Save Passwords Between Invocations**, ECCE stores the
-password and does not ask again. [TO CHECK: the title of the login window;
-`GETTING_STARTED.md` calls it "ECCE Authentication".]
+password and does not ask again. The title of the login window is
+"ECCE Authentication".
 
 If the account does not exist, ECCE shows "There is no account for ... on
 this ECCE data server". Create the account with `ecce-dataserver-adduser`
@@ -149,8 +149,7 @@ To use it from the first start:
    ```
    ECCE_LOCAL_DATA=$HOME/.ECCE-local ecce
    ```
-2. The Organizer opens without a login window. [TO CHECK: that no login
-   window appears in local mode.]
+2. The Organizer opens without a login window.
 
 To make it permanent:
 
@@ -167,10 +166,8 @@ Changing mode copies nothing. The calculations of the other mode stay where
 they are, and the new mode starts empty. The variable `ECCE_LOCAL_DATA`
 overrides the setting in Preferences.
 
-[TO CHECK: whether the structure and basis-set libraries are available in
-local mode when only `ecce-client` is installed. They are shipped in
-`ecce-server`; until this is confirmed, install both packages on a single
-computer.]
+Local mode reads the structure and basis-set libraries from the
+`ecce-client` package. It does not need `ecce-server`.
 
 ![The Organizer after the first start](img/organizer-first-start.png)
 
 
@@ -44,9 +44,11 @@ What happens next depends on the format:
 
 - XYZ: ECCE asks for the units of the coordinates (the file does not
   record them): Ångströms, Bohr, picometers or nanometers.
-- PDB: if the file has several models, alternate locations or chains,
-  ECCE asks which to use. [TO CHECK: the labels of this window.] A single
-  model such as glycine loads without a question.
+- PDB: if the file has several models or alternate locations, the window
+  "ECCE PDB Reader" asks which to use: **Read which model:**, **Alternate
+  location:** and **Select chain:** (`All` reads every chain). A file with
+  a single model and no alternate locations, such as glycine, loads without
+  a question.
 - CAR and MVM: loaded without a question. A CAR file such as `benzene.car`
   (12 atoms) loads directly.
 
@@ -112,9 +114,13 @@ Output files of other codes are not recognised and the import ends with
 ESPRESSO outputs cannot be imported, although ECCE can run those codes.
 GAMESS-UK, Amica and MOLCAS are not recognised either. Importers for
 Gaussian 03 and 98 exist but were not tested in this version.
-[TO CHECK: the ORCA importer reads only ORCA outputs for the keyword
-vocabulary ECCE itself generates (RHF, UHF, RKS, UKS with Opt, Freq,
-EnGrad) and recovers a basis set only when it is given by name.]
+From an ORCA output ECCE always reads the molecule. It fills in the
+theory and run type only when the keyword line of the ORCA input (the line
+starting with `!`) uses keywords ECCE knows: HF, RHF, UHF, RKS or UKS, the
+functionals B3LYP, PBE0, PBE, BP86, BLYP, TPSS, M06L and M06, and the run
+types Opt, Freq, EnGrad and NMR. Other methods, such as MP2 or coupled
+cluster, are not decoded. The basis set is recovered only when the input
+gives it by library name, not in a `%basis` block.
 
 ### Import the file
 
@@ -203,9 +209,12 @@ each code supplies is listed in its parse specification
 
 Orbitals need the molecular orbital coefficients and the basis set in the
 output. If the output has no orbitals, **MOs** is missing from the menu.
-[TO CHECK: whether each code's output has them by default; NWChem
-needs the `ecce_print` output or the `movecs` file, ORCA's default output
-may not print them.]
+ECCE reads NWChem orbitals only from its own formatted output, which the
+`ecce_print` line in the input ECCE generates produces; a plain NWChem
+output file has none. An ORCA output has the coefficients only if the
+input asked for them with `Print[P_MOs] 1` in a `%output` block, which
+the input ECCE generates does; ORCA's default output lists only the
+orbital energies and occupations.
 
 #### Vibrations
 
 
@@ -52,9 +52,8 @@ launch uses the new path. A job fails with "Path for
 <code> not found" if the code you run has no line in this file.
 
 You can also set the path in the Machine Registration window, in the
-**Applications** box (see below). [TO CHECK: that a path entered there for
-`localhost` is written to `~/.ECCE/CONFIG.localhost` and is used for the
-run.]
+**Program** field on the **Codes** tab (see below). It is saved in
+`~/.ECCE/CONFIG.<name>`, the same file as above.
 
 ### A code that needs environment variables
 
@@ -89,72 +88,67 @@ define its queues. Open it in any of these ways:
 - From a terminal, run `ecce -machine`. This opens only Machine
   Registration, without a session.
 
-The list on the left shows the machines you have registered. The machines
-supplied with ECCE, `localhost` and `dummy`, are shared by everyone who
-uses the installation and are not in your list. If you select one of them
-elsewhere and choose to register it, the form shows its settings; edit them
-and click **Add/Change** to save your own copy under the same name. Your
-copy is used instead of the shared one. **Delete Machine** removes your
-copy and brings the shared one back.
+The list on the left shows every machine ECCE knows, with a column
+**From**: `site` for the machines supplied with the installation
+(`localhost` and `dummy` among them) and `yours` for the ones you added.
+Site machines are shared by everyone who uses the installation. If you
+select one and click **Save**, ECCE stores your own copy under the same
+name, and your copy is used instead of the shared one. **Delete Machine**
+removes your copy and brings the shared one back. The window's tabs are
+described in [Register Machines](register-machines.md).
 
 ![Machine Registration](img/machine-registration.png)
 
-<!-- capture: Machine Registration window with one cluster registered, Applications and Queues boxes visible -->
+<!-- capture: Machine Registration window, Machine tab, one cluster registered -->
 
 ### Register a cluster
 
 1. Open Machine Registration.
-2. Click **Clear Form**.
-3. In **Machine:**, enter the fully qualified host name, for example
-   `login.hpc.example.edu`.
-4. In **Name:**, enter a short name for the list, for example `hpc`.
-5. Fill in **Vendor:**, **Model:** and **Processor:** as you wish.
-   [TO CHECK: whether these three fields are required.]
-6. Set **Total # Processors:** and **# Nodes**.
-7. Tick **ssh**.
-8. In the **Applications** box, enter the full path to each code on that
-   machine, for example `NWChem:` `/opt/nwchem/bin/nwchem`. Leave the codes
-   you do not use empty.
-9. In the **Misc. Paths** box, enter the path to Perl in **Perl 5:** if it
-   is not the default.
-10. Define the queues (next section).
-11. Click **Add/Change**.
-12. Click **Close**.
+2. Click **New Machine**.
+3. On the **Machine** tab, enter the fully qualified host name in
+   **Machine**, for example `login.hpc.example.edu`.
+4. In **Name**, enter a short name for the list, for example `hpc`.
+   **Machine** and **Name** are required; **Vendor**, **Model** and
+   **Processor** are optional.
+5. Set **Processors** and **Nodes**.
+6. On the **Codes** tab, select each code you use and enter the full path
+   to it on that machine in **Program**, for example
+   `/opt/nwchem/bin/nwchem`. Leave the codes you do not use empty.
+7. Define the queues (next section).
+8. Click **Save**.
+9. Click **Close**.
 
 ECCE connects to ssh machines with its own ssh library. It asks for the
 password, the verification code or an unknown host key in dialogs, so you
 do not need a terminal. If the machine needs a two-factor login that ECCE
 cannot answer, use the `dummy` machine instead (see below).
 
-Code settings that are not paths, such as `module load` lines, go in
-`~/.ECCE/CONFIG.<name>`. `GETTING_STARTED.md` describes the file and the
-variables it can use.
+Code settings that are not paths, such as `module load` lines, go on the
+**Job script** and **Codes** tabs; see [Register
+Machines](register-machines.md). `GETTING_STARTED.md` describes the file
+and the variables it can use.
 
 ### Queues
 
 A queue is a batch queue on the cluster, with limits ECCE checks before a
-job is submitted. You define queues in the **Queues** box of Machine
-Registration.
+job is submitted. You define queues on the **Queues** tab.
 
-1. In **Queue Manager:**, choose the batch system of the cluster, for
+1. In **Queue manager**, choose the batch system of the cluster, for
    example Slurm or PBS. Choose Shell if the machine has none.
-2. Tick **Allocation Accounts Used** if the cluster charges jobs to an
+2. Tick **Allocation accounts used** if the cluster charges jobs to an
    account.
-3. In **Queue Name:**, enter the name of the queue.
-4. Set **Min Processors:** and **Max Processors:**.
-5. Set **Max Wall Time:** in minutes (`min`). 0 means no limit.
-6. Set **Max Memory:** in GB. 0 means no limit.
-7. Set **Min Scratch:** in MB.
-8. Click **Add/Change Queue**.
-9. Repeat from step 3 for each further queue.
-10. Click **Add/Change** to save the machine.
-
-To change a queue, choose it in **Queues:**, edit the fields and click
-**Add/Change Queue**. **Remove Queue** deletes the chosen queue and
-**Clear All Queues** deletes all of them.
-
-[TO CHECK: whether **Add/Change Queue** saves on its own or only when
-**Add/Change** is clicked for the machine.]
+3. In **Name**, enter the name of the queue.
+4. Set **min**, **max** and **default** for **Processors**, and **max**
+   and **default** for **Wall time** (hours), **Memory** (GB) and
+   **Scratch** (GB). 0 means no limit for a **max**.
+5. Click **Add Queue**.
+6. Repeat from step 3 for each further queue.
+7. Click **Save**.
+
+To change a queue, choose it in **Queues**, edit the fields and click
+**Update Queue**. **Remove Queue** deletes the chosen queue and **Remove
+All** deletes all of them. Queue changes stay in the list until you click
+**Save**; nothing is written before that.
 
 Your queue definitions are stored in `~/.ECCE/Queues` and in a file per
 machine, `~/.ECCE/<name>.Q`. They are added to the queues defined for the
@@ -176,8 +170,9 @@ script on your computer and stops.
 6. When the job has finished, copy the output file back.
 7. In the Organizer, choose **File > Import Calculation from Output
    File...** to bring the results into ECCE (see
-   [Looking at a file](looking-at-a-file.md)). [TO CHECK: the dialog that
-   follows and which calculation the output is attached to.]
+   [Looking at a file](looking-at-a-file.md)). ECCE creates a new
+   calculation from the output; it is not attached to the calculation you
+   launched on `dummy`.
 
 ## Job submission settings