Repository navigation
To test
Last updated: 2026-10-28 11:08 CET. Hand checks waiting for a person at the screen. An item is added when a fix is merged and removed when it has been checked; anything that fails goes back to its issue. "In build" is the first commit that contains the fix: the package you test must be that build or later.
How to help: anyone can take an item. Install the build named (packages from the latest
CI run on main, or build as in GETTING_STARTED.md), follow
"How to test", and comment on the linked issue: what you did, what you saw, operating system and version, and a
screenshot if something looks wrong.
| Check | How to test | What to look for | Issue | In build |
|---|---|---|---|---|
| Reconnect with radium off | radium switched off or hung; select its calculation(s), Run Mgmt > Reconnect Selected, then Cancel | Cancel returns at once; ECCE stays usable; no 'using the built-in ssh client' line after a timeout; stored results of the calculation are still there | 65ce64d7 | |
| Progress box: Incomplete reason | Select an out-of-memory calculation (01_Ti2Nb8O28Na3H2O_8m_d2_na_orca_optfreq) | Last line of the Progress box: 'Incomplete: Out of memory: the scheduler killed the job (OUT_OF_MEMORY)'; buttons do not overlap the Viewer row | 65ce64d7 | |
| Allocations: Remaining on submission | Launch a small job to Kebnekaise (hpc2n2026-351), select it | 'Remaining on submission: 74.6 % of 30 000 CPU-h/month (past 30 days)' (or similar) under Allocation | #613 | 65ce64d7 |
| Tools > Allocations | Organizer: Tools > Allocations... (logged in to Kebnekaise first), Refresh | One row per machine and account: e.g. '25.4 % of 30 000 CPU-h/month used (past 30 days)' for hpc2n2026-351; no 'not recognised' | #613 | 65ce64d7 |
| Details dialogs on a small screen | Calculation Editor, Gaussian 16: Theory Details; drag the window corner smaller | Window resizes; settings scroll; Close and Reset stay visible | 65ce64d7 | |
| No internal names as tooltips | Gaussian 16 Theory Details: hover over the checkboxes and fields | No tooltips like 'ES.Theory.UseCartesian'; real help tooltips still shown | 9a40cc9f | |
| Mode composition: Add Selected | Vibrational Frequencies, Composition...; select 3 atoms in the 3-D view (a bend), press Add Selected | 'a-b-c' with the middle atom bonded to both appears in the box and in the table | 5bb3e9e4 | |
| Argon at four temperatures | Delete Examples/General_Chemistry, File > Open Example Calculations > General Chemistry; Gases, liquids and solids: open ar-slab-40K, then Compare with 100K, 130K, 200K; play each trajectory | 40 K a crystal slab, 100 K a liquid slab with a few atoms in the vapour, 130 K a spread slab with much vapour, 200 K atoms filling the box; MSD flat at 40 K and rising with T | #626 | 590d1e13 (speed distributions; not in deb 87851b10) |
| Register Machines narrower | Tools > Register Machines, Machine tab | No Vendor/Model/Processor row; the window is narrower; saving keeps the machine working | — | 380ea423 |
| GROMACS launch messages | GROMACS Dynamics: Dynamics tab, constant pressure on, pressure relaxation time 0.5 ps; Run settings, barostat interval 100; time step 2 fs (the default). Launch | The launch check stops on grompp's warning; the message says 'Dynamics tab, pressure relaxation time (tau-p)' and 'Run settings, barostat interval (nstpcouple)' instead of bare .mdp keys, and suggests a larger relaxation time | #527 | 169c1982 |
| ORCA opt not converged | ORCA Opt with a small cycle limit (e.g. %geom MaxIter 3 end in the input), Launch | Ends Unsuccessful; summary reason quotes 'The optimization did not converge' | #614 | d1c18d47 |
| Request lines per queue | Register Machines, Kebnekaise, Queues tab: put #SBATCH --constraint=zen4 in Request lines, Update Queue, Save; Launcher on that queue: job script preview; launch a short job |
The line is in the preview and in the submitted script; the job starts on a zen4 node. A Slurm machine without the memory option (KebnekaiseZ4) shows Memory in the Launcher | #624 | 8024cb5a |
| scancel on the cluster | Launch to Kebnekaise, then scancel the job on the cluster while it is pending | Within a minute or so the calculation shows Killed, reason 'cancelled on …'; Reconnect not needed | #636 | ef9daa18 |
| macOS GROMACS examples | Mac with a dmg from 85c44ab5 or later: delete Examples/General_Chemistry, reopen the set, open ar-liquid | Atoms drawn, Geometry Trace and MD analysis present | #630 | 85c44ab5 |
| Check | How to test | What to look for | Issue | In build | | Density Map zoom | Viewer, any calculation with orbitals (e.g. a larger ORCA run): Density Map, scroll the mouse wheel over the map several times quickly, then stop | No 'Evaluating...' dialog; while scrolling the line under the map says 'Coarse grid while zooming'; shortly after you stop the map sharpens, with 'Full grid: N %' in that line on a large molecule | #645 | b8de276b (not in a package yet) | | ORCA integration grid | New ORCA calculation, DFT; Theory Details: Integration Grid, choose DefGrid3; Save, then Final Edit (or Verify) | The input's '!' line has DEFGRID3 next to the functional; with Default there is none; closing and reopening the editor keeps DefGrid3 | #644 | 8024634a (not in a package yet) |
| Hand-edited input kept | Any new calculation: Save; Final Edit, change a line of the input and save it there; then change a setting in the editor and Save | A question 'The input file was changed by hand' with Keep my input (highlighted, Return chooses it), Regenerate and Cancel; Keep leaves your edited line in the input (Final Edit shows it); Regenerate replaces it | #622 | e7cdbcf3 (not in a package yet) |
|---|
| Check | How to test | What to look for | Issue | In build |
|---|