18
0
mirror of https://github.com/MaSzyna-EU07/maszyna.git synced 2026-11-20 13:54:58 +01:00
Commit Graph

554 Commits

Author SHA1 Message Date
jerrrrycho
93e2b85d08 fix(simulation): the test mode gives nothing to the outside and leaves the player's files alone
The test mode turned off what comes in from outside, but a few settings
of the user's eu07.ini still reached out of a test run:

- feedbackmode: 1 and 2 toggle Caps Lock / Scroll Lock by injecting key
  events into the whole system, 3 and 4 drive an LPT port or a PoKeys
  desk (and 4 reads the desk's switches);
- multiplayer: events are reported to a multiplayer server, and some
  events are held back for it;
- execonexit: a program is run with system() at the end of each run,
  which also waits for it;
- showsystemconsole: a console of the run's own pops up and stdout is
  reopened on it, so whoever started the test with a redirected output
  doesn't get the `test:` line (and a console the run was started from
  would be hidden without it);
- imgui.ini in the game folder was written by every run, and by several
  at once with run.py -j.

All of these are off in test mode. Nothing changes outside it.
2026-11-08 09:36:51 +01:00
jerrrrycho
4356e7c3fd fix(simulation): a test run that can't start ends with the error code and says why
- `--test` is looked for before the rest of the command line is read. An
  unknown option before it (e.g. `-s x.scn --test-timout 90 --test`) used
  to end the run with exit code 0, which a caller reads as a pass.
- `--test-timeout` without a value, or with one that isn't a number of
  seconds, fails the run instead of leaving the default or, through atof,
  turning the timeout off.
- These, a missing -s, a failed start and an exception caught in main()
  end the test through test_manager::conclude(), which writes the usual
  `test: error (exit code 5) ... : <reason>` line to stdout and the log.
  Before, `eu07 --test` alone exited with 5 and nothing anywhere: the
  error went to the log queue and the process ended before the log
  thread wrote it. In test mode main() now stops the log thread and
  waits for it before _Exit().

Outside the test mode nothing changes: conclude() does nothing there.
2026-11-08 09:36:28 +01:00
7667e5b2ec fix: the bugs and vulnerabilities SonarCloud reports on PR #216 2026-11-07 22:11:08 +01:00
jerrrrycho
6dc42e976f feat: keep the vehicle odometers between the sessions (#72)
Every vehicle gets an odometer (TMoverParameters::Odometer) counting along
with the per-session DistCounter, starting from its state saved in
eu07_odometers.txt in the user's settings directory, else from the file a
python screen kept it in (kilometry/<name>_przebieg.txt), else from a state
made up from the vehicle name. The store is saved every minute of
simulation time and on closing.

The cab's distcounter: gauge shows the odometer; its .mmd offset seeds a
vehicle with no known state. Python screens get it as
state["odometer_total"]; state["odometer"] stays the session distance the
logbook and mirror scripts count differences of.
2026-11-07 22:08:42 +01:00
jerrrrycho
30bdc3da61 feat: radio fuse, a cut-off in the radio's power supply (eu07/maszyna#438)
TMoverParameters::RadioFuse, closed by default, stands for a fuse in the
cabinet, a lever on the desk or the switch on the radio's power supply.
RadioPowered() needs it closed, so with it open the radio's buttons, calls,
messages and the radio-stop stop working; opening it at a standstill resets
the radio-stop as switching the radio off does, and the radio and radio-stop
lamps go out. New commands radiofusetoggle/radiofuseenable/radiofusedisable
(no default keys), cab controls radiofuse_sw: (follows the fuse state),
radiofuseon_sw: and radiofuseoff_sw:, and radio_fuse for the Python screens.
The AI turning the radio back on closes the fuse too.

Existing vehicles don't change: the fuse starts closed and opens only from a
new cab control or a bound key. SVN data, optional: cabs that have such a
fuse or lever get one of the new controls in their .mmd, in the form of
radio_sw: (e.g. "radiofuse_sw: { <submodel> rot 0 0 0 }"); lang/*.po can
translate "radio fuse".

(eu07/maszyna#438)
2026-11-07 21:33:28 +01:00
jerrrrycho
e8c4f3592e fix: fpslimit paced in integer clock ticks, FPS shown counts the wait (eu07/maszyna#416)
The limiter slept off the rest of each frame with sleep_for() given a float
duration. MSVC adds that to steady_clock::now() in float nanoseconds, so after
a few hours of system uptime the wake-up time is rounded to tens of ms: at
95 h uptime fpslimit 60 had no effect at all and fpslimit 30 gave ~22 fps, in
the full and the legacy renderer alike.

frame_pacer now schedules frames on a fixed cadence in steady_clock ticks and
the loop waits with sleep_until; an overshooting wait is made up in the next
frame. The FPS shown and the motion blur use the new mainloop_period stopwatch
(frame start to frame start, wait included); the targetfps/fpsaverage
draw-range controllers keep reading the work time as before.

(eu07/maszyna#416)
2026-11-07 18:47:41 +01:00
jerrrrycho
dd019d88ba feat(simulation): the test mode takes no input from outside (eu07/maszyna#541)
With --test the gamepad, mouse picking, UART, ZeroMQ and motion telemetry are
off, the hidden window gets no key, mouse, scroll or char callbacks, and the
window procedure that takes WM_COPYDATA commands from other programs isn't
installed, so nothing but the scenery drives a test run.
2026-11-07 17:38:37 +01:00
jerrrrycho
441caeed64 refactor(simulation): tidy the test mode code
The test manager is reached through simulation::test() instead of a global,
the result cell is owned by a unique_ptr until the memory table takes it, the
cell value is compared with a margin, strings are built with std::format and
the --test settings moved to apply_test_mode_settings(). The driver's signal
check is folded into UpdateSpeedPosition(), which updates a speed table entry
and reports a signal at stop the head of the train just passed. The test.*
keys are listed with the other settings at the start of the log.
2026-11-07 17:12:11 +01:00
d4120ea5db Merge branch 'master' into feat/renderers-cleanup 2026-11-07 16:44:51 +01:00
jerrrrycho
02e29f6971 feat(simulation): test mode, --test runs a scenery unattended and exits with the result (eu07/maszyna#541)
eu07 --test -s <scenery> [--test-timeout <s>] runs the scenery in ghostview with
the null renderer (models loaded, no textures), a hidden window, no sound,
network or python, and in fixed steps of 0.03 s of simulation time per frame
(three 10 ms physics steps) as fast as it can. Without -seed it uses a fixed
seed, so two runs give byte-identical physicslogs.

The scenery reports its result to the __test.result memory cell, which the
simulator always creates: value 1 passes once every vehicle has stood still for
1 s, value 0 fails at once. The run ends with exit code
  0 pass, 1 fail,
  2 a train passed a signal at stop (SetVelocity 0 0, as seen in its driver's
    speed table; SetVelocity 0 x on automatic block doesn't count),
  3 timeout (test.timeout, default 3600 s of simulation time),
  4 standstill (every vehicle still for test.standstill s, default 300, with no
    result; a pass with test.pass_on_standstill yes),
  5 error (no -s, the scenery didn't load or has no vehicles, an exception).
The test.* keys go in the scenery's config block; test.fail_if_stop_overran no
only logs a signal passed at danger. Outside the test mode a signal passed at
danger is logged.

The null renderer gets a ui backend which draws nothing, hands out the null
texture and material instead of throwing, and ends the ui frame. A GL context
is still made for the hidden window, without asking for a version, as parts
of the ui create textures with gl calls of their own. No crash dialog or
message box in test mode; the window code no longer needs a monitor.
2026-11-07 16:20:42 +01:00
jerrrrycho
798fd89dd0 test: scenario tests that run the simulator and check a vehicle's physicslog
tests/scenarios/run.py builds a test scenery from a game install's scenery
(own trainsets removed, case fragment added), runs eu07 in ghostview with
physicslog on, and checks what the vehicle did: starts, max speed, brakes,
stops. First case: EU07 on td.scn driven by AI on the td/rozklad timetable,
with the Stacja1 exit route set by an event launcher instead of shift+4.

eu07 gets a -ini <file> option: an extra settings file read after eu07.ini,
so a test run can switch to a window and turn on physicslog without touching
the user's config.

Local only: scenery and vehicle data are not in the repo.
2026-11-07 16:20:41 +01:00
Sebastian Krzyszkowiak
605a94cf2b uilayer: Force ui viewports to be disabled under gamescope and Steam Big Picture (#232) 2026-11-07 12:33:46 +01:00
maj00r
088eeaa743 fix(editor): a scenery saved under another name keeps one file of turntable controls, not the old one beside its own (#251)
The editor writes the turntable controls on each save to <scenery>_obrotnice.scm and includes it. A copy of a
scenery, saved under another name, still included the file of the scenery it came from. Its save wrote a file of
its own and included it too, so each turntable had its events twice.

Now the save leaves out of the scenery the other files of turntable controls it includes, told by the line the
editor writes at their start. Their files stay on the disk, for the scenery they were made for. A file of
turntables made by hand, without that line, stays included.

Co-authored-by: maj00r <maj00r@host>
2026-11-07 12:32:02 +01:00
bcda4aaa42 Merge pull request #244 from MaSzyna-EU07/split/1-small-fixes
Split/1 small fixes
2026-11-07 11:19:45 +00:00
3f6a10f1fe Merge pull request #237 from MaSzyna-EU07/feat/track-uneven
Feat/track uneven
2026-11-07 11:18:49 +00:00
maj00r
a6778c9fe2 refactor: the Sonar remarks on the small fixes taken care of
- the log stops through a flag kept in a function, not a global one
- the messages are put together with std::format
- a vehicle dropped on a trainset in the launcher goes in once the
  trainset is drawn, so the vehicles are gone through with a plain loop
  again; the drop split into a new vehicle and a moved one
- the scenario that can't be opened is caught as the runtime error it
  is, a file name outside of the code page as the system error it is
- a track of an unknown type is left out through the pointer owning it
- the keyword of a track type written out by a function of its own
- the picked node forgotten with ranges::replace_if, the empty
  Forget_Node of the base renderer told why it is empty
- FileCased takes the path it works on
2026-11-06 19:54:54 +01:00
maj00r
b7ff8bab73 Merge pull request #236 from MaSzyna-EU07/fix/imgui-panels-no-auto-detach
fix(ui): keep panels in the main window when resizing them at the scr…
2026-11-04 01:08:30 +01:00
maj00r
ed0ff36a69 fix(app): a scenario failing as it loads goes back to the launcher too, the files of other cases looked for only off Windows
A scenario failing past its opening popped the loader mode and went on to pop it again and push the driver mode with the scenery half loaded; it goes back to the launcher as one that can't be opened. FileCased() goes through the folder only where the file system tells the case apart: on Windows a missing file was looked for in its whole folder for nothing, and a name outside the code page threw from the parser. The Discord thread sleeps a quarter of a second outside the driver mode, not to hold the closing up.
2026-11-03 21:53:52 +01:00
maj00r
e230d7b902 fix: a deleted model is no longer the picked node, an orthophoto tile isn't swept as a grey stub, the ED brake arrays freed
Deleting a model in the editor left the renderer holding it as the picked node, the next pick used freed memory: the renderer forgets it now (Forget_Node). A texture filled straight from memory (internalsrc:) can't be read again, released by the garbage sweep it came back grey; it stays now, and the units bound with the old name are rebound. The arrays of the ED brake force sharing are vectors, they leaked once per update. The scene hierarchy and the editor menu items go through the translation. Python caches ignored.
2026-11-03 21:45:58 +01:00
maj00r
e75d849679 fix(app): a scenery file named in another case than given is found, a scenario that can't be opened goes back to the launcher
The command line gives the name of the scenery lowercased, and the sceneries made on Windows name their files and includes in any case: on a case-sensitive filesystem they weren't found. FileCased() finds the file whose name differs only in case; the parser and the scenario loader use it. A scenario that can't be opened logs the reason and goes back to the launcher instead of ending with an exception.
2026-11-03 21:45:57 +01:00
maj00r
b34238f33f fix(app): the log written to the end of the closing, each step of it with its time; the Discord thread idle outside the driver mode
The log service stopped with applicationQuitOrder, so whatever was logged while the simulator closed was lost; it now runs until StopLogService() and writes out what came before. The closing logs each of its steps with the time since it began, to tell where a closing that hangs stops. Outside the driver mode the Discord thread spun a core in its loop, it now sleeps a second between the checks.
2026-11-03 21:45:57 +01:00
maj00r
a68c535be9 fix(track): the cab keeps the shake a track doesn't replace, and the editor redraws all that an edit changes
The side sway from the lateral acceleration now goes off only on a track with a sideways value, and the random jolts on one with a vertical or a sideways value. Dips at the rail joints and cross level alone replace neither. On a switch it goes by the path set and what its joined tracks allow.

An edit in the editor also rebuilds the other switches of a joined group with the tracks they meet, and the tracks of a rail run whose joints moved. Removing a track does the same. While dragging, the tracks further off wait until the drag ends.

The clatter of a bogie's axles was right: the followers stand at the pivots and the offsets are the places of the wheel sets. The misleading name goes.

Also what Sonar had left: a parameter as a span, a const method, a struct and its member apart, commented out code, and a function of the grading tool split, with the same output.
2026-11-01 20:12:33 +01:00
Sebastian Krzyszkowiak
823807201c Fix driver UI "mode windows" menu (#226)
Make the keyboard shortcuts match the actually used ones and check
the correct panels for is_open.
2026-11-01 13:40:57 +01:00
jerrrrycho
d995674bee fix(physics): automatic couplers don't catch above the safe collision… (#217)
* fix: automatic coupler crashes, horns, compressor start delay, radio and cab lever sounds

fix(cab): the line contactors lamp goes out with the pressure switch (eu07/maszyna#470)

The lamp of the open line contactors, which tells the start can be made, went out at 1 bar in the brake
cylinders, while the pressure switch cuts the start off only at 2 bar (or below 3.6 bar in the main pipe), so
with the brakes partly applied the start could be made with the lamp out. With a pressure switch fitted the lamp
now follows it; without one it goes out at 1 bar as before. The same holds for the lamp of the second unit, which
used to keep its last state between 1 bar and the cut-off. The condition is
TMoverParameters::LineContactorsLampLit().

Tests: tests/unit/test_line_contactors.cpp.

refactor: Sonar findings of PR 217

- horn_supply::pneumatic shadowed the pneumatic enumerator of the sound flags (cpp:S1117), it's horn_supply::air;
  the supplies read with using enum (S6177),
- CouplerForce(): the automatic coupling refused or caught without another level of nested ifs (S134), the
  same outcome: too fast, the coupling is locked; caught, coupled and locked; otherwise left to try again,
- CompressorCheck(): the power the compressor doesn't draw yet noted without a TODO (S1135),
- tests: the previous crash damage setting in an in-class initializer (S3230), the loops which stop early with
  break rather than in their conditions (S886), the stand-ins of test::pair in a std::vector rather than an array
  (S5945), units in the comments of the example without /* */ (S1103).

test: the cases the changes of the branch left without a test

- Cntrl.: BDelay1 to BDelay4 reach the brake delays (now through std::format), the default start types of the
  ground and converter overload relays of a multiple unit and of other vehicles (out of the nested conditionals),
- the compressor: fed from the low voltage it waits from the low voltage coming on; the pressure switch stops it
  above the cut-out pressure, without one the excess air is vented (the vented fraction rewritten),
- the horns: a vehicle without any air sounds as before, the threshold pressure, an electric horn on 110 V,
- the positions of a cab control below zero, and a control without position sounds.

refactor(cab): the kind of the train heating switch as a flag

Read once with the vehicle definition, rather than compared as text in each of the switch commands.

feat(cab): train heating switch springing back (MaSzyna-EU07/maszyna#85)

TrainHeating=[toggle]|impulse in the Switches: section of the vehicle definition makes the train heating switch
an unstable one: it goes up to switch the heating on, down to switch it off, and back to the middle once let go,
instead of staying where it was put. The position of such a switch isn't set from the heating state when the cab
is entered.

Tests: tests/unit/test_switches.cpp.

feat(sound): mirror sounds (MaSzyna-EU07/maszyna#84, eu07/maszyna#322)

The sound definitions of a vehicle (.mmd) take mirroropen: and mirrorclose:, played as a mirror sets off
unfolding or folding, and the locations section mirrors: <offset>... end, the offsets along the vehicle of the
mirror pairs, as for the doors (negative ahead of the centre), each pair at the sides of the vehicle. As the
animation, only the mirrors at the occupied cab sound. The start of a move is mirror_moves::motion in
vehicle/mirrors.h.

Tests: tests/unit/test_mirrors.cpp.

feat(cab): joint oil and fuel pump switch (MaSzyna-EU07/maszyna#81)

A cab control oilfuelpump_sw: switches both the oil and the fuel pump at once, with the new command
oilfuelpumptoggle (no key by default). As a two-state switch it turns both pumps off when both are on and both on
otherwise; as an impulse switch it's held to start them, like the separate pump switches. Its position follows
the pumps when the cab is entered.

refactor(editor): text fields filled without strncpy (MaSzyna-EU07/maszyna#112)

The last three strncpy calls Sonar flags (cpp:S5816), in the road and track editors, copy the text into the
ImGui field buffers with std::string::copy and terminate it right after; the text gets cut to the buffer as
before.

fix: automatic coupler crashes, horns, compressor start delay, radio and cab lever sounds

- automatic couplers don't catch above the safe collision speed (#86): with crash damage on they couple at a
  closing speed of 15 km/h at most, a faster hit locks the coupling until the vehicles part and the collision
  runs its course. On the way: CollisionVelocity() read a head-on 10 + 10 km/h as 14.1 km/h (quaternions turn
  by half the angle), it's now taken from the touching ends; CollisionDetect() took the other vehicle's speed
  the wrong way round for vehicles facing each other. Head-on hits now derail sooner, at the intended limit,
- a single horn lever can't sound both tones at once (eu07/maszyna#392),
- a skid ends once the vehicle stops on locked wheels (eu07/maszyna#479),
- electronic destination signs are blanked at the end of the timetable (eu07/maszyna#446),
- horns need the air or the voltage to sound (eu07/maszyna#290, #76): HornType=[Pneumatic]|Electric and
  Horn3Type in Cntrl.; a pneumatic horn needs 0.2 MPa in the main reservoir (or the main pipe without one),
  an electric one the low voltage; the lever stays where it was put,
- compressor start delay (#67): CompressorStartDelay [s] in Brake., counted from the compressor getting its
  power, after the engine starts say; none by default,
- the channel and volume buttons of a radio without power do nothing (eu07/maszyna#429); the power check of
  the radio is TMoverParameters::RadioPowered(),
- a lever moved smoothly sounds the positions it passes (#83).

Sonar findings in the functions changed: nesting, nested conditionals, commented-out code, std::format,
references to const. Unit tests: test::pair for two vehicles running into each other, test_couplers,
test_horn, test_wheelslip, test_compressor, test_radio, test_gauge_positions.

* fix: review findings of PR 217

- the pressures of the mover are in bar (MinCP=7.0 in the vehicle definitions), so the 0.2 the pneumatic horns
  needed was 0.2 bar; it's 2 bar now, and the tests take their pressures in bar too,
- the lamps of the ground relay and of the motor overload relays follow the pressure switch as the line
  contactors one does, TMoverParameters::MainCircuitLampsEnabled(), so the lamps of one cab go out together,
- the joint oil and fuel pump switch leaves out a pump started automatically, which ignores the switching:
  with one, it couldn't turn the other off, and showed off with both running (OilFuelPumpsEnabled()),
- the kind of the train heating switch comes from the occupied vehicle on entering the cab too, as in the
  switch commands; in a multiple unit these are different vehicles,
- CouplerForce() put the force taken back by the other vehicle on its far coupler when the vehicles faced each
  other (Couplers[1 - End] rather than the coupler it's coupled with),
- the velocity difference along the coupler axis computed once, CouplerVelocityDifference(), for CouplerForce()
  and CollisionVelocity(),
- the attachcoupler flag added without the SetFlag() call with 0, the index of the mirror moves noted as by side.

Tests: the coupler force of vehicles facing each other, the joint switch with an automatic pump, the lamps of
the main circuit, the horn and compressor pressures in bar.

* fix: coupler force of vehicles facing each other taken back along the other vehicle's axis

CForce runs along the axis of its own vehicle, so the force the other vehicle takes back on its coupler of the pair
is turned around once more when the vehicles face each other (-CF * DirPatch()), as that vehicle computes it itself.

Tests: the coupler forces of the pair facing each other and facing the same way.
2026-11-01 09:12:36 +01:00
maj00r
9364b03e1a fix(track): what Sonar reports on the unevenness
The blockers are in the EP brake of a vehicle, older than this work: its
arrays made with new[] are vectors now, which also removes the deletes.

- The caches of the unevenness of a track are no longer mutable: what fills
  them isn't const any more.
- TSegment gets back to RaPositionGet in one piece, with what the path has
  to do with the others in one struct.
- Less nesting in the group of switches and in the fields of the editor,
  nested conditionals written out, and the attribute is read from a
  string_view. A test no longer counts in floating point.

No change in behaviour.
2026-10-31 20:17:47 +01:00
1ad6d8b2bd Merge remote-tracking branch 'origin/feat/track-uneven' into feat/renderers-cleanup 2026-10-31 19:52:29 +01:00
154c42cb66 Merge remote-tracking branch 'origin/feat/track-uneven' into feat/renderers-cleanup 2026-10-31 19:13:23 +01:00
maj00r
2b8d7d7cf3 refactor(track): what goes with unevenness in one place, readable
The rules stay in TrackUneven.h with no dependence on the engine, as they can
be tested. Their use on the tracks moves out of Track.cpp to TrackUneven.cpp,
split into steps: the group of switches, the values of its paths, the meeting
of two ends. One set of helpers for the paths of a track takes the place of
the same conditions written out in several places, and the value at the ends
of a switch is worked out in one place. The clatter of an axle leaves the long
Update of a vehicle. The fields in the editor say themselves when they show.

No change in behaviour.
2026-10-31 19:05:56 +01:00
maj00r
ec2d226688 fix(editor): the uneven attribute is written in the layout of the entry it goes into
Its numbers go by position, and the positions differ between a plain track
and a switch. A switch made in the style of a plain track wrote them in the
layout of the plain one, so they were read back as other values: the up and
down part as the value of the path p3-p4, and so on.

The values are now one struct with pure functions to read, write and resolve
them, which replaces eight fields and which the unit tests can reach.
2026-10-31 18:58:20 +01:00
24b805b348 Merge remote-tracking branch 'origin/feat/track-uneven' into feat/renderers-cleanup 2026-10-31 18:53:48 +01:00
maj00r
02abc7df0a feat(track): all of the unevenness in one attribute, with values for each path of a switch
Usage, before endtrack, in mm, 0-100, separated by semicolons with no
spaces:

  a plain track:  uneven value;up;side;cross;dip
  a switch:       uneven p1-p2;p3-p4;up p1-p2;side p1-p2;up p3-p4;side p3-p4

  uneven 18           the value alone
  uneven 18;;;25;6    with its own cross level and dip at the joints
  uneven 0;;30        sideways alone
  uneven 14;35        a switch, 14 on the straight path, 35 on the diverging

What's left out takes the first number, the dip goes by the up and down
one, and a path p3-p4 with no numbers of its own takes those of p1-p2.
This replaces unevenup, unevenside, unevencross and jointdip. The fields
in the editor stay, track panel, section Ride and sound.

At the end they share, the paths of a switch keep to the lower value, and
each comes to its own value along the way. Joined switches take the lowest
of the same path, a path joined to another path the lower of the two, and
a plain track comes to the value of the path it meets. See
track_uneven_doc.md.
2026-10-31 18:33:35 +01:00
0bc29af650 Merge remote-tracking branch 'origin/feat/switch-fastenings' into feat/renderers-cleanup 2026-10-31 17:13:45 +01:00
maj00r
e5266e1e6b feat(track): track unevenness declared in the track entry
Usage, optional attributes of a track entry, before endtrack:

  uneven <mm>       peak deviation of a rail on a 10 m chord, 0-100; normal and switch
  unevenup <mm>     up and down alone, instead of uneven; normal and switch
  unevenside <mm>   sideways alone, instead of uneven; normal and switch
  unevencross <mm>  cross level alone, instead of uneven; normal only
  jointdip <mm>     depth of the dip at rail joints, every sound distance; normal only,
                    default 0.4 of the up and down component

Values of uneven follow PLK Id-14 appendix 1: 6 mm is the limit at 160 km/h, 10 at 120,
18 at 80, 35 at 40; half of it for a track in a good state. Example:

  uneven 18
  unevencross 25
  jointdip 6
  endtrack

The same fields are in the editor, track panel, section Ride and sound. See
track_uneven_doc.md.

The unevenness is a displacement field over the plan of the scenery, with no random
numbers: one place strays the same every time, and joined tracks, the paths of a switch,
sleepermodel and axles at one place move together. Tracks meet at a joint at the lower of
their values. Joined switches take the lowest of theirs, the tracks joined to them come to
it, and a track with none of a component keeps the switch to its design in it. The clatter
comes from the dipping joints. Curve radius and grade for the physics stay as designed.

Tracks with it get geometry every 2 m on straights: 4x the vertices on Baltyk with all
tracks flagged, 0.2 ms more per frame on td.scn.
2026-10-31 15:47:10 +01:00
maj00r
5fa03304ef fix(ui): keep panels in the main window when resizing them at the screen edge
Pulling a panel's grip into the corner of the screen pushed it a couple of
pixels past the main window, and ImGui gave it an OS window of its own. Now
a panel only comes out when dragged with the cursor outside the main window.

A panel that stays in the main window needs its grips in reach, so a resize
stops at the edge, and a panel left past the edge or too big for the window
is moved back and shrunk to fit.

This is a new io.ConfigViewportsNoAutoDetach option in our copy of ImGui, so
it has to be carried over when ImGui gets updated. ImGui itself has no option
for this, as of 1.93.1.
2026-10-31 01:47:40 +01:00
765153ad91 Merge remote-tracking branch 'origin/feat/switch-blades' into feat/renderers-cleanup 2026-10-29 23:55:04 +01:00
maj00r
80379b991c feat(editor): the work of the editor sessions not committed yet, as it stands: a curve traced on from the end of the diverging track of a switch goes on at its radius with no transition in, or from a straight right at the switch (alignment start_held, fit_going_on), the plan taking an end adjoining R 10 km or more for a straight; the switches and slips laid into curves, slid along them and laid alone on one straight (frame_path, switch slide); the route plan reading curves going on from the tracks adjoining, the kept designs taken or forgotten; the orthophoto and terrain model settings, downloads and layers, the tracks onto the terrain; the curvature of a path in the plan (bezier::plan_curvature); the Polish texts of all of it. Not built as committed 2026-10-29 22:19:06 +01:00
maj00r
2eda26b94c feat(editor): a warning of the sudden change of the cant deficiency right behind a switch, as ST-T1-A6 sees it (9.11): a track laid on from the end of the diverging track of a switch, or into it, is checked by switch_curve::check_exit from the end of the curve of the switch, a straight frog's length before its end, to the first sudden change of the curvature of the track laid. Two of them nearer than the intermediate element of table 9.16 (6 m in track connections up to 60 km/h) are taken together, else each one alone, against table 9.15 (130 mm up to 60 km/h), at the speed of the laying or of the diverging track of the switch, the lower one. The view tells of it by the cursor while the track is extended or laid, the status once it's laid; a transition out of the frog straight makes no sudden change. Selftest: from a Rz S42-190-1:7,5 with a frog straight of 1.66 m a curve the other way of R 200 m warned of 194 mm together, the same way none; from a Rz 60E1-190-1:9 with 6.09 m a curve the other way of R 200 m none, of R 120 m warned of 157 mm 2026-10-29 22:16:19 +01:00
55372571b3 Merge branch 'master' into feat/renderers-cleanup 2026-10-29 20:42:39 +01:00
maj00r
b0918628b4 fix(editor): a curve traced from the end of the diverging track of a switch with one point clicked goes on with the curve of the switch as far as abreast of the point, seen from its centre, and ends on it there, instead of turning into a straight where the straight to the point left it: the curve is held to where it was clicked, a point inside it as well. Selftest: one point after P4 of a Rz 60E1-300-1:9 laid R 300 m on to 238.36 13.31 for a point at 240 8, to 243.31 14.88 for one at 240 25, the curvature at P4 going on within 0.5 % 2026-10-27 19:41:50 +01:00
maj00r
667e433af9 feat(editor): a slip is selected, slid and deleted whole: any track of it picked shows all of it selected, editor_mode::slip_of finding its switches by its record, the tracks between them and its four ends. Dragged along, the slip moves along its first track with the straights out to the joints of its template and its crossing track, as far as that one goes on into tracks with their ends free, the straights of its first track before it and past it lengthened and shortened, a metre of them kept; one bent into a curve isn't slid, nor one whose crossing track is joined on beyond it. Deleted, it leaves its two tracks crossing from end to end in its place, kept in the scenery as a crossing Skrz, as a switch leaves its main track; Shift+Del takes all of it, Ctrl+Z brings it back. Selftest: a Rkp laid with its crossing track slid 5 m on and 3 m back, the joints of its template with it, walked through; deleted into a crossing Skrz 1:9 walked through, back with Ctrl+Z; a Rkpd at a crossing of straights with free ends slid 2 m, deleted into a crossing 2026-10-27 19:40:56 +01:00
maj00r
7b11db566e feat(editor): a slip Rkp or Rkpd bent into a curve of a side track: dragged on a curve with a slip armed it's laid out on the tangent to the curve where it was pressed and bent onto the curve as a switch laid into one is, editor_mode::frame_path laying a path of a template into the curve for both, its crossing track coming along with its ends free, ahead to the side of the drag, the slip curve of a single slip on that side, the straights of both tracks out to the joints of its template. A crossing goes only where two straights cross, a curve as sharp as the slip curves is refused. As it's dragged the view tells the radii of its tracks and slip curves, its cant and what switch_curve::check_slip says, the 11.4 pkt 4b of ST-T1-A6 among it; laid, the status the same. A slip bent into a curve isn't replaced by one laid out of straights. Selftests switchcurve, bladetips: a Rkpd 49E1-190-1:9 in R 800 m with its slip curves of R 154 and 249 m and the warnings of 11.4 pkt 4b and table 9.2, one of R 300 m in R 800 m with no warning of table 9.2, a Rkp in R 4000 m, the blades from 1.248 m on all of them; refused in R 150 m, a Skrz refused 2026-10-27 19:39:38 +01:00
maj00r
7f6a8736ca feat(editor): a slip goes only where the straights cross at the skos of its template: a Rkpd 1:9 was laid at a crossing of 1:7 or 1:14 as well, shaped for the angle of the tracks there into no slip of the catalogue. The straights crossing at a skos more than 0.14 deg off it, half of what keeps 1:9.51 and 1:10 apart, are refused with the skos they cross at and the slip of the same kind made for it, of the same rail and radius first, or none; a double slip replaced by one of another skos likewise. The panel and the hints say the slip goes at its skos. Selftest: a Rkpd 49E1-190-1:9 laid at 1:9 and 1:9.08, refused at 1:7 with no template for it and at 1:9.51 with Rkpd S42-230-1:9,51 fitting 2026-10-27 19:35:01 +01:00
maj00r
f8a12107a0 feat(editor): a switch laid with no track there, and no switch laid by a click: a press on the ground with a switch armed lays the switch there, its start where it was pressed, the direction of its main track taken from the drag once it's 4 m long and held, the drag sideways then giving the side of the diverging track, back by the start the direction free again; pressed by a free end within 5 m it goes on from that end in its direction and grade, joined to it. Its ends are left free to lay on from, styled after the track it goes on from or as the lay tool lays its tracks (editor_track::create_switch with a path_style). A press let go within 10 pixels places no switch, a slip at a crossing included, which was laid at once on the press: it's laid once dragged, its straights shown meanwhile. The hints and the panel tell of both. Selftest switchfree: Rz 60E1-300-1:9 on the ground with its four ends free, from the free end of a straight joined and walked through; a click on a straight laid none, a drag laid z1 2026-10-27 19:33:23 +01:00
maj00r
1340017be3 feat(editor): a slip keeps in the scenery the radius of the main track it's bent with, as a switch does, the lists of the switches telling it as Rg: switch_store::main_radius_of gives it from the first path of a switch, classify takes it from there too; a slip laid on straights keeps 0 2026-10-27 19:23:31 +01:00
maj00r
476909581f fix(track): two switches laid right after one another, with no plain track between them, are joined, the vehicles no longer derail there: only the plain tracks looked for their neighbours, so the end of the main track of one switch at the start of the next one was joined to nothing, and an axle getting to it ran onto a track made to derail it. TTrack::join_switches joins the free ends of a switch to the ends of other switches in the same points, on both sides, with the codes of the ends the axles go on by: the end of either path of one at the start of the other, two facing at their starts, two meeting at their ends; done once the plain tracks are joined as the scenery is loaded, the old sceneries too, and as the editor joins a switch laid. Selftest pathwalk walks the tracks as an axle does: P2 to P1, P4 to P1, start to start and end to end go through either way 2026-10-26 20:40:51 +01:00
maj00r
bfb045f2ad fix(editor): traced in the mode of a curve from the end of the diverging track of a switch, the line goes on with the curve of the switch: with a single point clicked the ghost was a straight right from P4, laid so with Enter, and points clicked on the photo a little off the curve gave way to the line from a straight there. One point after P4 lays the curve of the switch going on as far as the straight from it heads for the point; a curve going on from P4 which turns its way stays however loosely the points keep to it, the line from a straight taken only where they turn the other way. Selftest: one point after P4 laid R 300.9 going on, the curvature unbroken at the joint, where it was a straight before 2026-10-26 20:40:44 +01:00
maj00r
ef592e6479 feat(editor): the drive of a switch stands by the tips of its blades, not at the joint before them, and goes with the switch deleted. A drive placed, changed or carried along with its switch is set where the blades begin, as they're shaped or where the diverging track is broken at the tip of a planed blade, along the main track there, its heading and its tilt taken there; the drives the earlier versions of the editor set at the joints are moved there once, as the scenery is opened and the templates are known, with no step of the undo history, to be saved with it. Deleting a switch, a double switch or a set of tracks takes the drives of the switches along in the same step, Ctrl+Z brings them back. The joint before the blades and the tips of the blades of the switch being placed or the one selected are marked across the track in the view, so the tips seen on the photo aren't taken for the start of the switch. Selftests bladetips, drivetips; on a copy of test42 the drive of z1 went from the joint to 1.248 m, deleted with z1, back with Ctrl+Z 2026-10-26 20:40:37 +01:00
e19c788f30 Catenary as plain node entries in scenery files 2026-10-26 20:23:26 +01:00
d794b1e736 Catenary editor: keep catenary parts out of node editing, pick and drag anchor poles 2026-10-26 19:51:15 +01:00
maj00r
aed2cfad2c feat(editor): a track deleted cuts the design kept for its line down to the tracks left along it, in the same step of the undo history: a design keeps the stretch of its line it stands for (//$h kept <from> <to>), a track deleted in the middle parts it in two, one at an end shortens it, a design with no track left is forgotten. The geometry stays as designed, only its reach changes. A design cut down isn't taken whole by the route plan nor merged with a new one nor carried along by a switch slid, and gives no designed curve to the tracks left of a curve it cut, which are recognized as they lie: a track laid by hand into the gap is never pulled back to the old shape. The status tells of it. The //$h click lines are the editor's own now: read back with their design, and written once on save instead of staying behind and piling up. Selftests on a copy of wgr.scn: a track of a compound curve deleted in the middle, undone and redone, one at an end, all of a piece, saved and read back 2026-10-26 19:50:35 +01:00