Since older builds of FRED parse the mission version as a float, the version can contain a maximum of one decimal point. These changes allow FRED and FSO to disregard the build number when loading, saving, or comparing versions.
There is also a small fix to qtfred's saving of the author field.
* enhance handling of motion debris
* dunno what i was thinking here
* address feedback
* require all four bitmaps to be defined in new parsing method
* address feedback
This finally enables the `$Version:` field in FreeSpace missions, which has been present ever since the beginning of FS1's development but has never been used for its intended purpose. This field now stores the version of FSO that is required to run the mission. If the version of the engine is not sufficient to run the version of the mission, then a) the mission is hidden from the mission listing, or b) a friendly notification is displayed if the mission is loaded anyway (for example, through a campaign). Furthermore, FRED will notify the user if the mission's version is migrated from any version other than the legacy 0.10.
It will be the responsibility of coders who modify the mission format to also update the `MISSION_VERSION` constant with the appropriate major/minor/build/revision of the format change.
PR #5058 made `count_incident_asteroids()` use `Asteroid_obj_list` instead of `obj_used_list`; a perfectly sensible change in and of itself. However, PR #4827 had previously made `asteroid_obj_list_init()` only be called if FRED wasn't running (presumably because of the iteration through `Asteroid_objs` to clear flags, which is wasted effort in FRED). The problem is that without `list_init()` being called on `Asteroid_obj_list`, its `prev` and `next` were `nullptr` instead of linking back to itself, which means the beginning of the list and the end of the list were no longer considered the same (the iterator's `begin()` calls `GET_FIRST(head)`, which is effectively `head->next`, while `end()` calls `END_OF_LIST(head)`, which is effectively just `head`). This meant that iterating over `list_range(Asteroid_obj_list)` was returning `nullptr` as the first element of the list, and trying to dereference that would cause a crash. This was still not a problem in and of itself, as long as nothing tried to iterate over `Asteroid_obj_list` while FRED was running, which theoretically shouldn't be necessary...
PR #3704 changed mission parsing so that when parsing a ship with a name that matches one of the `Asteroid_field.target_names`, it would call `asteroid_add_target()` to initialize the relevant data for the asteroid field to fling asteroids at that ship. `asteroid_add_target()` calls `count_incident_asteroids()`, which (since PR #5058) iterates over `Asteroid_obj_list`, causing a crash due to dereferencing a null pointer on any mission that defines `$Asteroid Targets:` (and then actually has such a target be parsed).
While this could be solved by unconditionally initializing `Asteroid_obj_list`, there's really no need to do anything with `Asteroid_obj_list` in FRED, so instead the `asteroid_add_target()` call has been gated behind a `!Fred_running` check, and `Assertion()`s have been added to make sure `Asteroid_obj_list` has been initialized at the beginning of any function that does anything with it (besides initialize it), to catch similar problems in the future.
Fixes#5092.
Follow-up to #5040. This extends the checks to most iterations over `obj_used_list` in FSO. There are a few exceptions such as `waypoint_add` and `waypoint_remove`, plus a few others. And none of the object iterations in FRED have the guard because FRED does not use the same object life cycle as FSO.
When iterating through `Ship_obj_list`, check the `Should_be_dead` flag before using the object. This fixes#4990 and adds defense-in-depth checks to guard against similar errors elsewhere.
I considered moving the `Ship_obj_list` removal from `obj_delete` to `ship_cleanup`, but decided it would be prudent to avoid fundamental changes to lifecycle management.
* set-asteroid-field sexp
* errors
* scope
* set-debris-field sexp
* Clarify sexp documentation
* num_debris_sizes
* allow any order and activate display name feature
* unify naming scheme
* modular -ast.tbm
* styling
* decouple sexp parsing and asteroid field creation methods
* remove this
* fix merge oversights
* Remove this check
* reorder
* use scp_string instead
* spacing
* Only setup display name if it's empty
* Use define instead of magic numbers
* Ensure asteroid types get the correct spawn weight
* Warn if spawn weight is set for asteroid type
* typo
* Correctly handle asteroid split types
* iiiiiiiiiiiiiiiiiint
* address feedback
Adds gravity! Applies on a per-mission basis, and weapons, ships, debris, and asteroids now have a configurable "gravity const" which multiplies the effect of gravity on them, by default everything except debris is 0 (meaning no effect even when the mission has gravity). AI and the lead indicator will also lead for ballistic trajectories as well, given a gravity-affected weapon. No FRED support or gravity direction control yet, to come in later PRs.
* Fix crashes with sssi
We need to init in both places to avoid purging goals before ssi is set.
* Make sure we adjust for the wingp->special_ship--
Make sure we adjust sssi on ship death
Load campaigns using the full path provided by the user, not just from the configured mod paths.
Also tweaks initialization based on the campaign having already been cleared in `mission_campaign_clear()`.
Parse objects have their path masks temporarily set to -1 during parsing, since at that point the mask represents an index, not a bitfield. In all other cases, the path mask defaults to 0.
Fixes#4873.
Gather all the mission initialization code into a common function for clarity and ease of expansion. Previously, the various parsing routines contained initialization as well as parsing. As a bonus, quite a few redundant initializations are streamlined.
When the top-level operator of a SEXP is not found, flag the error instead of crashing. This is helpful for missions that contain unsupported scripted SEXPs.
Also be sure to set the bad node in the event of a NONOP_ARGS syntax error.
* implement modular nebula.tbl
* Address feedback and skip > 32 poofs
* read/write nebula poofs by name
* qtfred changes
* variable name
* log -> warning
* nullptr and make sure to only set bitmap_name when it's actually defined
Some refactoring with asteroid fields:
1. Moves asteroid throwing targets into the Asteroid_field struct, in anticipation of potentially multiple fields in the future
2. Moves some initialization code into `asteroid_level_init`
3. Calls `asteroid_level_init` when a mission is cleared
4. Loads asteroid targets as part of the asteroid field update/init logic. This fixes a FRED bug reported in Discord where asteroid targets stay in the dialog when a new mission is loaded
5. Remove `Default_asteroid_throwing_behavior` flag as it is equivalent to checking that the target list is empty
Goober5000