[pSimon]

Not quite sure what the issue is, but I get a crash in the first mission, which is just as beams fire for the first time/fighter wings launch.

The debug build gives some errors concerning missing beams, so that's not too surprising.

As other people don't have this issue, I'm assuming the problem is at this end. I don't see this with other mods though.



[attachment DELETED!! by Strong Bad]

[Taranis]

What missing beams are you getting with the debug build? There should be an error message for the MRed which I never fixed since no ship in the campaign uses it. The beam fired in this case is a regular BVas. I do have a different checksum to you for some of the mediavps files so the issue could be there.

[tomimaki]

MRed is used in Walk-shp.tbm for SSD Bhaal. Check turret18 nad 19. 

Anyway this build crashes when Typhon jumps in and launches fighter wing: http://www.hard-light.net/forums/index.php?topic=91436.0
and this not: http://www.hard-light.net/forums/index.php?topic=91428.0
Can someone confirm this?

Further testing.
On linux I got this backtrace:

Code: [Select]

Program received signal SIGSEGV, Segmentation fault.
0x000000000060b50b in vm_vec_sub (dest=0x7fffffffd610, src0=0x4, src1=0x10)
    at math/vecmat.cpp:170
170     dest->xyz.x = src0->xyz.x - src1->xyz.x;
(gdb) bt
#0  0x000000000060b50b in vm_vec_sub (dest=0x7fffffffd610, src0=0x4, src1=0x10)
    at math/vecmat.cpp:170
#1  0x00000000006df615 in get_submodel_delta_angle (sii=0x0)
    at model/modelread.cpp:632
#2  0x00000000006ee66d in model_get_rotating_submodel_list (
    submodel_vector=0x7fffffffd750, objp=0x1377b28 <Objects+570216>)
    at model/modelread.cpp:4427
#3  0x00000000007a496e in ship_ship_check_collision (
    ship_ship_hit_info=0x7fffffffd970, hitpos=0x7fffffffd940)
    at object/collideshipship.cpp:259
#4  0x00000000007a7f17 in collide_ship_ship (pair=0x7fffffffda70)
    at object/collideshipship.cpp:1132
#5  0x00000000007ad472 in obj_collide_pair (A=0x1376e38 <Objects+566904>,
    B=0x1377b28 <Objects+570216>) at object/objcollide.cpp:1653
#6  0x00000000007ac228 in obj_find_overlap_colliders (
    overlap_list_out=0x7fffffffdb50, list=0x7fffffffdb70, axis=2, collide=true)
    at object/objcollide.cpp:1270
#7  0x00000000007ac01f in obj_sort_and_collide () at object/objcollide.cpp:1239
#8  0x00000000007b466e in obj_move_all (frametime=0,0279998779)
    at object/object.cpp:1514
#9  0x0000000000417257 in game_simulation_frame ()
    at freespace2/freespace.cpp:4013
#10 0x0000000000417e1f in game_frame (paused=false)
---Type <return> to continue, or q <return> to quit---
    at freespace2/freespace.cpp:4405
#11 0x0000000000418af7 in game_do_frame () at freespace2/freespace.cpp:4816
#12 0x000000000041ae01 in game_do_state (state=2)
    at freespace2/freespace.cpp:6507
#13 0x00000000004e4b8c in gameseq_process_events ()
    at gamesequence/gamesequence.cpp:409
#14 0x000000000041bce1 in game_main (cmdline=0x1e2b0a0 "")
    at freespace2/freespace.cpp:7074
#15 0x000000000041bef3 in main (argc=1, argv=0x7fffffffdfa8)
    at freespace2/freespace.cpp:7209
And git bisect shows this:

Code: [Select]

975e108841e74f64ba43b1c7bbf8ba1109ec7ed2 is the first bad commit
commit 975e108841e74f64ba43b1c7bbf8ba1109ec7ed2
Author: Goober5000 <[email protected]>
Date:   Mon Jan 25 01:29:51 2016 -0500
 
    further accommodation of the dumb-rotation movement type, including, finally, proper collision detection!
So maybe it is bug in FSO.

[CT27]

I had this campaign (and Templar when I played that) crash on a recent build when a Typhon started launching fighters/bombers; though I was playing on Windows.  I was able to fix my issue at least by playing on an earlier build.

[niffiwan]

I saw an address sanitiser issue recently that IIRC was something to do with the new dumb/intrinsic rotation code and a wild memset, I was hunting a different bug at the time so didn't dig into it at all.  Needs more investigation, but it certainly sounds like an FSO bug so far.

[CT27]

I saw an address sanitiser issue recently that IIRC was something to do with the new dumb/intrinsic rotation code and a wild memset, I was hunting a different bug at the time so didn't dig into it at all.  Needs more investigation, but it certainly sounds like an FSO bug so far.

For what it's worth, some of the more recent nightly builds that I've had mysterious crashes on have had "3_7_5" in the file name.  The older build I'm using that works without the crashes had "3_7_3".

[niffiwan]

That's expected, once the 3.7.4 RC was released, the nightlies were switched to version 3.7.5

[CT27]

I meant that maybe there's something in these recent nightly builds that's causing the issue.

[niffiwan]

oh right, yeah, there's two separate bugs being investigated in recent nighties

This one is probably the cause of the issue in this campaign:
https://github.com/scp-fs2open/fs2open.github.com/issues/572

This doesn't seem to be directly related, but might cause random crashes:
https://github.com/scp-fs2open/fs2open.github.com/issues/574

[GTVA Qlurzach]

Hmm I like the nebula, but sadly before the mission progresses I encounter the same problem of crashing when the Typhon launches fighters in mission 1.

It'll be good to add some conversation between the rookie pilots at the start of mission 1 to give the scenario a more in-game feel (although so far I've only played till the crash occurs)

Otherwise great work Taranis thank you

[tomimaki]

GTVA Qlurzach, crash is fixed in recent nightlies after 2 April.
Try this build http://www.hard-light.net/forums/index.php?topic=91868.0

[GTVA Qlurzach]

This is totally awesome thank you tomimaki

is it just me though or are all the nightly build links dead? leads to an error 404

[niffiwan]

is it just me though or are all the nightly build links dead? leads to an error 404

It's not just you, there's been a mishap with the server hosting the builds, alternative arrangements are being investigated.

[Mito [PL]]

A simple, short, bug-free and pretty interesting minicampaign. Mentioning the balance - it was pretty easy (Medium difficulty), but not too easy. It's just good

[Goober5000]

I played this just now.  A fun, short, interesting campaign - but that last mission proved incredibly difficult.   All of my fighters, and myself, kept getting swarmed and wiped out no matter what I did.  I finally ended up flying a Myrmidon with a secret Helios payload and assigning the rest of Alpha to be Erinyes, which helped.  I still completed the mission as the only survivor, besides the "Phonecia".  Never did see the Bhaal.

[Mito [PL]]

IIRC, the only thing player has to do there is removing enemy beams facing friendlies, comm subsystem and then flying anti-bomber defence... I went in there to bomb the **** out of the Demon, but a Trebuchet spam (can't remember if available) might be sufficient...

The Bhaal arrives

Spoiler:

if you didn't destroy the comm subsystem.

[Taranis]

Thanks for playing Goober, glad you enjoyed it.

One thing to try out on M4 if you haven't is retasking beta wing to help kill the Karnatoth and deal with the comms subsystem yourself. The AI is horribly inefficient at destroying subsystems but they can help knock out the cruiser quite a bit earlier. Another thing to keep in mind is that your Alpha wing are equipped with trebuchets and can take out the second main beam on the destroyer in short order if you can keep them alive. 

I actually found M3 to be harder in testing with the default loadout but that's more due to the deficiencies of the Rhea.

[AndrewofDoom]

I feel stupid for not playing this last year. I totally missed it. But, remedying that I played it and it was actually pretty decent. Nothing groundbreaking but that's fine.

[Novachen]

Because i am uploaded this campaign on Knossos and fixed the Phonecia spelling error on this way. (and added the Phoencia nameplate also) i played through it too, in this process.

Well it is hardly believeable that a whole HoL destroyer can hid itself for nearly 20 years (since Operation Templar at least) but nonetheless i liked this mini story. Nothing spectacular, but not bad either.

I actually do not think, that the missions are too short in any way. Also i actually had the most problems in the second mission

Actually i had no problems with the difficulty in this campaign either. Even on Insane. But actually i had more problems in the second instead of the other missions. Actually i did not have much problems in the last mission with the Beam cannon and the comm system thanks to Maxim and/or Trebuchets you can use in this mission.
Even i had to say, that the bad end is much more interesting in this campaign than the good one . That bad end reminded me much of Deep Blood 2... unfortunately with the same cliffhanger without any continuation.

But as i said before, a good small campaign i can recommend to every player who is especially short on time.

[captain austin]

Because i am uploaded this campaign on Knossos and fixed the Phonecia spelling error on this way. (and added the Phoencia nameplate also) i played through it too, in this process.

Speaking of the Knossos upload; I think there is an error with that version of the mod file. I've never had any other problems with mods on Knossos in the past, but every time I try to download A Walk in the Sun it fails to finish extracting both the root and animation bundles before telling me the mod only installed partially. When I open the mod, AWitS doesn't appear in the campaign room. Anybody else running into this problem?