[Trivial Psychic]

I was wanting to attempt to make a species_defs table file, but unfortunately, the Wiki has no data except that it exists and what its for.  Could someone in the know, add data regarding the parsing, for it?  What entries need to be included, what entries can be included, and perhaps in this thread, what pitfalls may undermine my efforts.

Thanks.

Later!

[phreak]

talk to woomeister.  he's done some stuff with it for inferno.

[Goober5000]

Or check out the default table embedded within the code:
http://fs2source.warpcore.org/cgi-bin/cvsweb/cvsweb.cgi/fs2_open/code/species_defs/species_defs.cpp?annotate=1.15

[Trivial Psychic]

This should help, although I'm hoping that there are more options available than simply the default.  I think I may talk to Kazan regarding this.

Thanks.

[Trivial Psychic]

What's the maximum number of species supported, and why are the error message warning me of subsystem and weapons data in the icons table file, when no such entries are there?

[Goober5000]

A maximum of 8 species is supported.  The default table is the general format; anything after a : can be specified.

What error messages do you mean?

[Trivial Psychic]

Switched to 8 species (down from 13... whew!), and while the error messages are still there, they've changed.

Error 1 (relevant data)

Error: icons.tbl(468):
Error: Required token = [$Name:], found [#End] .

File:C:\projects\fs2_open\code\Parse\PARSELO.CPP
Line: 543
[This filename points to the location of a file on the computer that built this executable]

Error 2

Error: icons.tbl(468):
Error: Required token = [$Name:], found [e: Oops 1] .

File:C:\projects\fs2_open\code\Parse\PARSELO.CPP
Line: 543
[This filename points to the location of a file on the computer that built this executable]

Error 3

Error: icons.tbl(469):
Error: Required token = [$Name:], found [$Message: Hey! What the hell are] .

File:C:\projects\fs2_open\code\Parse\PARSELO.CPP
Line: 543
[This filename points to the location of a file on the computer that built this executable]

Error 4

Error: icons.tbl(470):
Error: Required token = [$Name:], found [+Persona: Wingman 2] .

File:C:\projects\fs2_open\code\Parse\PARSELO.CPP
Line: 543
[This filename points to the location of a file on the computer that built this executable]

It continues on like this, then gives:

Error: Unable to parse icons.tbl!  Code = 1.

File:C:\projects\fs2_open\code\Mission\MissionBriefCommon.cpp
Line: 881
[This filename points to the location of a file on the computer that built this executable]

... then another stream of errors, and onother "unable to pare Icons.tbl", and so on.  The stupid thing is, that the "ERRORS" its finding, are from the messages table, starting with the 2nd in the why-are-you-shooting-me category and going down from there.

I tried "OK"-ing all the error messages, hoping to clear through them, but then it switches over lines in the ships table file, again... saying that they're problems in the icons.tbl.  Something's seriously screwed up with the code there.

Also, using a debug build gives the following "Assertion Failed!" error:

Assert: True_NumSpecies < MAX_SPECIES
File: C:\Languages\Visual Studio Projects\Visual C++\fs2_open-CVS\code\species_defs\species_defs.cpp
Line: 206
[This filename points to the location of a file on the computer that built this executable]

If I'm reading that right, then it seems that I'm exceeding the maximum number of species, when in fact I've got it down to 8.  Could the code be either A. still set to a lower number (like 3), or B. automatically generating terran, vasudan, and shivan species, if they're not in the species list, and thus exceeding the limits?  Anyway, FRED Debug crashes after than error message, but it's still kinda old for a debug FRED, (07-15b).  I am using a newer release build of FRED, but no debug was included with its release.

[Trivial Psychic]

OK.  I kinda overcame these problems, but I'm still not there.  The error was that I needed 8 icon entries in each category in the icons table file.  It was expecting more entries than were there, so it continued on to another table file and started looking for the entries there, thinking that it was still looking in the icons.tbl.  Fixing this up stopped the error messages, but now it simply crashes without anything beyond the this-program-is-crashing-do-you-want-us-to-report-it-to-Microsoft? message.  The debug build still gives the same error though.  I guess I'm kinda stuck for now... unless someone can shed some light on my problems.

[Goober5000]

Hm.  That Assert should be <=, not <.  That needs to be fixed.  We'll also need to add some error messages for parsing icons.tbl.

Try it with 7 species and see what happens.  We might have an off-by-one error somewhere.

[Trivial Psychic]

I'll give it a whirl with 7 and see what happens.  Thanks.

[Trivial Psychic]

OK.  I've cut back to 7 species, altering the species_defs, sounds, ships, and icons tables.  FRED release 08-28 still crashes without any useful messages.  FRED debug 07-15b no longer gives the error messages about number of species, but it now gives this:

Assert: num_icons < MAX_BRIEF_ICONS
File: C:\Languages\Visual Studio Projects\Visual C++\fs2_open-CVS\code\Mission\MissionBriefCommon.cpp
Line: 889
[This filename points to the location of a file on the computer that built this executable]

The debug log indicates that it scans as far as the end of the entries for sentries, then stops and throws up this error message.  Its stopping short of the jumpnode and transport wing entries.  I've checked several times and I have confirmed that there are 7 entries in each category.  It would be useful to have a more recent debug build of FRED to ensure that the errors I'm encountering are not outdated.

[Goober5000]

Post a new bug on Mantis with the appropriate info, then upload the files you're working on.

You should have a new set of EXEs to test in a few days.

[Trivial Psychic]

Thanks.  Just to note, I haven't loaded this up in an FS2 build yet, as I haven't had the time, (should tomorrow).  I'll post the mantis report as soon as I have done this, just to be sure I'm giving the most up-to-date feedback.  Also, since I have to include the tables that I'm working on, it'll soon become common knowledge (to anyone who is willing to look), that I'm attempting to enable multi-species for TBP.  I know... guy who only has TBP Avatar, working with multi-species stuff... nothing obvious about that.

[Trivial Psychic]

Alright.  I finally tested this with FSO (redmenace's 09-17 build).  The release build gave no troubles, beyond the tendancy to crash after every mission (not encountered in standard config, same build).  Fortunately, in a campaign, this doesn't hinder the ability to move on to the next mission.  This occurs whether I'm using an older pilot, or a new one created under multi-species environment.  I tried in both 7 species and 8 species configurations, and experienced no (additional) problems in either.  In Debug mode however, the 8-species gave me the error about number of species, so that definitely needs to be fixed.  When I switched back to 7-species for debug, no errors about species number, no errors about icons, but it gave me the following for asteroids:

Assert: Num_asteroid_types == Total_asteroid_types
File: c:\fs2_open\code\asteroid\asteroid.cpp
Line: 2082
[This filename points to the location of a file on the computer that built this executable]

I assume I also have to alter the asteroids table to add debris for each new species.  I'll make that change, then see if the debug works.

Later!

[Trivial Psychic]

Triple post.

I just tested with the revised asteroid table.  In debug mode, it didn't crash... until I tried to load a mission... no useful error message.  However, errorlog.txt reported that the program crashed as a result of a "breakpoint", which I'm not familiar with.  Also, the last entries in the fs.log file, are:

Int3(): From C:\fs2_open\code\MissionUI\MissionShipChoice.cpp at line 3630

This was just after the first briefing icon was loaded.

With release builds, it will crash as soon as it attempts to load a second mission, or restart one in progress.  Only once did I manage to load a second mission, as part of a campaign.  The first one was a red-alert mission, while the 2nd one ended in a red-alert sexp.

In any case, I have enough data now to formulate a proper Mantis report.

[Edit]
4 Mantis bugs now logged.
FSO Debug - number of species error
FRED Debug - Icons error message
FRED Release - Crash on load
FSO Release - Crash loading 2nd mission
[/Edit]

[Kazan]

yikes lotsa bugs introduced into this feature over the years from the lack of people using the feature