[Perro Seco]

Hello. I'm trying to make a grenade, using the gamevar PIPEBOMB_CONTROL with the values 2 1, and the gamevar GRENADE_LIFETIME with X 1, but sometimes the grenade doesn't explode. I'm doing something wrong or it's a bug?

Thanks in advance.

[The Commander]

[Fox]

I tested and it seems to be true. But I couldn't figure out what's the problem.

[Perro Seco]

If I delete the gamevar GRENADE_LIFETIME it works fine, but it takes a long time to explode.

[gibfrag]

I think it's cool if it doesn't explode every single time.. It makes it more realistic

Next stop.. gun jams!

[Fox]

I believe I have found out what's the problem, it's possible for a Pipebomb to spawn with a negative htg_t 7 and is likely related to this section of the source:

                        actor[j].t_data[7]= ltime
                                            + mulscale(krand(),lv, 14)
                                            - lv;

[TerminX]

Should be fixed in r4019. I didn't test it.

[Perro Seco]

Sorry for reviving this old post, but what is r4019? And is there a way to fix the bug using only CON?

[Player Lin]

Perro Seco, on 27 December 2013 - 06:52 AM, said:

Sorry for reviving this old post, but what is r4019? And is there a way to fix the bug using only CON?

No, that looks like engine bug, fixed at eduke32 synthesis build r4019 and later.
Get a new latest one from there.
http://dukeworld.duk...ke32/synthesis/

But seems no one tested that fix so try it yourself.

[Perro Seco]

Thank you. Now all the grenades explode, but there's another bug; sometimes they explode immediately when I launch them and they kill me...

[Jimmy]

I regret nothing.

[Helixhorned]

Fox, on 17 August 2013 - 02:41 PM, said:

I believe I have found out what's the problem, it's possible for a Pipebomb to spawn with a negative htg_t 7 and is likely related to this section of the source:

                        actor[j].t_data[7]= ltime
                                            + mulscale(krand(),lv, 14)
                                            - lv;

Oh, sigh, that's clearly wrong. Here, 'ltime' holds the value of *_LIFETIME and 'lv' that of *_LIFETIME_VAR. The way it's now, the ultimate lifetime is in the range

LIFETIME - LIFETIME_VAR + [0 .. LIFETIME_VAR/4].

It's not clear what the author of that code intended; if he assumed that krand() returns a value in [0 .. 32767] (it's really [0 .. 65535]), it would make some sense: the range would then be

[LIFETIME - LIFETIME_VAR .. LIFETIME + LIFETIME_VAR]. (1)

Then again, "var(iance)" could be meant to mean "extent" (max - min), in which case the intent was to make the range

[LIFETIME - LIFETIME_VAR/2 .. LIFETIME + LIFETIME_VAR/2]. (2)

I think fixing the code for meaning (1), i.e. using mulscale15 is most sensible. It changes the meaning retroactively, sure, but personally, I don't care for EDuke-era mods or WW2GI/NAM, presumably the only CON code where the "variance" was used at all.

[Helixhorned]

Helixhorned, on 05 January 2014 - 08:41 AM, said:

The way it's now, the ultimate lifetime is in the range

LIFETIME - LIFETIME_VAR + [0 .. LIFETIME_VAR/4].

Oops, it should read:
LIFETIME - LIFETIME_VAR + [0 .. 4*LIFETIME_VAR].

Perro Seco: what are the values of GRENADE_LIFETIME and GRENADE_LIFETIME_VAR you use? After all, as the above shows, it's perfectly expected that they occasionally explode immediately if LIFETIME_VAR >= LIFETIME.

[Hendricks266]

I checked the EDuke 2.0 source code and the statement in question has not changed.

                                        {
                                                long lGrenadeLifetime=GetGameVar("GRENADE_LIFETIME", NAM_GRENADE_LIFETIME, -1, snum);
                                                long lGrenadeLifetimeVar=GetGameVar("GRENADE_LIFETIME_VAR", NAM_GRENADE_LIFETIME_VAR, -1, snum);
//sprintf(g_szBuf,"Lifetime=%ld Var=%ld snum=%d",lGrenadeLifetime, lGrenadeLifetimeVar, snum);
//AddLog(g_szBuf);
                                        // set timer.  blows up when at zero....
                                        sprite[j].extra=lGrenadeLifetime
                                                                + mulscale(krand(),lGrenadeLifetimeVar, 14)
                                                                - lGrenadeLifetimeVar;
                                        }

[Perro Seco]

Helixhorned, on 06 January 2014 - 01:05 AM, said:

Oops, it should read:
LIFETIME - LIFETIME_VAR + [0 .. 4*LIFETIME_VAR].

Perro Seco: what are the values of GRENADE_LIFETIME and GRENADE_LIFETIME_VAR you use? After all, as the above shows, it's perfectly expected that they occasionally explode immediately if LIFETIME_VAR >= LIFETIME.

Hey, thanks for your help, but when I saw the GRENADE_LIFETIME in Eduke32 Wiki I didn't saw the GRENADE_LIFETIME_VAR, so it hadn't any value and maybe it was the problem. Now GRENADE_LIFETIME has the value 30 1 and GRENADE_LIFETIME_VAR 1 1, and it works great.