FAangband 0.3.6

[thapper]

trap of protection?

While fighting abyss spiders in Nan Dungortheb they created traps all around and underneath me a couple of times. The trap underneath me seemed somehow to protect me from further attacks and I could calmly stay around attacking whatever came near. I also didn't trigger the trap by waiting (5) on it...

thapper

[Nick]

Was the port tested on real hardware?

I just retested it on my DS, and it has no problem. I deleted the .raw files, and it took about 10-20 seconds to initialise the arrays. I am using a CycloDSEvolution. It almost sounde like you are running out of storage on your card - have you checked that?

[Nick]

While fighting abyss spiders in Nan Dungortheb they created traps all around and underneath me a couple of times. The trap underneath me seemed somehow to protect me from further attacks and I could calmly stay around attacking whatever came near. I also didn't trigger the trap by waiting (5) on it...

This is very odd, and I can't replicate it. Traps should not be created underneath you. If you have a savefile where this is happening, I'd like to see it.

[Nick]

There is a trivial way to remove all inconsistencies about ID when selling to stores.

I think that having no selling (or selling for no money) would make the FA advanced races even harder - particularly for low stealth classes. Even if gold drops are increased, an Adan Paladin, for example, will have a hard time killing enough monsters to buy the consumables necessary to get started.

This is something I may revisit at some point, but I don't think I'm ready to take the plunge yet.

[thapper]

This is very odd, and I can't replicate it. Traps should not be created underneath you. If you have a savefile where this is happening, I'd like to see it.

It was a splinter file from the competition. That particular one, as all my other attempts so far, has already died. If it happens to the next one that make it so far I'll send you the savefile.

[PowerDiver]

I think that having no selling (or selling for no money) would make the FA advanced races even harder - particularly for low stealth classes. Even if gold drops are increased, an Adan Paladin, for example, will have a hard time killing enough monsters to buy the consumables necessary to get started.

Then start them with more money and/or consumables?

An alternative is to start them with several ?recall and give a recall point to an area safe enough for the early game.

I think the suggestions I made about safer paths might help if you decide to go in that direction.

I probably shouldn't be pressing this, since I find your approach to starts a bit dubious yet do not have any idea about a viable in-theme solution, but I simply can't help myself when someone posts a problem trivially solved by no selling.

[Nick]

I probably shouldn't be pressing this, since I find your approach to starts a bit dubious yet do not have any idea about a viable in-theme solution, but I simply can't help myself when someone posts a problem trivially solved by no selling.

I'm aiming for any race/class combination to be possible, but not to balance difficulty - I'm more interested in variety of experience.

I am certainly thinking seriously both about your path ideas and the no-selling, but I currently have a list as long as both my arms of things to do for the next version, six or seven of them major changes, and I'd like to finish it in this lifetime. Maybe I'll pass it on to my children like some kind of family curse.

[Mr.Peanut]

I just retested it on my DS, and it has no problem. I deleted the .raw files, and it took about 10-20 seconds to initialise the arrays. I am using a CycloDSEvolution. It almost sounde like you are running out of storage on your card - have you checked that?

Card has plenty of space (90% free). I am using a Supercard Lite(slot-2). I have reinstalled and repatched and am testing again...
... no luck. Still buggered. Are you sure there were no changes that might have caused this? The last version worked fine.
Okay. Now patching and installing on Supercard SD...
...Worked on the SCSD, so it looks like the problem is the SC-Lite dldi patch, or possibly my micro-SD. ARG!

[Nick]

Okay. Now patching and installing on Supercard SD...
...Worked on the SCSD, so it looks like the problem is the SC-Lite dldi patch, or possibly my micro-SD. ARG!

Phew - SEP

[Mr.Peanut]

Good for you but bad for me!
Can't find anything wrong with my micro-sd... sigh...

edit: I still find it strange that it fails always on writing object.raw. Are you sure no code was changed there? Nothing a quick diff or two wouldn't reveal.

edit2: load.c line 252-253 commented out i.e. byte old_dd; byte old_ds
These variables are still used, but a quick grep of 036 shows they are not defined anywhere.

[Nick]

edit: I still find it strange that it fails always on writing object.raw. Are you sure no code was changed there? Nothing a quick diff or two wouldn't reveal.

There were some changes to object.txt, but they shouldn't have broken anything.

edit2: load.c line 252-253 commented out i.e. byte old_dd; byte old_ds
These variables are still used, but a quick grep of 036 shows they are not defined anywhere.

Actually, they're only used inside a block that's got #if 0 ... #endif around it.

[Mr.Peanut]

Found it(I think). The object.txt file for 0.3.6 is in dos format not unix.

[PaulBlay]

Found it(I think). The object.txt file for 0.3.6 is in dos format not unix.

That sort of thing's a pain. It's even worse when you have to deal with multiple-byte character encodings as well (I have to keep switching between Japanese(Shift-JIS) and Japanese(EUC)).

[Mr.Peanut]

Arg! Converted lib/edit/*.txt files to unix, but it still hangs on object.txt...
I am getting tired of being wrong.

[Mr.Peanut]

It turns out the previous version I was using was the 2.53 from Immir's site not the 2.50 from here. I will retest 2.50 again later today, but IIRC it did not work for me either. That would mean the problem is from merging Immir's code into the main trunk?