Question Running on a Windows machine?

I appreciate your kind words, Sasha. But even when my computer is working, I find I run through cycles when I'm helpful and when I'm not. There are a few things going on now that just make it a bit harder. I'm expecting to get past this around spring and by 2019, I'm expecting to have even more free time. In fact, thinking about 2019 is probably slowing me down a bit

I hope everyone is having a nice season and looking forward to the new year!

thomas wrote: Unfortunately something is still not right after one does this - the mud has a bug with these settings causing it to crash after a couple of minutes. I have still not been able to pin it down - apparently somewhere we're either clobbering memory or freeing something we're not supposed to. This doesn't happen on linux or when compiled in cygwin, so it might be some library included by visual studio that's causing it.

So, I recommend that instead of going through all that trouble, you run tbamud under cygwin.

Okay, here I am, half a year later, thinking I should dust off that 3.57 and play with it a bit in hopes that it would encourage me to do a bit of work on my zones. However, it was before any of my zones went in so I decided to add my zones in manually.

That's when I found that the nice download zones section is gone now. So...in my simple-minded way, I grabbed the lib\world section of the 3.66 and swapped it out. I expected to get bugs but figured it would sort of maybe work in a general sort of way, right? I was sort of surprised to see that older zones (I expected that all the bugs would be in the newer zones because of new code that would be missing) suddenly had a bunch of mobs that were flagged NOT DEAD YET! Anyway, the mud started up and off we went! ...until it crashed... But still, not unexpected, right? So I started looking around for information about start-ups and DGB and crashes and anything that would help me figure out what was conflicting.

I read Thomas' post a few times when it struck me that the situation sounded very familiar. I swapped back the old lib\world and there were no crashes. "AHA!" I told myself. "Maybe it's not the code, maybe it's the zones!"

Therefore, I went back to Visual Studio and compiled 3.66. Sure enough, there were the mysterious crashes. I swapped out the world file for the earlier one and...WOW, no crashes. That was good enough for me! I compared the zone files of the 3.57 and the 3.66 and...um...oh, *expletive deleted*, all the changes I could see on a fast comparison were ones I'd put together! 211 wasn't in the old one but it was in the new one. Hmm, my Tarot zone. 355 and 356? Those were the Ultima that I did as an experiment because someone had asked if it were possible to port it from a ROM based base. 653 and 654? Those were my player housing zones.

I manually entered them into the old lib. First Tarot, then housing, then Ultima, restarting each time and waiting for the crash. It's now been 2 hours and the 3.66 is still running.

I still believe that it isn't Visual Studio and it isn't anything about the codebase. I don't know why it doesn't show up in linux or Cygwin but switching the world file makes 3.66 run with Visual Studio without crashes.

Next step: THE WORLD! *maniacal laugh* Or at least a trial of TBA 2018.

Glad you're up and running again!
I suppose the moral of the story is ¯\_(ツ)_/¯

WhiskyTest, I like your moral and totally agree with it.

Using the same 3.66 changes for Window along with the tbamud-2018.1, I made the same changes in Visual Studio. This gave me one error which I forgot to record but which went something like this:
error C2065: 'HUH': undeclared identifier

Since I couldn't find a way to fix it, I cheated and commented out the line in db.c so it looks like this:
3786 /*CONFIG_HUH = strdup(HUH);*/

This gives me a working tba2018 with 11 warnings and, when using a nonsensical command, an output like this:
> alphabet me
(null)

When I ran it with the current zones, it crashed every few minutes. I swapped out my amended world files (3.57 + Tarot, Ultima and housing) and it's been running for hours. I compared some of the file sizes and found that Whiteknight has added two zones (343 and 346 are extensions to God Complex) and changed one (274 changed Smurfs to St. Brigid) but since these were not in my crash_world files of 3.66 I didn't bother to swap them out to test them.

I'm much happier with a more up-to-date mud to play with and I can live with the (null) thing, especially since I know what it is, but I'd be even happier if someone could fix it for me

Re: crashes

I've found that the mud will run using all the current world files except 0.obj. If I replace that one file with a copy from 3.57 all the crashes stop. I've looked at the files and done an eyeball comparison. The crash file is missing #45, a pile of gold coins.

I thought that must be it and added it in manually but the crashes continued. I also found differences in #71, a Chinese finger trap and #0, a bug, was missing. After I copied in the changes, the mud still crashed. The only way I could stop the crashes was to replace the file.

This is a bit of a disappointment to me since this file has a bunch of quoted text. The problem with quoted text is that when you /fi, two spaces appear before the end quote marks. "I am the dragonbane." turns into "I am the dragonbane. "

These are all fixed in the current file but not in the old file. However, that's a small price to pay for a non-crashing mud. What I don't understand is why it only crashes under Visual Studio and not Linux or Cygwin.

maybe there is some invisible formatting in 0.obj?

Try making a new empty txt file in notepad, copy the contents of 0.obj that is suspect into it, save and rename to 0.obj