Gharnef
-
Posts
59 -
Joined
-
Last visited
Content Type
Profiles
Forums
Events
Posts posted by Gharnef
-
-
Hooray! At least we see some progress! Now if only Twilkitri could at least update soon.
Twilkitri's MIA
Bwahahaa I think I fixed the text overrun issue. Gonna do more playtesting.
So I guess what I'd really like to do is make the stuff in FE4 consistent with the current etymology and mythology. Off the top of my head, here's what I know I'd change (aside from Serlis):
- Holsety -> Forseti
- Eltshan -> Eldigan
- Scathach -> Ulster
- Alvis -> Arvis
- Aless -> Ares
- Lakche -> Larcei
- Sety -> Ced
- Swanchika -> Helswath
- THammer -> Mjolnir (though I love reading THammer...)
- Falaflame -> Valflame
- Hmm, Narga vs. Naga?
Thoughts on what the community at large would like for an update? I don't have the official list of stuff that's changed or been updated in the etymology, this is what I've gathered from FE Awakening. Any playtesters interested in trying a new patch?
-
I'm close here ... with some hackery I compiled a patch where the only change I made was Serlis -> Seliph. Some of the text doesn't line up correctly now though, and while I haven't done anything more than rudimentary testing that's the only issue I see. TBH I just did a batch replacement of Serlis to Seliph so likely that is what caused text alignment issues.
Text overrun:
Serlis no more!
-
Well... in case anyone cares, I hacked at the 65816 cross compiler and with some other edits and the font compression in Windows thing, built a patch that is mostly correct.
-
I realize that post was a mess :) SF, Let's start over.
One thing I don't understand about compiling the project is newfont.bin and the compression. The package includes newfont.bin, which if I understand correctly, is just a new uncompressed font created for inclusion in the patch which overwrites the old Japanese tiles in the ROM. In Linux, fe4enc segfaults during while compressing this file. I can compress it in Windows using fe4enc.exe from a post by twilkitri's then copy the compressed one back into Linux to make it build, but that seems to screw things up bigtime when I make an IPS.
Has anyone gotten fe4enc to work on Linux?
-
I blew up the "Improving the Genealogy Path" thread with this and it was suggested I come over to the Hackers sub ... before we get started, I'm very comfortable with building software from source in a *NIX environment and hopefully you are too.
OK, now to the nitty gritty - I haven't even made script changes to the patch, I'm just trying to build it vanilla in a SLES11 SP2 Linux environment. First issue - fe4p/fe4p.c, the #define macro for row and col vars doesn't like to compile. I manually moved it inline and I can compile this OK.The next big issue is in making the fonts.
s11sp2:~/FE4/fe4transsrc # make menufont
...
./fe4enc newfont.bin zfont.bingmake[1]: *** [zfont.bin] Segmentation faultIf I run this command through GDB I see the issue is here:
Compressing... 3.3%Program received signal SIGSEGV, Segmentation fault.0x08048d6f in FindMatch (bestmatch=0xbffff514, readahead=0x8061239 "\347\271\ב\177\221o") at lz77.c:8080 if(*match != *readahead ||So I tried to fake this by using the Windows fe4enc.exe I've seen floating around by manually generating a zfont.bin from the newfont.bin included in the source package. I had to do this for another font file as well.
I eventually got an .ips (yay!) but when I loaded it, everything was messed up in the game (boo!).
OK, starting over. "make spotless" then just did a "make". All goes well until
gmake[1]: Entering directory `/root/FE4/fe4transsrc/dcenc'../bin/65816 -c -hi dcenc.asm...[entering second pass]assembling code at $91FAB0...finished at $91FBD3.*** glibc detected *** ../bin/65816: double free or corruption (fasttop): 0x08063b20 ***...Soo... Has anyone successfully compiled this from source in Linux to make a working .ips ? Anyone know why you'd make script, or make dump? -
instead of fooling with the patch itself, there was definitely a makefile for an accompanying text editor that you could probably use pretty easily
Any idea where that might be? I've combed this site, grandbell and the web in general - didn't find it. This is my first foray into FE hacking so it's possible I missed common resources. Thanks.
-
it'd be a tad harder than rebuilding the software from the source (although i vaguely remember a makefile for some text editor that would be substantially easier) since you're essentially hacking into an archaic game format with hardcoded stuff everywhere, trying to insert data that isn't meant to be there
Sorry, I didn't phrase it well. I'm not trying to edit the binary patch directly, that definitely won't work :) I just want to edit the scripts and rebuild the patch from source.
I'm close. There are some errors in fe4p/fe4p.c relating to the row/col preprocessor macros that I fixed myself. This enables compilation at least. But then during cross comp of the .ips it blows up:
../bin/65816 -c -hi dcenc.asm
...
[entering second pass]assembling code at $91FAB0...finished at $91FBD3.*** glibc detected *** ../bin/65816: double free or corruption (fasttop): 0x08063b20 ***And it pukes out the BT too. I don't mind making requested edits and sharing this if it works, but it's gotta work first - I was hoping to not be the only soul brave enough to try this at home.Still smh @ Serlis :) -
Not that I want to take ownership of the patch but "Serlis" bugs me so much that I wanted to edit and rebuild it. Has anyone successfully rebuilt the patch from source? I'm very familiar with *nix and building software from source (I'd better be, I do it for a living) and I'm getting a pretty specific and repeatable error I can share if anyone with technical aptitude and familiarity with the patch wants to see it.
Seriously, Serlis ... ? :/
Improving the Genealogy Translation Patch
in NES and SNES Era
Posted
I'd go Lord -> Great Lord if anything.