The storm is upon us… the storm of activity, that is! Conditions improved enough for me start working again. I have certainly taken advantage of it! It’s funny how one thing leads to another. Shortly after returning to work on Tenshi No Uta, I was met with some additional non dialog text to dump. None was stored in a similar manner to any other text in the game. I grew a bit frustrated that I lacked some reliable generic tools to handle some of these text formats. So, I decided to take some time and start developing a better dumper, which led to my Generic Dumper project, which will likely be released to the community in some form. Read below for more details. While I started developing this project, I was beefing up my table file library. I realized we don’t have any official standard on table files and their features. So, work ignited on creating a rough draft for standardization of table files for myself and the community. Details can be found below. Next, the additional activity prompted Mattias to contact me regarding applying UVWF (Universal VWF project) to Terranigma and doing the text reformatting work. And thus one thing lead to another and another until a flurry of activity began! Now, on with the status updates!
For years, ROM hackers have had to deal with countless incompatibilities amongst table file uses for various utilities. Even when developing my own utilities, I didn’t have a clear definition of the features supported in the table file format. I decided to do something about it for myself, and hopefully the benefit of the entire community. The document’s purpose is to document and standardize the table file format in hopes that ROM hackers can progress forward with compatible utilities, clear indication of format, and standardized features.
I have already written a first draft. The draft was created by reviewing documentation and features of various utilities throughout ROM hacking history, reviewing their relevance to the primary purpose of a table file, and update them to meet current needs. Basically; collect, improve, and standardize what we’ve been doing while touching on some tough issues such as UTF-8 encoding.
Obviously, for this to be of any use outside of myself, it needs to be adopted and accepted by others. I urge you to read the document, make some suggestions, present some ideas, and comment. It is my hope that we can start having compatible utilities, standard features, and pave the way for advancement!
I’ve been working on a GUI based flexible, generic dumping (and hopefully insertion) program in recent weeks that would be able to do most everything that Cartographer can do and then some. I’m trying to eliminate the need for a custom dumper in as many cases as possible. I thought Cartographer and Atlas made some good progress in that direction, but lacked simplicity and ease of use.
Implemented Features so far:
- Follows Table File Format 1.0 Draft for supported features.
- Raw Dumping
- Pointer Table Support
- Variable Pointer Sizes (up to 32-bit Big and Little Endian)
- Odd Pointer Spacing
- Relative Pointers
- Fixed Length Strings
- Fixed Length Lines
- Atlas Compatibility
During development of UVWF, I threw around the idea of Terranigma as a nice test case for useful implementation in a game other than Tenshi No Uta. I wasn’t big on doing the script reformatting work to make a project out of it. As a result, it didn’t go very far. Recently, Mattias (of Terranigma Swedish Translation fame) contacted me about doing the grunt work. He has the experience necessary, so I started taking a look at this.
Terranigma uses sprites for the intro text and some of the routine is shared with the standard dialog and menu text. It also takes advantage of color manipulation. The UVWF is not yet mature enough to handle these more complex situations. I do find it an interesting case to continue to look at for study and potential support for continued development of UVWF.
I think I may have to draw the line somewhere on the extents this project will be able to cover and what would require a custom VWF. As I’ve seen before, the more flexible I make it, the slower it becomes for time sensitive implementations. I’ve been playing with the idea of handling some cases like this in a modular fashion where various versions of a ‘RenderChar’ function would exist and you would use the appropriate one for the job. However, that complicates the work involved in implementing it and adds significant size. Originally, I had hoped it would be able to be implemented with very little assembly work by the user.
I will continue to try and adapt my design to meet my needs with these situations and see where it goes. It’s really just been one big experiment to take this concept for a ride and see what I could do with it. Certainly limitations and out of scope cases are going to exist. I just need to try and define what that is going to be. We’ll see where it goes from here!
I have a bad habit of underestimating the amount of misc. text/menu work to be done in my projects. I suppose that’s probably because it’s my least favorite part of the project. I found several sets of misc. and/or menu text that needs to be dumped in various formats. I grew pretty frustrated that I didn’t have tools to handle all of them. That spurred my Generic Dumper project. I did a little bit of work on dumping some of the needed text as test cases for the program. I have gotten off on a tangent with some other work mentioned in this news update, but expect to get back to Tenshi as soon as my Generic Dumper progresses.
You know how these things go in the world of hobbyist work. Things don’t always go according to plan, time lines are thrown out the window, and surprises lurk around every corner! I certainly didn’t expect to be working on the things I am now. It has been a nice change of pace and some fresh air around here now that I’ve been back to more active work. I’m hoping conditions will allow me to continue this nice pace of activity!
-Nightcrawler
Greetings TransCorpians! 2009 is drawing to a close and 2010 will soon open. I for one am looking forward to a fresh new year in hopes of improved conditions for myself and TransCorp! It’s been a very rough year for me on a personal level with some physical disabilities impeding progress. However, progress has been made nonetheless when conditions allowed. While life’s lemons have seen to it that I could not bring you a completed translation to stuff in your stockings, I can sprinkle a little news and eye candy your way!
Tenshi No Uta has seen much menu hacking progress. When we last left off, we needed a new thin font and many tile map text routines converted over to VWF, using my ‘Universal VWF’ system in development. Thanks to Gemini, we have ourselves a fitting thin font that will allow us to take maximum advantage of all available screen space. It’s definitely necessary as some screens are fully used with no room for any expansions. Though I don’t show it in the screenshots above (some text systems are still not converted), there are many screens that utilize near 100% of the screen, mostly when full item lists are given along with several complex menus on screen at the same time. A thin font was a necessity to ever hope to squeeze full English in there.
We’re now up to seven different text engines in the game that have been hacked to use my ‘Universal VWF’ system. Five of them were tile map based. As I mentioned in my forum post in November, the practicality of the universal routine is starting to hit some walls. To keep it at a high enough level to be flexible and game independent, there’s quite a bit of overhead that gets added. Then there’s the fact that there are just so many possibilities and ways games handle text printing in tile map based routines. Again, you go higher up on the abstraction ladder adding more flexible code. What’s the result? Something that ultimately gets slower and slower. This isn’t too bad except for cases such as item lists where up to a full screen’s worth of data must be dynamically redrawn quickly and interact with user input. That’s when the lag starts to really show. So, while my routine is doing the job I set out to do with it, it may be more practical in some instances where speed is important to use something more game specific. At least I can rip and reuse parts of my system for these instances.
I’ve still got a ways to go with it. There’s a few changes I know I can make that will help in the speed department and I still haven’t fully hacked Tenshi’s menus to have full multi-routine screens in place to really benchmark and optimize the system. So, we’ll see where it goes. Its usefulness for menus and items will likely be a bit limited, but for everything else, I believe it should be able to be applied to other games easily. It’s been a fun venture so far nonetheless. I can also examine such things as further modular breakdown and other ‘big picture’ items when I get Tenshi fully converted and working. I haven’t hacked the battle menus yet which will have 3 or 4 more text engines from what I have seen so far. So Tenshi No Uta really is a spectacular test subject having more individual text engines than I’ve seen in any game I’ve hacked before. The total when it’s all said and done will probably be 10-12 routines.
I will certainly push forward with Tenshi when I can. However, I will likely be unable to work for a few months while I recover from an upcoming surgery. However, I’ll be back at it whenever I can. This old dog isn’t dead yet and passion for the hobby is as strong as ever. It may be at a snail’s pace, but I continue to march forward. Thank you for your ongoing support. It definitely helps me get through these down times quicker! TransCorpians, you rock!
-Nightcrawler