If you're just a user of RPGMaker Trans, please ignore this message. Translators can look forward to a proper GUI and stuff. However, developers... please click through and read on.
I'm having some trouble with current work in RPGMaker Trans; specifically with some internal architecture choices. Basically, it boils down to the nature of the patch structure: originally it was "one-file, one-patch file" basis. This failed due to the RPG_RT.LDB file being massively oversized. Presently it's a sort of cludged together method that splits the RPG_RT.LDB file. That's inelegant, but it works. Another cludge was the problems with multiple meanings associated to the same translated text, partial fix implemented by "contexts". Again, inelegant.
The problems start with the communications part of my speedup architecture - communicating data between processes is somewhat tricky. However there's other problems, for instance sharing data across files when desirable, and more fine grained control over replacement policies i.e. allowing instances to be translated etc. Final problems come with porting the tool to RPGMaker XP/VX, which use variable width fonts - meaning giving advice on line sizes is very, very hard in a txt file.
Finally there seem to be some user experience issues, due to users not reading instructions properly. Some users appear to think that RPGMaker Trans is an automatic machine translator as it's default behaviour is to create a patch folder if it can't find a patch. Clearly I need to do something about this.
Basically, the more I look at it, the more it seems that the current method of text based patches is wrong. Hence the v3 RPGMaker Trans patch format will almost certainly not be human readable, and instead there'll be an editor program. Whilst the format will likely be Pickle based, this is not helpful for 3rd party utilities.
In addition, because of the license for RPGMaker Trans, after this patch format is released it will not be permitted to run older versions of RPGMaker Trans which use the v2 format.
Now to be clear on all fronts, this is still a little way out there (2-3 months, maybe). But I want to give advance warning to anyone working on tools: I am without doubt, that all 3rd party tools will be broken by these changes and it will require a vast effort to restore them.
No comments:
Post a Comment
Note: only a member of this blog may post a comment.