cliffman wrote:joe_moog wrote:jeepo wrote:Why not remove the lower case and special characters to free space? I'm sure people will understand and actually prefer new functionality over lower case characters.
+1, although I'm not sure what this would "break". I have the Librarian Editor software and it might not play well with that, unless the OS knew if it saw lowercase/special chars coming from the application to upcase/replace with spaces, etc.
What version of the Librarian do you have? Mine doesn't recognize the lower-case stuff at all, it just shows blanks for patch names - is there a newer version?
I made this suggestion to Amos a while back. It was a bit of a laugh as I rib him from time to time about the fact that he should stop being so lazy and spend some time optimizing.
When I was a young man (kid), I taught myself assembly language on a machine with an architecture very similar to the Little Phatty. Well, not quite but it shared one very important limitation, that being 64k of memory of which some was accounted for.
There is only so much that can be done.
From a Character set perspective, it is likely that upper/lower case will make NO difference and I make this statement under the assumption that the character set is embedded in the display controller. If indeed it is a pure bitmap, then sure, he'll save by cutting out at least the 26 lower case letters x the bit matrix which is something like 8x8 or so which is 8 bytes saved +/-. So 8 bytes x say 32 characters and you've got a hefty 256 bytes back.
EDIT: Just checked, the character matrix is only 8x6 pixels so the saving will be even less!
The other approach which I suggested to Amos (both can be used) is to squeeze down the known bits into an encoded format such that every 6 bits represents a single character of storage (but the set only has 64 characters in it). So you can pack 8 characters into 6 bytes (instead of 8 characters into 8 bytes). Multiply that by 13 characters in a name x 100 presets and you start adding a few more hundred bytes.
But wait! You need some code to unpack these names for display and also need to get ALL of the librarian software to be able to interpret the new format.
In short, this is very geeky stuff and it's a losing proposition unfortunately.
But it's the life of an 8 bit programmer.
Keep up the good work, Amos; very admirable and generous of you and Moog for doing all of this for us.