Current Progress

[Iambian]

http://iambianet.freewebpage.org/project0.html

Current version: 3.07

Features currently supported (taken right out of the readme) :

[Iambian]

Alright. I've decided to open this topic up for discussion. In the meantime, I've been working on a few things.

1. A version of the line read function that'll read the binary of a program and output hexadecimal digits. A version of the line write function to write binary given a hexadecimal string. And a version of the delete line routine that will delete X amount of bytes N bytes from the start of the file. This set would work well for the more ASM-inclined programmer.

2. A relatively simple version of the readline function that will output the entire contents of the program into a string, newline chars and all.

The following are still planned.

1. A version of the lineread routine that will read multiple lines given a second numeric input (will be decided later on, but it'll most likely be the same way you have to input the second number for the TOKENCODE command)

When I've tested some of the features already in the code to work, I'll be releasing Celtic 3.08. Hopefully, this won't take very long (maybe a day or two)

[Iambian]

Okay. It took me longer than a day or two, but it's finally out. Celtic 3.08 is now released under the "open beta" status, so you can test away without fear of too much being changed around.

The key features of this release are as follows:
1. Support for multiple line reading, replacing, and deleting
2. Binary reading/writing/deleting for those ASM inclined.
3. A final fix for the 9999 input limit, though the TOKENCODE command hasn't be rolled back to reflect those changes. (I'd like to know if you like the a+bi method better for input)
4. Reimplemented the LINEREPLACE command, and again, powered up with the multiple line replacing.
5. Actual checking to ensure you don't attempt to insert more memory than there is free RAM. (Why didn't I do this earlier?)
6. A command that will allow the contents of an entire file to be copied to a string.

There should also be a note that for the binary reading routine, there are no bounds checking to ensure that you aren't reading from memory areas outside of the intended variable file. Seriously, I don't want to put in any checks for that since it really isn't hurting anything, but that also means that you might be able to use Celtic III for purposes for which it was not intended :ninja:

The updated file is currently on my website. See post at top of thread for link.

You'll want to read the update information there, too. There are some more technical things that aren't mentioned here, including changes that never made it to the readme.

EDIT: And as always, if you see any bugs in the program, please report them. I'll never know unless you tell me.
------
EDIT2: I've also noted that the project is growing pretty large, and it may soon become RAM-prohibitive to keep as an ASM project. What do you guys think about this thing becoming a flash application?

[justusdude]

Sounds like a good idea. That way you can create your own tokens and whatnot.

[pugboy]

I second that. It is nice to have an 84+

[Harrierfalcon]

If it's going to enter app form, I'll make a few suggestions. First, use the imag( command . Second, implement any features that even slightly look good, because someone'll use it sooner or later. Make use of your 16K!

Third, this is getting excellent. Keep up the good work!

[pugboy]

Hey, I was going to use the imag( . I thin it is a good idea to use the imag(. Everything uses real(, and it gets annoying to type in real(100,...

[Weregoose]

For the love of all that is holy, please do not use imag(). Think about it.

[pugboy]

It just outputs a 1 or 0...

[Iambian]

Well, I do have a few ideas that might be app-specific.

Nice idea about the imag( token being used for more than one thing. I know it has uses, but I also know that any hook will pick up on the fact that if there are more than one arguments placed into imag(, the OS will throw an error if nothing's done about it. That's where the command input will take charge, and I suppose it'll set itself apart from all the other apps that make use of the real( command.

In addition, I'm thinking about allowing the use of the sub( and the inString( commands to allow extracting string parts and whatnot out of program files instead of strings. Don't worry. I'll have a way to distinguish between legit use and Celtic use. I'll have those ideas outlined later.

As far as the app goes, I might also include an option, since I have 16K to waste, to have the app generate an ASM program that can be used to autoinstall Celtic. Anyway, you can put those ideas in the feature request thread.

[DJ Omnimaga]

I think he made celtic detects if imag( is used for its normal purpose or if you use it for celtic, thus, you can still use imag( fine when celtic is installed

[calc84maniac]

[vuurrobin]

I am not sure whether I like the idea of celtic becoming an app. since people already can create custom builds, they could create 2 programs with each different functions to keep the size down and the speed high. it's not like you need every function all at the same time...


the reason that most people use real( is that there is a workaround using imag
(. if you use imag(, that workaround can't be used. I suggest to stick with real(, or to use ipart( or int(, since they do almost the same.


if this does become an app, make sure it is compatable with xlib (and maybe some other apps). and that installing program would be usefull to (xlib could use a prog like that to)

[calc84maniac]

How about det(? I think it is a one-byte token, so that will cut down on program size. Just check and see if the first argument is a matrix, and if so, return control to the OS.

[DarkerLine]

I also suggest making the chosen function return a value even when there's no meaningful value to return (just return 0 in that case).

This shoudl allow the programmer to put it into a seq( loop or an expr( command, among other things.

[Weregoose]

[DJ Omnimaga]

[Iambian]

I plan on having Celtic return values according to whatever function it is.
For example, say we'd use a command to extract a line from a program. The result of that would then be a string. For commands that have no return value, it'll just mirror what Ans had.

And thanks for the suggestion about using det(. Yes, it is a one-byte token (ti83plus.inc says it equates to $B3).

I'm just happy that there's a bit of support going for this project turning into an flash application, so I'd better get to work with the conversion. If I happen to add any new features that aren't app-specific I'll still update the RAM-resident version to reflect the changes.

[Areign]

have any of you used symbolic, its seems that it makes its own commands, why cant this be the same? instead of ruining programs that make use of commands why not make your own? if that cant work, i also dont understand why people cant implement the actual command into the app like if xlib, in addition to doing whatever it did, also returned the actual real value of any varialbes that were passed in? i dont know asm so i dont know if its feasible but to me it seems easy, at least the second part. like in the case of det( people are passing in a matrice. so if the person passes in a matrice, have the actual det( do its work, otherwise your lib will take it. idk if anyone has ever done that, but that would make it alot easier to use for those of us who actually use the calc for math programs as well.

[DarkerLine]