Would it be a good idea to have braces act as syntactic sugar for the "block" keyword? Ie. instead of block( ... ) just: { ... } Or is the savings too little...
I think that IO really needs better error messages. A lot of the error messages that I have been getting are completely useless to me. Could these be...
... I can't speak for Steve, but I believe it's in the spirit of linguistic purity and minimalism, a la Smalltalk. There's a certain syntactic uniformity to ...
Daniel A. Koepke
dkoepke@...
Jan 3, 2003 7:48 am
376
I know that it is unfinished work, but the VM shouldn't segfault [IoVM version 20021211] Io> "HttpQueryParser.io" doFile <Block 0x8073f08> Io> HttpQueryParser...
... the ... It's an interesting idea, but yeah, I'm not sure it's worth it. I'd like to keep the syntax as simple as possible. I'll give it some more thought...
... I agree. I'll think I'll work on that when I get back from the holidays. ... error ... I like that idea. I've been considering printing the source line for...
... Io doesn't really have any key/reserved words other than "=". "block" is just an slot in the Lobby that can be overridden like any other. This helps keep...
... Or changing the parser to use "block" in the case of no message name: Account deposit = (amount, self balance += amount) ... redesigned. ... true and ... ...
... Debian Testing with gcc-3.0. An strace is attached. Regards, Markus -- http://reauktion.de/archer/ execve("./io", ["./io", "min.io"], [/* 18 vars */]) = 0 ...
... for sake of those reading archives, this comment was made by Danial Koepke. ... that seems to make more logical sense. I am trying to come up with a...
... This breaks at least part of the expression-based semantics of if, as you noted. I'm not sure if that's important or not. We could keep the semantics of...
Daniel A. Koepke
dkoepke@...
Jan 3, 2003 11:43 pm
388
The build system for IoDesktop is OS/X-specific. To compile on Linux and, I assume, any other platform, requires a bit of work. There doesn't appear to be a...
Daniel A. Koepke
dkoepke@...
Jan 4, 2003 12:40 am
389
Markus Fritsche wrote: Btw., how do I get a stack trace? I don't know much about debugging with C. Here's some output of my "first contact" with gdb: min.io: ...
... I should spend more time googleling before posting :-) mfritsche@server:~/IoServer-2002-12-11/IoVM$ gdb ./io gdb: Symbol `emacs_ctlx_keymap' has different...
On Sun, 5 Jan 2003, Markus Fritsche wrote:
> #1 0x0805b667 in ByteArray_replace_with_ (self=0x8075440, > substring=0x8076638, other=0x8078068) at...
Daniel A. Koepke
dkoepke@...
Jan 6, 2003 8:24 am
392
... Thanks for the patch Daniel! I've added it and will do some testing. One of the reasons I've kept the base functionality separate from the VM is so other...
In an earlier thread, I suggested the possibility of adopting Lua's idea of not requiring parentheses around certain types of function arguments for unary...
Daniel A. Koepke
dkoepke@...
Jan 8, 2003 9:37 am
395
... Ok, what about chained messages like: list at(x) print in the proposed syntax this could also be written as: list at x print How is the above distinguished...
... That's a toss-up. There are two solutions, assuming that we're going to be unifying the semantics: 1. Use the operator semantics for all messages (a...
Daniel A. Koepke
dkoepke@...
Jan 9, 2003 1:29 am
397
I'm not a windows user myself but there are folks on the iolanguage mailing list that are running Io on windows under cygwin. I'm replying to the mailing list...
Does anyone have a screen shot of some of the graphics for IO. I can't get the IoDesktop to compile on windows. I'd like to see what I'm working towards. ...
... I suppose there's also a third (not mutually exclusive) option that I hadn't really considered: require an operator character at the end of a message name...
Daniel A. Koepke
dkoepke@...
Jan 9, 2003 12:13 pm
400
... So: list at x print would be parsed as: list(at) x(print) And name space chaining like: Extensions Algebra Matrix = Object clone would be parsed as: ...
