... One possibility is to override the Object method method to eval it's arguments in such a way as to accept the syntax you use in your example. A simpler...
... Also, I think it's worth noting that no language that I know of does this in a way that doesn't seem ugly in some way. I know of only two solutions. 1....
... That's true, but the assumption that "o v" is faster makes assumptions about the implementation and the hardware on which the implementation is running. ...
I reformated the python proposal to be consistant with the other ones and moved them all into a new page. I also added the python proposal to the synthetic...
... ones and ... to the ... I've been following this discussion with great interest, but couldn't contribute due to time constraints. Here are some of my ...
... Thanks for sharing your opinions! It's good to have more people involved in the discussion. ... ...(snipped opinion that locals should have lexical scope,...
... 3. Mark locals and or variables with a wart, so just by looking at the name you know what scope it is: Ruby: aLocal = 1; @ObjectMember = 1; @@classMember =...
I put together another table of the proposals: http://www.iolanguage.com/Library/locals.html Does it look correct? It looks like there is some general...
... I think we all agree that we are trying to distinguish creates from updates, *and* we are trying to distinguish local scope from member slot scope. The...
... Cool. I'm not sure the synthetic presentation really helps. It was not easy for me to understand how it worked, and it is still difficult to get out of it...
... Very nice! It's really easy to see. But I'm not sure it has quite enough information. ... The self EIA column looks incorrect for most of the proposals,...
... Actually, 7 should DO an error-if-absent on non-locals update. This is the one feature I like. Also, 5 could read: "creation should specify explicit...
... Right. And while I keep trying to propose using OTHER characters, nobody answers to that because they're considered "ugly". Allowing : in names comes from...
... I wouldn't worry about that. Colons allowed in names is just a hack for the Objc bridge. If it's a problem, I'd rather change the bridge (which is just a...
... Right. I was lost with the number of proposals, each one being a slight variation of another. I did this for myself at first, but I thought it might be...
... I only partly agree with you. First, I don't think "intuitive" should be used to describe things that seem comfortable after you have already learned them....
... I wouldn't say cleaner, but certainly shorter. How would it work outside a method, say for an object? Would it be like this? "New": new a = Object clone a...
... The self EIA should have an X for most of the proposals. They generally raise an exception if neither a local nor slot exists. "Kevin" certainly does, and...
... This is not an important feature to have, I agree. This is only a summary of the mailing list discussion, where it seems that people were more vocal about...
... Yes. ... Outside methods, it doesn't help. Inside methods, not only is is "slot" shorter than "self new", but it might be easier to implement the...
... The way I understand the slot/local proposal versus the "new" proposal, using the above example, but encased in a do(), is as follows: ... new a = Object...
... Ok, I'm convinced: New3 is worth looking at. ... I'd like Steve to give us his take on this. ... What's a "module"? How would you define it? Eric...
... Kevin is right: most proposal forward to self when updating and local is not found. So, locals EIA should be either almost the same as self EIA, or almost...
... I just tried it, and do() doesn't seem to have locals: it's not a block() nor a method() and you cannot do something like: d = do( "abc" print ) So that...
... Obviously, I meant: new a = Object clone do( new initVal = 1 // not a local new v = initVal new inc = method( v = v + 1) ) Which is exactly the same as: ...
... In the implementation I was thinking of, "local" would be either be implemented as a keyword similar to return, so: Lobby local a = 1 would be parsed as: ...
