Mailing List Archive: 49091 messages
  • Home
  • Script library
  • AltME Archive
  • Mailing list
  • Articles Index
  • Site search


From: carl::cybercraft::co::nz at: 11-Aug-2008 18:52

On Sunday, 10-August-2008 at 21:46:48 greg.schofield wrote,
>People like me are a pest and I know it. > >I am nowhere near being a REBOLer, but I am extremely impressed by what >little I do know. > >I have come up from REXX, I script occassionally and only when I need som >ething done, it can be anywhere between 6 months to a year or two between > projects. Relearning a language is a major problem, hence the simplicity > of REXX is a great asset for that language as well as its natural limit. > >I can see REBOL doing both and doing it well, mostly because of the power > of dialecting. > >REBOL is full-on, especially with the datatypes, and I admit I just can't > really get some pretty simple stuff, for instance I still can't really g >et my head around blocks (how stupid does that make me!).
It just means you're not a true block-head yet. ;-) Blocks are just containers, or that's how I look at them. Consider a string. It's just a container for bytes/characters, right? Well a block's just a container for REBOL values. ie... if a = b [do this stuff] IF expects two arguments, the second being a block. And if the first argument returns true, REBOL will evaluate what's inside the block, otherwise it'll ignore it. But that block's no different to this one... blk: [some stuff] Meaning a block with code in is no different to a block with data in. And you'll notice IF's behaving like a function, not a command or statment. And just like a function that accepts a string, it's accepting a block. It's that simple.
>My suggestion is REBOL-SIMPLE as a very simple dialect of REBOL resemblin >g REXX and heavily Function based, but using as many characteristics of t >rue REBOL to make transition from it, to the real stuff, as seamless as p >ossible. > >A data-type-less language where as far as the user is concerned every var >iable is a string (using the functions themselves to sort things out). I >would also suggest versions of stem-variables as a simple way to manipula >te related data. There are a few things I can suggest, but the idea is ju >st make it simple (ie every "do" has an "end", every "if" etc.,. simople >and clear often means verbose scripts).
It's an interesting idea, though I'm not a dialect expert and don't know REXX, either. However, wouldn't a simple language like that, (only having string variables, the need for "end"s, etc.), make the transition to REBOL proper just as hard as it's been for you with moving from REXX? -- Carl Read.