On Mon, Aug 17, 2009 at 12:37 AM, Per Bothner<[email protected]> wrote: > On 08/16/2009 08:30 PM, Eduardo Cavazos wrote: >> >> Per Bothner wrote: >> >>> You might consider implementing similar hooks for Ikarus. If you >>> have any questions about what is needed, feel free to ask. >>> I'll be happy to merge them into the reference implementation. >> >> Hi Per, >> >> This implementation of SRFI-64 is not Ikarus specific, but is part of a >> project to implement R6RS-portable implementations of various srfis: >> >> https://code.launchpad.net/~scheme-libraries-team/scheme-libraries/srfi >> >> So I think they're aiming for not including implementation specific >> things. It's conceivable that some extension which only assumed R6RS >> could be added, but the drawback is that somebody making use of that >> extension might be surprised if they port the code to a different >> implementation of SRFI-64 an then *not* see the extension behaviour. > > Well, for a testing framework it's really nice to have line numbers > when reporting an error. And there is no portable way to provide > line numbers. So the SRFI-64 implementation provides a basic > portable implementation, but provides hooks to provide extras > like line numbers. I assume some but not all of the optional hooks > would be unnecessary if one can assume R6RS. > > This isn't all that different from SRFI-1 or SRFI-13 which contain > extensive notes about porting the reference implementation, both wrt > prerequisites and optimization. > -- > --Per Bothner > [email protected] http://per.bothner.com/
Did one if you want to update the R6RS version with an abstraction layer for reporting line numbers? I am just wondering what came of this thread.
