2014-08-21 18:56 GMT+02:00 Kevin Wortman <[email protected]>: > > IMO there are enough use cases where the growing feature is unneeded and > it's overhead is unwanted, that the best compromise is to have both a raw > fixed-size array type (i.e. an R7RS-small vector), _and_ a higher level > self-resizing array. It is not hard to write a self-resizing array on top > of a fixed-size array. That's what e.g. C++ and Java do. >
Maybe we could come back to the original question how to avoid unnecessary initializations during the the creation of a new vector by parts of other vectors. The standard requires to use make-vector and vector-copy which initializes the new vector twice. Is here a broad consent that it is not worth to care about, because one should disregard vectors in favour of gap buffers? Sascha
_______________________________________________ Scheme-reports mailing list [email protected] http://lists.scheme-reports.org/cgi-bin/mailman/listinfo/scheme-reports
