Re: I wonder if I'm missing something...
"David McClain (as dbm at refined-audiometrics dot com)" <[email protected]>
| Newsgroups | gmane.lisp.lispworks.general |
|---|---|
| Message-ID | <[email protected]> |
Yes, I see that we are entering the divide that seems to happen with Common Lisp - do the problem solution correctly, which means a huge amount of work and far reaching consequences. I guess what I’m striving for is an immediate Q&D solution to a problem that you run into in the lab when just trying to get some results for data collection. I don’t need to solve everyone’s problems. I just need to solve my immediate problem. And for that, C TypeCasts work beautifully, even if obscene. And hence the divide between the practicing Empirical Physicists and the Computer Scientists. (and hence, why we get abominations like Python.) > On Nov 8, 2025, at 08:32, Tim Bradshaw <[email protected]> wrote: > > On 8 Nov 2025, at 13:43, David McClain <[email protected]> wrote: >> >> And since we are talking about read-only access, I really don’t understand the aversion to multi-type overloading access. > > It is clearly extremely possible for an implementation to read and write vectors of immediate data (which could be slightly different than vectors with upgraded element types which are not T but likely is not) with very high performance without copying data in many cases (copying the bytes of a multi-byte object is unavoidable in general if you care about network byte order). And to do that without type punning. > > That's what you asked for. _______________________________________________ Lisp Hug - the mailing list for LispWorks users [email protected] http://www.lispworks.com/support/lisp-hug.html