Re: [reformated] Re: proposed Eistring interface

On Tue, Apr 18, 2000 at 09:53:36PM -0400, Olivier Galibert wrote: Found the time, after all, even if I should be doing something else. On the whole, I like it. Some remarks, though: - static inline functions are infinitely preferable to macros - alloca() doesn't go well with string concatenation (no realloca) - alloca() bites you if you try to return one of these strings to a caller or change a string which was given to you through a function call - we'll need a kind of sprintf - dynamically allocated strings tend to add a whole lot of overhead in all mutating operations to check whether they need to be extended. But what I'm the most afraid of is that it is going to add yet another copy of data in a number of cases, which will slow down the system. In particular I dread the Lisp_String->eistring->cstring->system call path which may be seen as an easy way to code this. But like I said, on the whole, I like it. OG.