<div class="gmail_quote">On Fri, Jun 10, 2011 at 20:59, James Porter <span dir="ltr"><<a href="mailto:jvporter@wisc.edu">jvporter@wisc.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div id=":h1">Is your goal merely to make the function fail when the result would be a<br>
non-empty array?</div></blockquote><div><br></div><div>Yes, that is the intended semantic. The zero length is calculated and the function should fail if we calculated wrong.</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div id=":h1"> I can't think of any other case where you'd be passing<br>
in an alloc value of 0 and expect iMesh *not* to allocate for you.<br></div></blockquote><div><br></div><div>See my and Jason's examples.</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
<div id=":h1">
<br>
In any case, a better spec would simply be "if you want iMesh to<br>
allocate the array for you, pass NULL as the list pointer", (i.e. get<br>
rid of the case where alloc=0 means "allocate the array for me"). I'm<br>
not really sure why we need two ways to tell iMesh to allocate the array<br>
for us.</div></blockquote></div><br><div>Yes, this is what I am advocating.</div>