diff options
author | Eric S. Raymond <esr@thyrsus.com> | 1996-10-31 09:16:39 +0000 |
---|---|---|
committer | Eric S. Raymond <esr@thyrsus.com> | 1996-10-31 09:16:39 +0000 |
commit | f773cf68c2b60a003966d8b69a6f90536fad8bd1 (patch) | |
tree | 865db31843e253b06c8931e08f900994a68350de /socket.c | |
parent | 1d95131a945586b4eb6596583c364e5eb60025a8 (diff) | |
download | fetchmail-f773cf68c2b60a003966d8b69a6f90536fad8bd1.tar.gz fetchmail-f773cf68c2b60a003966d8b69a6f90536fad8bd1.tar.bz2 fetchmail-f773cf68c2b60a003966d8b69a6f90536fad8bd1.zip |
Add FIXME notes.
svn path=/trunk/; revision=456
Diffstat (limited to 'socket.c')
-rw-r--r-- | socket.c | 4 |
1 files changed, 3 insertions, 1 deletions
@@ -86,6 +86,8 @@ va_dcl { } /* + * FIXME: This needs to be recoded to use stdio, if that's possible. + * * If you think these functions are too slow and inefficient, you're * absolutely right. I wish I could figure out what to do about it. * The ancestral popclient used static buffering here to cut down on the @@ -95,7 +97,7 @@ va_dcl { * The right thing to do would be to use stdio for internal per-socket * buffering here (which is why Socket() returns a file pointer) but * this causes mysterious lossage. In case someone ever finds a way - * around this, a note on the original implementation said: + * around this, a note on Carl Harris's original implementation said: * * Size of buffer for internal buffering read function * don't increase beyond the maximum atomic read/write size for |