On Sun, 6 Aug 2000, Doug Kaufman wrote:
> try to redo the patch so that it works as intended. The problem
> appears to actually be in the Ascii transfer after all (unless someone
> thinks that lynx should specially handle CR CR LF).
Here is a revised patch. I hope this doesn't break anything.
Doug
--- lynx2-8-4/WWW/Library/Implementation/HTFTP.c Fri Jun 2 18:01:06 2000
+++ lynx2-8-4/WWW/Library/Implementation/HTFTP.c.new Sun Aug 6 23:35:58 2000
@@ -3064,21 +3064,12 @@
format = HTFileFormat(filename, &encoding, NULL);
}
format = HTCharsetFormat(format, anchor, -1);
+ if (server_type == CMS_SERVER) {
binary = (BOOL) (encoding != HTAtom_for("8bit") &&
encoding != HTAtom_for("7bit"));
- if (!binary &&
- /*
- ** Force binary if we're in source, download or dump
- ** mode and this is not a VM/CMS server, so we don't
- ** get CRLF instead of LF (or CR) for newlines in text
- ** files. Can't do this for VM/CMS or we'll get
- ** raw EBCDIC. - FM
- */
- (format_out == WWW_SOURCE ||
- format_out == HTAtom_for("www/download") ||
- format_out == HTAtom_for("www/dump")) &&
- (server_type != CMS_SERVER))
+ } else {
binary = TRUE;
+ }
if (!binary && type && *type == 'I') {
/*
** Force binary if we had ;type=I - FM
__
Doug Kaufman
Internet: [EMAIL PROTECTED]
; To UNSUBSCRIBE: Send "unsubscribe lynx-dev" to [EMAIL PROTECTED]
