Kermit insists on treating PDF files as text - Protocols
This is a discussion on Kermit insists on treating PDF files as text - Protocols ; Desktop side: Kermit 95 version 2.1.3 on WinXP
Server side: C-Kermit 8.0.209 on Linux
Using SSH sessions
I type "bin" on both sides, but Kermit stubbornly transfers the PDF
file as text.
The transfer from Linux to Windows is fine, ...
-
Kermit insists on treating PDF files as text
Desktop side: Kermit 95 version 2.1.3 on WinXP
Server side: C-Kermit 8.0.209 on Linux
Using SSH sessions
I type "bin" on both sides, but Kermit stubbornly transfers the PDF
file as text.
The transfer from Linux to Windows is fine, the uploads are the ones
getting messed up.
-Ramon
-
Re: Kermit insists on treating PDF files as text
On 2007-03-05, Ramon F Herrera wrote:
: Desktop side: Kermit 95 version 2.1.3 on WinXP
: Server side: C-Kermit 8.0.209 on Linux
: Using SSH sessions
:
: I type "bin" on both sides, but Kermit stubbornly transfers the PDF
: file as text.
:
: The transfer from Linux to Windows is fine, the uploads are the ones
: getting messed up.
:
See:
http://www.columbia.edu/kermit/ckermit70.html#x4
As you can see, by default Kermit scans each file to see whether it is text
or binary, and within this default, the default is to scan the first 48K.
Very infrequently -- but mainly with PDF or certain PostScript files -- it
can turn out that the first 48K are, indeed, text, and the "binary" part
happens later, and the default file scan comes up with a false positive
for text.
The cited material shows the many available workarounds. The simplest one,
if you want to control the transfer mode yourself is, is:
SET TRANSFER MODE MANUAL
and then use SEND /BINARY and SEND /TEXT. Other options include using
filenamename associations instead of scanning the file's content, and
increasing the length of the scan.
- Frank