On a fax document, the date/time and ID of the sending machine are part
of the transmitted image; added to the image by the sender.
In keeping with current fax, I would think that we still want this
information as part of the image. So I guess I think that answer is "no,
the information in the dictionary is not quite enough". If the doc is
printed and the soft copy discarded then this information is lost.
The digital signature does validate the data from the originating
sender, but what if the receiving user decides to try to re-fax what he
just got? Do we replace the signature and the URI stamp?
A digital Signature gives you a tamper-proof way to time-stamp a
document, but besides that...
What about using the information in the "Document Information
Dictionary" (See Table 9.2 in the PDF Reference 1.4)? This dictionary
is already part of the PDF/is spec. and should be all that is needed.
Of course, it's not tamper proof unless the document is also digitally
Is this what you were looking for?
It's occurred to me that with PDF we have level of portability that we
didn't have before. However the IFX spec requires that the sender add
the URI of the sender:
The Sender MUST place the Sender's URI, i.e., the value of the
"sender-uri" attribute (see section 8.3), along with the date
and time, in one of the following places, DEPENDING ON
1. On a cover page automatically generated by the Sender that is sent
before the rest of the document.
2. Merged with the first page of the document.
3. At the top of every page of the sent Document.
The Sender MAY include additional data (Sending User, Receiver
identity, etc.). As for regular FAX, it is RECOMMENDED that
this information be represented as bit map data, so that it is
more difficult for it to be modified before it gets to the
My first thought was to add an optional field to the PDF that indicated
where the Stamp was located. If the doc was resent, then the second
sender could replace the Stamp with its own stamp. But if the doc was
digitally signed then either the Signature would be lost or invalidated.
Anything thoughts on how we should handle this?
This archive was generated by hypermail 2b29 : Mon Feb 03 2003 - 16:13:16 EST