On Mon, Jun 4, 2007, [EMAIL PROTECTED] said: > > On 6/4/2007, "Michael Monnerie" <[EMAIL PROTECTED]> > wrote: > >>On Montag, 4. Juni 2007 Tom Allison wrote: >>> So if I add one line at the end of a word doc, this will be >>> considered to be the same file unless the file format includes >>> something unique within the first 8MB of the file? >> >>No, as the size is also used for the primary key. >> > > But I thing I made my point... > Appending or Editing documents can cause problems. > > trons100:/home/rzlsnk $ ls -l test1 test2 > -rw-r--r-- 1 rzlsnk gm 32 Jun 4 12:57 test1 > -rw-r--r-- 1 rzlsnk gm 32 Jun 4 12:57 test2 > > trons100:/home/rzlsnk $ cat test1 test2 > Any users may request that they > All users shall be required to
Indeed, this seems pretty obvious. Whomever had pointed out that the hashes are all streaming functions, and therefore we can generate the hash on-the-fly while the messgae is arriving is also quiet correct. The thing I'm worried about is accurately catching MIME boundaries on the fly, at least not without some major refactoring of how messages are currently received. Anyhow, I think on this thread we're at the point where it's an implementation detail and not a major issue. We could argue forever about which hash to use, but imho, as long as there's some sanity in the decision, whomever writes that code gets to decide what gets written. Aaron _______________________________________________ DBmail mailing list [email protected] https://mailman.fastxs.nl/mailman/listinfo/dbmail
