Re: Need test data

On Mon, 11 May 1998, David Kilzer (210334) wrote:

> 1. Some mail software attaches a "file" with message attributes. I
> would consider these MIME attachments as "garbage" and not bother
> saving them into a file.

Well, we can always add support for different content-types. I could imagine hypermail offering different programs to take care of different content-types, in a style similar to mime!! :-)

> BeyondMail
> Content-Type: application/octet-stream; name="ATTRIBS.BND"

If they use a generic type like that, it'll be impossible to know what kind of data that's in there.

> 2. Mail clients that send TWO copies of the same message...one as
> text/plain and one as text/html. If you trust the HTML encoding,
> this can be a Good Thing as most of your work is done for you.

The current version will include both versions, and yes, it will trust text/HTML attachments. (Of course it could be a subject to make an option for.)

> The tricky part is knowing when this feature has been activated or
> deactived for a given message.

I'll see if I can come up with some more details around this...

> 3. And finally, if you're going to use the attachment's real filename,
> note that some attachments may not have filenames, so make sure to
> have a backup naming scheme just in case.

Of course. It needs not have a name, the name may be illegal for the system (due to illegal letters or length), it may already be in use because of multiple attachments using the same name, etc. All those are reasons why the current version uses random names for all stored attachments. Another reason, not mentioned before, is that the file name that the articles is saved as is not yet determined at the time when the attachment is saved...

--
             Daniel Stenberg - http://www.fts.frontec.se/~dast
   ech`echo xiun|tr nu oc|sed 'sx\([sx]\)\([xoi]\)xo un\2\1 is xg'`ol