[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: lynx-dev thoughts on cookie file w/multiple sessions
From: |
brian j. pardy |
Subject: |
Re: lynx-dev thoughts on cookie file w/multiple sessions |
Date: |
Wed, 18 Nov 1998 08:56:03 -0800 (PST) |
On Wed, 18 Nov 1998, Leonid Pauzner wrote:
> > As things are presently:
>
> > Two sessions, A and B are running. Both load the same set of cookies
> > from COOKIE_FILE.
>
> > Session A gets a new cookie from some site, then exits, writing all
> > the original cookies and the new cookie to COOKIE_FILE.
>
> > Session B quits, overwriting COOKIE_FILE with all the original
> > cookies, but blowing away the new cookie that session A received.
>
> > So here's my thought -- prior to writing cookies in memory to COOKIE_FILE,
> > Lynx should *reread* COOKIE_FILE, ignoring any cookies that are already in
> > memory, but adding cookies from the file that are not in memory.
>
> > The only thing coming to mind so far that could be a problem here would be
> > getting two cookies with the same 'name' attribute from a server in two
> > different sessions (two unique identifiers from advertising sites?), in
> > which case the one which is written to COOKIE_FILE first (thus read into
> > memory in the other Lynx session, overwriting the existing one) will be
> > maintained, and the others dropped.
>
> Remote server may change cookie expiration date,
> which date should be correct from session A or from session B?
With the way I have things up there, whichever copy of Lynx is exited first
will be the session with the cookie that is saved.
So it's pretty much an even shot either way. I don't like this, and can't
really think of a way to do handle this without (for example) timestamping
individual cookies (possibly with comments in the cookie file?).
The more I think about it, the more appealing a directory to store cookies
sounds.
--
GPG & PGP public keys: <URL:http://www.psnw.com/~posterkid/keys/>
PGP fingerprint: 42 57 B3 D2 39 8E 74 C3 5E 4D AC 43 25 D2 26 D4