Re: interception proxies

[Vernon Schryver <vjs at calcite.rhyolite.com>]

> From: Joe Touch <touch at ISI.EDU>

> ...
> > The problems draft is interesting and depressing.  All of the problems
> > listed are technical nits.  
>
> This was a choice - in some larger sense, if sourcing other-owned IP
> addresses or TCP connections is considered an architectural problem,
> needs to come down from above, rather than up from WREC. f

Call me a non-team playing scab, but I refuse to the honor the old guild
work rule that limits the questions I can consider.  If sourcing
other-owned etc. or anything else is an architectural or other problem,
then professional pride ought to force one to raise the issue insetad of
waiting for the AD, IESG, IAB, or a plenary to redirect things.  But I
realize that's a minority view, and not just in IETF working groups or
even the IETF.
If only I could have had one pre-IPO share for every time I've been told
"we can't think about that; we'll have to ask to marketing" ...  on the
other hand, most of those shares would be merely expensive wallpaper.



> >                    ... That there is no mention of the problems that IP
> > fragmentation can cause interception proxies is depressing.
>
> The problems of IP fragmentation are not unique to web caching or
> replication proxies. They affect all interception proxies. The issue of
> inteception proxies was around long before WREC, and is more than just a
> caching or replication issue. 

Which is why it was depressing.  Oh, well, perhaps a future version of
the Problems draft will consider that issue and say as others wrote, it's
not a problem and can be fixed with big buffers watching IP ID's, avoiding
UDP, assuming good MSS's or discovery prevent TCP/IP fragmentation, or
whatever.


> ...
> That's the property of WGs in general, by construction. These questions
> sometimes get addressed in BOFs, but there is also often too much
> momentum or political interest in establishing a 'standardizing
> presence' in an area. By the time a WG is formed, the time for 'whether'
> has often passed in favor of 'which'.

Which was exactly the lament the other day.  By the time a Last Call rolls
around, it's months and $B of market cap late to worry about "whether?"
There are always screams about the unfairness of raising "whether" at such
a late date and vague reference to nearly completed implementations that
will have billions of installations by the end of the quarter, or when
IPv8 replaces IPv4 at the latest, exactly as we heard in response to the
initial comments about draft-cerpa-necp-02.txt.

I don't have a fix for the problem, except to steadfastly refuse to heed
cries from partisans that "whether?" is out of order until it really is.

 ...


] From: "BookIII, Robert" <Robert.BookIII at cwusa.com>

] Joe,
]        Am I to presume by your statement that you are of the mind that the
] time for considering whether vs. which has already come and gone? Is there
] anyone on this list who thinks that?

I don't speak for Joe, but it's clearly over.  Unlike the wiretapping
question, interception proxies are too near to the technical interests
(and pocketbooks) of too many IETF participants.  You must admit that they
sound like cool hacks, unlike merely forwarding copies of bits.


Vernon Schryver    vjs at rhyolite.com