-----BEGIN PGP SIGNED MESSAGE-----
Mark J. McArdle wrote: <totally ad hominem attack from a PGP staffer
omitted>
I have no wish to post off-topic to this group. Mark's misrepresentation
of my concern is what is off-topic. Fairness and a setting of the record
straight requires a response.
To put it simply and directly, by negating RSA in PGP 5.53, PGP
invalidated all the prior RSA web of trust signatures and keys for users
of the new version and created either a split in the web of trust or a
situation of forced conversion and re-signing for those who wish to use
the latest version. This seems both incontrovertible and unarguable, and
ad hominem attacks on those who ask about it are non-responsive. My
concern is to see that Open PGP is robust over such "improvements" to
algorithms. The current worked example by PGP Inc. is relevant to
understanding the issue.
As to the lies which have gained strong currency, I need to say a word
to clear the air and in fairness. I am a PGP user. I was one of the
earliest PGP users until I found it was infringing. I then worked
behind the scenes (as Jim Bidzos will confirm) to try to get him to
license RSA for PGP. That effort initially failed apparently because of
what Simson Garfinkel describes in his book on PGP as Phil's
intransigence, not because of any lack of effort on my part or Eric
Hughes part or even Jim Bidzos' goodwill. Subsequently I was one of the
first adopters of MIT PGP and both Jeff Schiller and Derek Atkins
signed my key. I believe PGP has its strong place when web of trust is
appropriate to the circumstances, and rigid heirarchical (as practiced
with S/MIME and CAs such as Verisign in Netscape and Microsoft
Explorer) has its place. I believe there is no all-purpose solution to
trust models that won't either lose some of the advantages of web of
trust, or some of the advantages of rigid heirarchical. Some discussion
on this point by all-or-nothing fans has also led to attempts at
demonization and marginalization when substantive counter-arguments
were unavailable to the proponents. In particular, I believe we need
two solutions for the Internet--one of each trust model, and I
understand this group and the S/MIME 3 group are under instructions to
coordinate in areas of commonality.
I remain a PGP user as well as an S/MIME user. I take strong exception
to some acts of PGP Inc. and PRZ but those have nothing to do with this
group except as they affect such things as the robustness of Open PGP
over algorithmic improvements--an important topic--or as they might
affect true IETF configuration control. I do not plan to raise prior
concerns here except as they may be relevant to Open PGP. Some who hold
either Phil or PGP Inc. as an avatar have used my criticisms of some
acts of PRZ or PGP Inc. elsewhere as an occasion to demonize or
marginalize me when they were unable to respond to such substantive
criticisms with rational counter-arguments. Shoot the messenger remains
a popular sport in some quarters.
Finally, with respect to this group my interest is solely in making what
contribution I can to Open PGP's robustness and independence. I take
very seriously the concept of IETF configuration control, and if it IS
going to be a pacing principle, it must be complete and effective, and
not window-dressing for competitive advantage-seeking by PGP Inc. (in
the case of Open PGP) or RSADSI (in the case of S/MIME 3).
David Sternlight, Ph.D.
Mark J. McArdle wrote: <totally ad hominem attack from a PGP staffer
omitted>
I have no wish to post off-topic to this group. Mark's misrepresentation
of my concern is what is off-topic. Fairness and a setting of the record
straight requires a response.
To put it simply and directly, by negating RSA in PGP 5.53, PGP
invalidated all the prior RSA web of trust signatures and keys for users
of the new version and created either a split in the web of trust or a
situation of forced conversion and re-signing for those who wish to use
the latest version. This seems both incontrovertible and unarguable, and
ad hominem attacks on those who ask about it are non-responsive. My
concern is to see that Open PGP is robust over such "improvements" to
algorithms. The current worked example by PGP Inc. is relevant to
understanding the issue.
As to the lies which have gained strong currency, I need to say a word
to clear the air and in fairness. I am a PGP user. I was one of the
earliest PGP users until I found it was infringing. I then worked
behind the scenes (as Jim Bidzos will confirm) to try to get him to
license RSA for PGP. That effort initially failed apparently because of
what Simson Garfinkel describes in his book on PGP as Phil's
intransigence, not because of any lack of effort on my part or Eric
Hughes part or even Jim Bidzos' goodwill. Subsequently I was one of the
first adopters of MIT PGP and both Jeff Schiller and Derek Atkins
signed my key. I believe PGP has its strong place when web of trust is
appropriate to the circumstances, and rigid heirarchical (as practiced
with S/MIME and CAs such as Verisign in Netscape and Microsoft
Explorer) has its place. I believe there is no all-purpose solution to
trust models that won't either lose some of the advantages of web of
trust, or some of the advantages of rigid heirarchical. Some discussion
on this point by all-or-nothing fans has also led to attempts at
demonization and marginalization when substantive counter-arguments
were unavailable to the proponents. In particular, I believe we need
two solutions for the Internet--one of each trust model, and I
understand this group and the S/MIME 3 group are under instructions to
coordinate in areas of commonality.
I remain a PGP user as well as an S/MIME user. I take strong exception
to some acts of PGP Inc. and PRZ but those have nothing to do with this
group except as they affect such things as the robustness of Open PGP
over algorithmic improvements--an important topic--or as they might
affect true IETF configuration control. I do not plan to raise prior
concerns here except as they may be relevant to Open PGP. Some who hold
either Phil or PGP Inc. as an avatar have used my criticisms of some
acts of PRZ or PGP Inc. elsewhere as an occasion to demonize or
marginalize me when they were unable to respond to such substantive
criticisms with rational counter-arguments. Shoot the messenger remains
a popular sport in some quarters.
Finally, with respect to this group my interest is solely in making what
contribution I can to Open PGP's robustness and independence. I take
very seriously the concept of IETF configuration control, and if it IS
going to be a pacing principle, it must be complete and effective, and
not window-dressing for competitive advantage-seeking by PGP Inc. (in
the case of Open PGP) or RSADSI (in the case of S/MIME 3).
David Sternlight, Ph.D.
-----BEGIN PGP SIGNATURE-----
Version: 2.6.3
Charset: noconv
iQCVAwUBNH8FEkwgH+NYrQ81AQEE6AQAg1NJ2qJilE98sqY40xgX8zsZDxuFL+09
S1OnDiz9TR4kgz3j8IX0caE2jJpEbdRM4mu53FP4cNwCVXW4zUNHBQIZ0VT45NzQ
0Xs0ucJcRqu06p5Epj9/rHNy2x3JIMhvNBEpHK0ZeLyTTYrfEIghUoB7CeigTxi0
tXVJk3YPM0A=
=8qJq
-----END PGP SIGNATURE-----