The NGP moderation team will shortly be upgrading the PGP signing key
for news.groups.proposals. The current signing key is a 1024-bit DSA key created in 2006, which is now considered weak. It will be replaced with
a 2048-bit RSA key that has been signed with the current key.
The upgrade is planned for 24th November 2025, i.e. in one month's time.
Wouldn't it be better to use a 3072-bit RSA key nowadays? As theIs there any reason not to use rsa4096?
rationale for the change is security weakness, a 3072-bit key will
last a bit longer. 2048-bit RSA keys are expected to be weak near 2030 according to NIST SP 800-57 Part 1 Rev. 5 publication released in
2020. See table 4, page 59, about security strength at:
The NGP moderation team will shortly be upgrading the PGP signing key
for news.groups.proposals. The current signing key is a 1024-bit DSA key
created in 2006, which is now considered weak. It will be replaced with
a 2048-bit RSA key that has been signed with the current key.
The upgrade is planned for 24th November 2025, i.e. in one month's
time.
Wouldn't it be better to use a 3072-bit RSA key nowadays?
As the rationale for the change is security weakness, a 3072-bit key
will last a bit longer. 2048-bit RSA keys are expected to be weak
near 2030 according to NIST SP 800-57 Part 1 Rev. 5 publication
released in 2020. See table 4, page 59, about security strength at:
https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-57pt1r5.pdf
Wouldn't it be better to use a 3072-bit RSA key nowadays? As the
rationale for the change is security weakness, a 3072-bit key will
last a bit longer. 2048-bit RSA keys are expected to be weak near 2030
according to NIST SP 800-57 Part 1 Rev. 5 publication released in
2020. See table 4, page 59, about security strength at:
Is there any reason not to use rsa4096?
Hi,
The NGP moderation team will shortly be upgrading the PGP signing key
for news.groups.proposals. The current signing key is a 1024-bit DSA key created in 2006, which is now considered weak. It will be replaced with
a 2048-bit RSA key that has been signed with the current key.
The upgrade is planned for 24th November 2025, i.e. in one month's time.
Wouldn't it be better to use a 3072-bit RSA key nowadays?
As the rationale for the change is security weakness, a 3072-bit key
will last a bit longer. 2048-bit RSA keys are expected to be weak near
2030 according to NIST SP 800-57 Part 1 Rev. 5 publication released in
2020. See table 4, page 59, about security strength at:
https://nvlpubs.nist.gov/nistpubs/SpecialPublications/NIST.SP.800-57pt1r5.pdf
Naturally, the Board can go on with a 2048-bit key now and generate a
new key in 2030 :)
Personally I doubt the ability to forge Usenet control messages is worth millions of dollars to anyone.
Even if someone did crack the key, it wouldn't let them do much except briefly disrupt this group. All we use the key for is to confirm that
posts to NGP were properly approved by a moderator, so that we (or
anybody else who wants to) can detect if someone tries to forge approved posts to the group.
This seems like a lot of work for a very
slim chance of doing something that, as you say, isn't going to be worth anything close to the likely cost of doing it.
You're right, a bigger key would be more secure and last longer.
However, since I already made the announcement, it's probably less
confusing to continue with the upgrade as announced, and upgrade the key again later when it becomes necessary.
NGP is also not, I suspect, a high-value target :) Being able to forge a signed posting to the group won't compromise national security or makeSo true!
money for anyone, so I doubt there are adversaries willing to throw huge amounts of computing resources at cracking the group approval key. I
don't think we even have any particularly motivated kooks these days...
Incidentally, if a post is forged in this newsgroup, will a cancel
article or a NoCeM notice be sent automatically?
As they contain an X-Auth header field mentioning PGPMoose, maybe it is
the case. Is the group monitored by Tim's bot? (I don't recall having
seen any NoCeM article from his bot since a long time; maybe there had
not been any forgery for ages? or the bot is no longer active?)
In article <mmfdh6F7n0lU1@mid.individual.net>, iulius@nom-de-mon- site.com.invalid says...
Is the group monitored by Tim's bot? (I don't recall having
seen any NoCeM article from his bot since a long time; maybe there had
not been any forgery for ages? or the bot is no longer active?)
I'm not sure if Tim's bot is still active or monitoring this group, but
I've just sent him a email to make sure he knows of the planned change.
The NGP moderation team will shortly be upgrading the PGP signing key
for news.groups.proposals. The current signing key is a 1024-bit DSA key created in 2006, which is now considered weak. It will be replaced with
a 2048-bit RSA key that has been signed with the current key.
The upgrade is planned for 24th November 2025, i.e. in one month's time.
A reminder that the PGP signing key for this group is scheduled to
change on 24th November (Monday).
This is only important if you're verifying posts to NGP to check for
forged approvals; most readers of the group will have no reason to
notice or care.
The new public key is available from:
- The first post in this thread
- The PGP keyserver at pgp.mit.edu
- The Big 8 Management Board wiki at
https://www.big-8.org/wiki/Signing_key_for_news.groups.proposals
The email address ngp-approval-key (at) ngp.big-8.org will auto-respond
with the current key until the changeover happens, and will thereafter respond with the new key.
| Sysop: | Amessyroom |
|---|---|
| Location: | Fayetteville, NC |
| Users: | 74 |
| Nodes: | 6 (0 / 6) |
| Uptime: | 45:35:11 |
| Calls: | 1,100 |
| Files: | 1,339 |
| Messages: | 275,479 |