Slashdot Mirror
Mozilla's Proposed Conclusion: Game Over For WoSign and Startcom? (google.com)
Reader Zocalo writes: Over the last several months Mozilla has been investigating a large number of breaches of what Mozilla deems to be acceptable CA protocols by the Chinese root CA WoSign and their perhaps better known subsidiary StartCom, whose acquisition by WoSign is one of the issues in question. Mozilla has now published their proposed solution (GoogleDocs link), and it's not looking good for WoSign and Startcom. Mozilla's position is that they have lost trust in WoSign and, by association StartCom, with a proposed action to give WoSign and StartCom a "timeout" by distrusting any certificates issued after a date to be determined in the near future for a period of one year, essentially preventing them issuing any certificates that will be trusted by Mozilla. Attempts to circumvent this by back-dating the valid-from date will result in an immediate and permanent revocation of trust, and there are some major actions required to re-establish that trust at the end of the time out as well.
This seems like a rather elegant, if somewhat draconian, solution to the issue of what to do when a CA steps out of line. Revoking trust for certificates issued after a given date does not invalidate existing certificates and thereby inconvenience their owners, but it does put a severe -- and potentially business-ending -- penalty on the CA in question. Basically, WoSign and StartCom will have a year where they cannot issue any new certificates that Mozilla will trust, and will also have to inform any existing customers that have certificate renewals due within that period they cannot do so and they will need to go else where -- hardly good PR!
What does Slashdot think? Is Mozilla going too far here, or is their proposal justified and reasonable given WoSign's actions, making a good template for potential future breaches of trust by root CAs, particularly in the wake of other CA trust breaches by the likes of CNNIC, DigiNotar, and Symantec?
This seems like a rather elegant, if somewhat draconian, solution to the issue of what to do when a CA steps out of line. Revoking trust for certificates issued after a given date does not invalidate existing certificates and thereby inconvenience their owners, but it does put a severe -- and potentially business-ending -- penalty on the CA in question. Basically, WoSign and StartCom will have a year where they cannot issue any new certificates that Mozilla will trust, and will also have to inform any existing customers that have certificate renewals due within that period they cannot do so and they will need to go else where -- hardly good PR!
What does Slashdot think? Is Mozilla going too far here, or is their proposal justified and reasonable given WoSign's actions, making a good template for potential future breaches of trust by root CAs, particularly in the wake of other CA trust breaches by the likes of CNNIC, DigiNotar, and Symantec?
Why in the hell would anyone trust certificates signed by a Chinese CA to begin with?
The world's burning. Moped Jesus spotted on I50. Details at 11.
It's a system built on trust. If a CA is anything less than completely trustworthy, it's useless. A year long suspension looks like a slap on the wrist, when the obvious action is to drop them completely.
Required reading for internet skeptics
I don't think a big American company would get the same treatment for making the same mistakes.
Long live SSL? Or should we quicken its demise and end the empire of SSL certificate charges?
Who in security thinks they are trustable? Too many breaches of trust and who knows how many covered up ones by state actors?
I don't know these CAs, I don't trust them, and Mozilla shouldn't be in the business of deciding who I trust by default.
...depends upon the flawed root CA system. These companies have repeatedly failed to do their primary job of cooperating with established rules and protocols. They've failed to report breaches, they've issued certificates erroneously for other domains and then not reported it. This has been done repeatedly, and is the PRIMARY function of a CA. I don't consider it "draconian" at all, it seems pretty charitable for their timeout to be only one year instead of permanently. It's also an example to other certificate authorities that the rules actually have some teeth.
Sure it will cause a number of customers pain in having to search for another provider and again yes it will cause major pain to WoSign and StartCom, however in an age where secure communications are becoming more and more pivotal to ensuring a safer experience we need to make sure that issuers are aware that they also have a responsibility to uphold.
Sure if an end-user or a web host fails to secure their communication it should be on them, but when a provider fails to provide a secure service or are acting in a questionable manner then we need to make sure that they are held accountable too!
Look, I realize the population is chunked up into groups of zombie types but...
Those groups can certainly teach such a simple process to their minions.
Just use a simple ssh first try auth scheme.
Self sign certificates and request authorization on the first try. If it changes say hey somethings wrong, the websites or ssl servers can add a message that says it's a new certificate.
You could even go as far as having peer groups for trusted certs, more like a blockchain than a signing authority though.
You could also use something like Electrums hash to keywords thing for letting people keep track of their keys reasonably but it's probably easier to just show them the meta info held in the key saying the url or ssl socket.
Seems simple, I can't think of a good reason a very authority would be made in the first place. Maybe they didn't know or didn't slow up and think or maybe they were malicious. Either way, fix it right.
But you trust your OS vendor then?
What's draconian about not trusting someone proven to be untrustworthy? Is it because their only job was to be trustworthy?
Don't waste your vote! Vote for whoever you want, unless you live in a swing state it won't matter anyways
The CA model is broken.
The fundamental difference between a CA and a web of trust is that in a CA model, only the CA signs your certificate while in a WoT, the certificate can be signed by as many signers as you want, which mean you don't have a single point of failure.
For example StartCom may not be worth your entire trust but it is still better than nothing. And complimented by, say, a few independent, free authorities, it starts getting good because the attacker now have several different targets. This is not an option with CA as we have now, that's blind trust or nothing.
Guess I need to get my certs moved over to someone else. Fortunately there's some other free options that look promising.
https://letsencrypt.org/
Lawyers, MBA's, RIAA? A jedi fears not these things!
How about don't use a CA at all? Self sign your certificates in your organization. Expect everyone you do business with to verify and install your certs rather than trusting Mozilla to trust a third party. Oh... and staff up your help desk to answer questions like, "I didn't need to do this with Amazon. Why are you guys so stupid?"
If you can't remove problematic certs by a vendor or penalise them for misdeeds, then they have no constraints. User trust is more important.
For every problem, there is at least one solution that is simple, neat, and wrong.
In my opinion, this does not go far enough. These entities are in the business of trust. Once you break that trust ONCE, it should be game over! No warnings, slap on the wrist, suspensions or other nonsense. You break that trust and you should be removed permanently.
what about the whole Bluecoat thing? Or when the other big CA's did wrong? It's just an issue when it's a non-U.S. based CA, is it?
go into advanced settings and distrust those to CA's, it takes less than a minute.
Snowden and Manning are heroes.
Is Mozilla big enough (in the form of Firefox) for the rouge CA in question to care what Mozilla does? I've no idea whose numbers are reliable, but the first set I found indicated that Firefox has less than an 8 percent share of the browser market, with IE @ ~27% and Chrome @ ~53. If that's even close to true, is Mozilla taking an action like this relevant? Or will it just push people into dropping Firefox?
A thousand pounds of wood moving at 300 feet per minute. Don't get in the way.
Which will impact less then 8% of the market: https://www.netmarketshare.com...
please read the article... only the NEW certs will be distrust, old ones will keep valid until they expire. You might have problems only on renews...
If they behave well and follow all the rules, in one year they may be trusted again... if they keep trying to issue certs using past dates, they will be totally removed and if they ever try to reenter the CA business, they will have to follow again all the audits, tests, checks, etc... takes ages, log of money and in the end, mozilla can still say "NO"
Higuita
I thought the 'punishment' was an interesting take to show a loss of trust, after a certain date and the ability to regain it after a period of time. I found it slightly more interesting that Mozilla would also choose to no longer accept audits conducted by Ernst & Young. That could potentially be huge as it shows (at least in some manner) that their auditors were not conducting a thorough audit or did not have the technical prowess to fully audit a CA.
think before you write, it'll save me moderator points.
It's not just a few CAs, it's the whole system. The CA system is built on trust and there has been no trust left in the system in years. The whole idea of encrypted communications between web browsers and web servers needs to be reworked and somehow decentralized so that rogue CAs will eventually die out.
-SR
The certificate business is big money. It's possible some companies may be able to purchase certs from multiple vendors but it adds up very quickly, and coordinating activities like expiration dates have to be aligned among the vendors which is tricky with multiple large contracts. Only the biggest companies will be able to do this, leaving the rest to single and/or smaller CAs.
Yet does that really make an entity's presence on the public Internet inherently more trustworthy? If I was to get certs from Verisign, Thawte and Let's Encrypt, that's not saying much since Let's Encrypt does DV and not EV certs. If you have a breach of one CA but not the other, who do you trust and why? What does that result even mean? Best two out of three or three of five? It's not entirely out of the realm of possibility that smaller CAs could be simultaneously compromised, which is why the larger companies mostly go to that company based in Northern Virginia that has been rock solid if nothing else.
I think smaller lesser-known entities like these Chinese CAs will be perpetually more risky to obtain certs from. It's just what it is. As you go up the chain the certs get progressively more expensive but more trusted as well. As long as there is a commercial interest in selling certs, I don't think the current situation will change. It's just another warning just like Diginotar and others have demonstrated and Mozilla is IMO being overly lenient and perpetuating the problems currently supporting the "list of trusted CAs in the browser" model.
They may be rouge, but their money's green!
Take a look at Mozilla's trusted CAs. It is a joke. They need to be harsher. First abuse cut them off. It also needs to be easier for users to remove trusted CA's from Firefox.
Read the account of how WoSign handed out the key to githib.
As if WoCom and Startcom are any less trustworthy than the rest of the despicable commercial CA signers.
In an industry where trust is essentially the product, and critical to the system, Mozilla should have permanently banned them along with a lifetime ban on the executive level management. Punishment for abuse of the trust system should be harsh if an independent audit shows wrongdoing.
If you disagree, please post your argument. (-1, Overrated) isn't your personal censorship tool for views you don't like
Reading through the incidents most of them are bugs or glitches with no malice intended, or at least no obvious malice with a little bit of neglect, all of which you could also claim about Mozilla browsers. Is there something not listed on the linked page to justify blocking them? By the same standard should users all be stopping using Mozilla browsers for the next 12 months till they are not the most insecure browser on the net?
If browser vendors were really serious about certificate security, we would have RFC 6698 as standard in browsers already.
Change is certain; progress is not obligatory.
I removed trust for all Startcom certificates on my machine soon after reading this piece. I did not find WoSign roots, so I'm guessing their path leads up to Startcom.
One of these days I will have to spend some time and find a list of the "good" certificates (track record of zero policy breaches), and only leave those on my machine.