Slashdot Mirror
Opera Screeches at Mozilla Over Security Disclosure
The Register is reporting that Mozilla's handling of a recent security exploit that affected both browsers has drawn an unhappy response from the Opera team. "Claudio Santambrogio, an Opera desktop developer, said the Mozilla team notified it of a security issue only a day before publishing an advisory. This gave the Norwegian software developers insufficient time to make an evaluation. [...] Santambrogio goes on to attack Mozilla's handling of the issue, arguing that it places Opera users at unnecessary risk."
At least Mozilla told them of the issue. I personally don't think it's their ultimate responsibility. Definitely obligated to do something... but imagine the kind of action Opera would have if Microsoft found the security flaw.
to fix the exploit wins!
:(){
As far as I can tell, Firefox had a flaw, they fixed it and notified Opera that they had the same flaw the day before Firefox's fix was announced. Sounds to me like the only thing that Firefox did wrong was notice that it affected Opera at all, because if they hadn't Opera would have been left with egg on their face and nothing to bitch about.
Let's imagine that the Mozilla developers had modified the release notes for 2.0.0.12 so that it wasn't obvious what they'd fixed. Would that have been any better? Of course not. I can grab the code, diff against 2.0.0.11, take note of the changes, and presumably figure out why they were made. Now I can craft a working exploit against 2.0.0.11. After testing it on Firefox, what's the first thing I might try? How about... see if other browsers have the same problem?
So keeping in the fix but not mentioning it in the release notes is out. What, then... not patch the flaw? Yeah. Right.
Opera might be a nifty browser, but apparently its authors are whiny bitches.
-=rsw
From the Opera developers' description it appears that the Mozilla foundation could have handled things more professionally - Opera was only notified the day before a public advisory was published, and since that time the Mozilla foundation have opened most of the bug reports containing exploitation details to the general public. Judging from the emoticons on Opera's blog, the latter action by the Mozilla foundation is the primary issue here, not that they published the advisory.
And so we go, on with our lives
We know the truth, but prefer lies
Lies are simple, simple is bliss
Seems if they'd kept their whiny mouths shut, nobody would have realised from the vulnerability disclosure that the issue affects Opera. Now EVERYONE knows, from the kiddie scripting 'sploits to the IT manager planning the software deployment for the next few months, who is now seeing why closed-source Opera isn't really such a great choice after all. Even the CVE entry doesn't disclose Opera's vulnerability to this bug. Still, it makes good comedy if nothing else...
Resistance is futile. Reactance buggers it up.
I think we all know already that disclosing the exploit is what brings the motivation to fix the hole.
The fact that they hid the bug reports at all should be enough to make the Opera kids grateful. After all, the Mozilla foundation operates in a pretty open and transparent fashion. The most honest (and destructive) way to go would be to never hide the bug reports.
But just to cover that old ground once again; when code changes, diffs happen automatically, and people know just precisely what changed. You can be sure that some of those people are malicious hackers looking for new ways to screw us all; there's good money in it. So by hiding the details of the exploit, you make sure that only the more skillful and malicious hackers have the exploit. Does that sound like a good idea to you?
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
Best episode of Oprah ever!
Tsukasa: All I really want, is to be left alone...
I don't see how expressing dipleasure at something on a blog is an overreaction. "Screeching" is stretching it pretty fucking far, since it's basically saying what happened. Where in the blog entry is there screeching, perhaps the bold on "responsible", or maybe the ":("? Wouldn't it be better to link to the blog entry directly and not some dumb opinionated elreg article? Really, did you even read the original source before deciding "the developer needs a chill pill"?
At the end of the day, Mozilla would have acted better by keeping the exploits closed for a few more days, as they would hope anyone else would do for them. By not doing so, they upset people, and others expressing that upset is perfectly understandable. There's no mass outcry at Opera, no press release or open letter saying the Mozilla team are dicks, there's a few words saying what happened and a couple of emoticons on a developer blog entry.
You know, looking at Mozilla's release, they didn't seem to mention anything to anybody about Opera having a problem too. Looks more like Opera screwed themselves.
Unless for some reason they use the same engines, what's the problem with this practice? Opera's security isn't Firefox's responsibility. The fact that they notified opera at all went above and beyond what they needed to do, and asking firefox to be less open with their community is asking them to risk their image for the sake of opera and its users. Unless I'm missing something here, Firefox was being polite and Opera's throwing a world class hissy fit.
But never hiding bugs is silly. For example, if you provide an strace of ssh crashing, you'd want to mark that private at least.
Maybe, maybe not. You never know what the black hats already know; as a _user_ of ssh, if you disclose then I can take steps to limit damage--e.g. if I'm allowing full ssh access from outside my network (so that employees can work on the go), I may decide that the small benefit of doing so doesn't merit the risk. I'd rather turn off external ssh access for a few days until there's a fix.
When you hide the bug, you're hiding the ability for the users to take steps to protect themselves. You're forcing me to run with exposed systems for several days, and hoping that nobody "bad" knows about the bug. And you're making that judgement for your users rather than giving them the ability to make that call themselves; that's almost impossible given that the judgement might hinge heavily on whether I'm a large financial institute or a personal blog site that backs up daily. Just guessing that most users are happy with your security through obscurity is bound to be wrong in some cases, and those cases are likely to be some of the more financially significant ones.
(That's on top of the pressure to issue a real fix that full disclosure brings. Before things like BugTraq, it was common for people to sit on severe security bugs for literally _years_.)
rage, rage against the dying of the light
Now, wait a second. If I am developing software package "A", and you develop competing package "B", and I find a hole in A and fix it, then just for laughs test to see if your product has the same hole and then I am kind enough to let you know that it does, then I announce that there is a hole in A, how am I responsible for the security of B at all? I've done you a favor by performing the test and giving you a heads up in the first place! I don't owe you anything.
I think we all know already that disclosing the exploit is what brings the motivation to fix the hole. You haven't given a specific example of Opera needlessly hiding an exploit.I'm not sure what you think that has to do with anything. The Mozilla foundation didn't even announce to the public that there was a hole in Opera. The announcement is that there is a hole in Firefox. Why not try reading the advisory? There is NOTHING in there about Opera's susceptibility. You can't even view the bug report without a Mozilla bugzilla account with the proper access - I just logged into my account, and that doesn't include me, so it's not like even the report is generally available. Also, as per the advisory:
So it seems as though the Opera team has had some warning about problems similar to these in the past - along with the rest of the world.
Could I find and fix a bug in one of my pieces of software in a day? Probably, because all of them are very simple. If I had a development team and a security response team (they do have one of those, don't they?) then I bet "I" could find and fix known security problems in larger software products in a day, too.
Actually, a number of security holes in the Linux kernel have been found, announced, and fixed on the same day, now that I think of it.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
Somebody posting to Slashdot says that somebody at The Register says that an Opera blogger screeches about Mozilla. Even for Slashdot, this is a pretty weak title.
:) ), but not happy that there was only a day before it was made public. Nobody is particularly happy when they only have a day from learning there's a security hole to everybody else learning about it, thats not enough time to get a fix rolled out, so this is hardly surprising.
What they actually say is that they only had a day between notification and public disclosure. He's actually happy that Mozilla told them at all (hence the
I know Mozilla can do no wrong around here, but come on. Even the Mozilla devs would be happier getting more then one day before public disclosure of a security hole.
-- "So they told me that using the download page to download something was not something they anticipated." - Bill Gates
They've had twelve days to fix it. Have they? If you RTFA, you'll see not only have they not, they've expended a greater amount of energy trying to whip up support for their malcontent with Mozilla. So, in reply, yes it does seem that they would rather cover this up than fix the issue in a timely manner. Their actions scream it, even if TFA doesn't.
Resistance is futile. Reactance buggers it up.
The problem usually isnt coding time. It's organizational response and resource allocation issues.
For example, Opera is on a very differen timezone from the US, so initial publication may happen overnight from the POV of the Opera staff.
So then a day starts. When people start their day, they have a pile of things to respond to. The incoming messsages have to be triaged. Someone has to make a decision that this is important enough to escalate or take action on.
Then you have to find people with the capability to test whether its a real problem. This may take a couple hours. People go on vacation, get sick, etc.
Then you have to take the time to do the research, test whether this is a real problem, what versions it affects, etc. This takes a couple hours.
Then yuou have to stop a coder from working on something else, bring them up to speed on the problem (if its not the same person doing the testing), and get them started on the fix.
Then even with a fix you have to do regression tests. Not sure about Opera, but many mature apps have full test suites that can take a couple hours.
Then you have to write release notes, update the web page, do a new deploy package, and update your update servers to notify Opera that there is a new update.
As you can see, very little of the time here is coding.
Many large orgs have taken steps to create a 'short path of decision making' to streamline this process, always have one coder on call who can do this work, etc. But even then if anything is out of whack or the wrong person is sick or on vacation or on another urgent item, a whole day could pass without response.
No one is suggesting that Mozilla should have delayed the fix (in order to hold back disclosure).
No, it would have been open and responsible and good if someone at Mozilla had thought to send an email to the Opera dev team a week or two ago saying:
Roses are red, violets are blueWe're fixing this exploit and think you should too.
Lots of Love,
Your secret big red monster Valentine.
No need to coordinate releases, but given that it took them a while to patch it, they should assume it'll take Opera a wee while to, and in the meantime they're leaving members of the public open to exploit.
Members of the public that used to use Firefox, but had to stop because Mozilla never fixed the memory leak and these users were using old machines (NT4, 32 meg RAM) and Open Source was supposed to mean never being obsolete, but it was only the non-open, free Opera browser that offered me a fully-patched, fully working browser.
HAL.
Got them moderator blues I blieve I walk out the do', With these mod-points I been gettin', I 'most never post no mo'