Slashdot Mirror
Critical Mozilla, Thunderbird Vulnerabilities
d3ik writes "An advisory has been issued on several buffer overflow exploits in the Mozilla and Thunderbird code. Coincidentally, one of the exploits takes advantage of a unchecked buffer in the bitmap parser, very similar to recent Microsoft JPEG vulnerability.
The good news is that if you have an updated version (Mozilla 1.7.3, Firefox 1.0PR, Thunderbird 0.8) you won't be affected."
Afterall, it's Microsoft's fault when their users don't keep up to date with security patches.
"Ask not what your country can do for you." --John F. Kennedy
.....you can patch without fear of breaking a gazillion programs.
-Randy
Here's why:
Software is written by humans. As a result, mistakes are bound to be made. Various software design strategies merely mitigate and minimize those risks, but it's bound to happen. This is a fundamental fact of life. Deal with it.
However, OSS permits investigation and transparency in the resulting software. This leads to better code reviews (hopefully) and more bug fixes. In addition, there is nothing that a software development team or company can hide behind (a la IP rights) all the while shouting, "Shut up! Shut up! I can't hear you! la la la la!"
If I use Internet Explorer, I can deploy patches to every amchine on the domain automagically using software like Shavlik's HfNetChk - with Moz I'd have to take a trip round the desktops, forty or fifty upgrades is something I don't fancy.
The Moz team should be looking with urgency at how corporate customers can keep it up to date - I'm sure that would also make it a much easier sell to business.
Except the similar MS bug is already patched. And yet people were still quite pissed about it a few hours ago.
I cannot ask my father to uninstall his browser and reinstall a new one every so often. If Firefox wants to be accepted by the large crowd out there it definitely needs an automatic update.
michael at slashdot.org: The real answer is that a couple of the slashdot authors are sick.
I wasn't notified of this critical vulnerability until I checked slashdot. Perhaps FFox/Moz should have a feature that automatically checks for updates and recommends them appropriately?
I hate to download yet again all 11 megabytes just because of a single bug.
The owls are not what they seem
And here's the additional difference:
We're going to fix this Firefox bug, and it doesn't matter if it wipes your preferences and breaks your extensions. Your loss for using beta software.
We're going to fix this IE bug and try to make sure it doesn't break existing installs.
I use Firefox, but haven't upgraded from 0.8. I got tired of having to reset my preferences and extensions with each update. I'll take the time to upgrade when it gets to 1.0.
Yeah, that is a loss of using beta software. If you're using firefox you're a beta tester, which comes with all sorts of drawbacks like that.
They're at the stage where they make large sweeping changes quickly. Once they hit production they should no longer do that... but until then, it comes with the terroritory... personally I'm amazed, and think it speaks greatly to the quality of Firefox and the lack of quality of IE that Firefox has such a showing in a beta state.
Well actually buffer overflows are inherent problems in C/C++ because they allow programmers to make those kind of errors.
Java on the other hand does not allow programmers to make that error. If more people used better tools it would mean less security problmens.
Just saying it like it are.
If you RTFA, and scroll to the botttom, you'll notice they link to all of the relevant Bugzilla entries for the reported problems.
Read them. Do you know how these flaws were found? By people looking at the source code and reporting them. The people who detected the problems couldn't have found them if the source was closed.
This is Open Source at its finest. On the other hand, we have the flaws in IE that are all too often found after someone has created an exploit and it's in the wild.
Personally, I wouldn't mind one bit if Mozilla users and Open Source developers found a security problem once per hour and got the problem fixed quickly. It's vastly better than the closed-source alternative where you have to hope that someone without access to the source reports the fault when they find it, and that Microsoft doesn't take their own sweet time fixing it.
Once again, Open Source at its finest.
Yaz.
The safest and best thing is to use a real VM, like the JVM. Another alternative is to use something like Cyclone which also doesn't allow unsafe memory operations.
To all the ditto-heads who keep on saying "if it's not in C, it's too slow", wasn't there just an article on Slashdot a few days ago about full-motion video players written in pure Java? Surely a jpeg here and there shouldn't be too much of a problem?
so when are we going to be able to update firefox/thunderbird without reinstalling the entire app? I'm sick and tire of this because I also have to reinstall every single extensions and themes I use. Sure I can do this easily, but it's a pain in the ass when I have to tell my not-so-tech-savvy friends to upgrade. it's tedious and stupid. and god bless those poor souls who have to upgrade a whole network of machines.
mozilla.org really needs to include a link to their Security Centre on their front page.
As FireFox and Mozilla become more widely used, we will truly see how well the open source community can keep up. After all, I honestly believe that the reason more bugs and fulnerabilities are found in IE is that it is more widely used.
I see the day not too far off when FireFox could overtake IE in the market...so will the majority of problems then be in FireFox, or is microsoft really writing bad code? It will be interesting to see.
I believe the open source community will be up to the task of maintaining the bugs as they come in, but I think we will see that there will still be a lot of these types of serious problems that crop up once there are thousands of people dedicating their lives to exploiting them.
Grab a chair, sit back and watch the fun.
"The good news is that if you have an updated version (Mozilla 1.7.3, Firefox 1.0PR, Thunderbird 0.8) you won't be affected."
And the good news is if you have the updated version of Windows (Windowws XP SP2) then you aren't affected by the similar critical flaw either but it's different when it's OSS huh?
We've found a bug in firefox, we're really sorry. Anyone using old versions of firefox will be affected.
We've found a bug in internet explorer, we're really sorry. We'll fix it... eventually.
The only difference here is when they decided to announce the flaw.
Mozilla decided to keep it secret until a new version was released. Don't you find that at least slightly scary?
Look when this security exploit was filed: #226669.
Beware: In C++, your friends can see your privates!
All those critical bugs have been detected by reviewers from the "Security Bug Bounty Program", as described on mozilla.org. The Mozilla Foundation has offered a $500 bounty for each security bug found, and already has secured a $10,000 budget to do so.
Thus, all those bugs should not be seen as a proof that the Mozilla code is badly written, but rather that the Mozilla Foundation is aware that secure code is hard to write, and that a good review process is critical to reach this goal.
And thats why Open Source is better! find it one day patch it the next.
Nimbda and Code Red both came out after patches had been available for months. I don't see this as positive or negative for Open Source.
At the end of the day--regardless of platform, it comes down to someone actually installing the patch!
OSS permits investigation and transparency
Without design specifications and a complete, well written documentation, the only way people could check a program is by reading the whole code and understanding the whole thing. Do you know a lot of people who would waste hundred of hours to look for bugs (apart from the ones who are developing the program) ?
OSS permits investigation, but no one is doing it because most OSS project have very little documentation. The result is most OSS project are extremely buggy.
And even worst, since most people who "work" on OSS project do it as a hobby, they prefer to add new shiny things rather than fixing bugs. Take the address book in mozilla/Thunderbird for example. I regularly lose contacts. Also, I once deleted a contact, and it gave the address of the deleted contact to the preceding contact - which means I was sending mail TO THE WRONG PERSON. Last week I tried to copy 34 address from one address book to another, it said 34 address copied, but then there was only 33 address. Found the missing address, tried to copy it (drag and drop), but no, I had to enter it manually. It's a real joke but no one is fixing it.
So who's shouting "Shut up! I can't hear you! la la la la" ?
HAHAHAHAHAHAHAHAHA!!!
Somebody mod that guy up as Funny!!!
Or, if you're not trying to be funny, you've clearly never worked in QA, or... maybe you've just explained that there are few GOOD pieces of commercial software...
Anyway, let me assure you that I worked a lot of QA gigs, and in every single one of them, the QA team was dwarfed by the dev team, rarely had good specs to plan from, and found their test time was viewed the most expendable part of the product cycle ( it's the first one to shrink in case of a slip elsewhere ). And those automated tests? Those paths you automate aren't likely to have *glaring* problems- at lest not ones the automated tools can catch - it's just the cases QA didn't have time to code up that'll fail... and of course, you can't automate something until the program is available, can you ? In practice, automated tools are only *really* useful for regression testing.
The most important thing I learned working QA is that the best QA in the world won't save you from a poorly planned or managed project, poor design, coders who don't unit test, or marketing guys who promise the sky and give a fixed do-or-die ship date to go with that sky. Code review is usually better than QA at finding non-design-related bugs. If the coders are good, QA ends up finding usability issues, rather than functionality issues, which is your best-case scenario, even though it means your prototyping and design phase was lacking.