Update: Mt. Gox was compromised, the database of users was released. I believe the thread here was removed, but many people will probably be able to verify it.

Change your passwords now.

^{I told you so}

There is only one solution that will work in any situation to solve the problem of infiltrants/provocateurs. It will work, regardless of the hierarchy of a forum, even regardless of the used communication medium. It will work regardless of whatever factor may influence the environment you are using to communicate. The best thing is it won't cause any technical issues or restrictions, and keeps everything accessible for everyone.

It's called individual common sense.

Javascript is a legitimate technology that is pretty much a basic cornerstone of the web as it is now. You can't just take that away. A way better option would probably be if browsers by default protect against CSRF attacks, like they do with XSS now.

Voorzover ik kan zien is het een combinatie van dingen. Nadat de psychologische grens van $30 was bereikt, verkocht iemand een berg bitcoins en daarna werd het een soort van sneeuwbal-effect. Combineer dat met het nieuws over de US overheid vs. Bitcoin, de diefstal van wallets, de inbraken op Mt. Gox, de DDoS-aanvallen op een hoop bitcoin-gerelateerde sites, etc.... en dan is het niet zo vreemd dat de boel fluctueert
Dat komt wel weer goed

You could easily bring a backup of the wallet.dat on your USB thumbdrive. That being said, I think there should be an intuitive interface to manage things like USB thumbdrive backups etc.

There is always the risk of a wallet stealer running on an internet cafe computer, but then again - there's also a risk of a software or even hardware keylogger being present!

Seeing as a considerable amount of Bitcoin-related sites was being attacked (I believe Mt. Gox had to turn to Prolexic?), it appears to be a person or group that doesn't want Bitcoin to exist.
Probably not Anonymous, probably not Lulzsec either, maybe a bank/government agency, but for all you know it may be one single person who owns a large botnet and just happens to dislike Bitcoin. There's not really a way to tell.

Do some research before you claim things. That was an attempt to frame Lulzsec that has been widely publicized...

Not entirely true. Linux is absolutely more secure by design, and even *if* more than half of the world was using Linux for their desktop machines, it would be considerably harder to write successful hardware for Linux systems than it would be for Windows systems.
I do agree however that a wallet stealer would be just as successful on Linux, seeing as your wallet is stored in your /home directory, and is thus accessible freely by anything you run. A "wallet stealer" really isn't anything more than something that emails/uploads a file in your user directory.

If anything, the wallet needs to be encrypted by default (through the client, and not by third-party software, so that you never need to have an unencrypted copy accessible as is the case with Truecrypt etc).

Then I suppose you have the IP address of the person who logged in to the accounts?

Also, I have not yet received a response to my tickets #957/#1797, nor to the PM I sent you on this forum.

By refusing to accept tainted coins you are indirectly forcing others to refuse them as well, because there is a risk they can't spend them on what they intended to spend them on. I am not even talking a centralized blacklist here, just the fact that there is a risk in that.

Not to mention that it's entirely impossible to implement this because there are plenty points (merchant wallets, exchanges) where coins get mixed up. You would be refusing coins for the single fact that they have at one point been used by someone with bad intentions. How does that make any logical sense?

In comparison, right now you can just steal the wallet file from a mounted Truecrypt partition in 2 seconds and be done with it. Comparing that to needing a keylogger for a prolonged time to be able to decrypt the wallet in the first place... an always-encrypted wallet would be a VERY good idea.

^{There's always the input-through-mouse PIN system.}

Without a massive-scale timing attack making use of so many nodes that you are connected to every single legitimate node on the network, there is, as far as I am aware, no way to determine how many bitcoins a person with a specific IP address holds.

There will always be a psychological barrier in systems that are so wildly different from the "conventional" systems. I believe it's better to educate people about Bitcoin and how it works, and help them get over that psychological barrier, rather than taking the "easy way out" and changing the system to meet what people are used to.

On the other hand, if you use a web wallet you have to trust that they will adequately protect your funds whereas with a wallet on your own pc you can make it as secure as you want.

There are already several webwallet providers that provide exactly this service. I believe they are also mentioned in various places in the wiki. I don't think that, for the sake of decentralization, it would be a good idea for bitcoin.org to run a web wallet service.

I know there's a psychological barrier involved in downloading the client, but from an ease of use perspective Bitcoin is not any more difficult to get started with than any other software package. It's literally install and go.

The issue is not with Bitcoin. It is perfectly possible for someone to set up a Bitcoin bank, that has insurance against theft etc, just like "conventional" banks. The issue here lies with Mt. Gox, which is only a single independent exchange. Bitcoin itself (as an idea and protocol) is technically sound. The only thing I am missing is wallet encryption in the client by default, but that can be overcome for now by storing a wallet on a machine that is not connected to the internet, using third-party encryption software.

Bitcoin is much like digital cash, with the difference that you can encrypt a Bitcoin wallet, while you can't encrypt an IRL wallet.

Not just that. There has been a lot of media attention for Bitcoin, not just regarding Lulzsec etc, but most importantly things like Silk Road.
Lulzsec never claimed to have been responsible for that transaction. They simply tweeted the link.
Bitcoin is definitely not just fueled by "hacktivists" (also, Lulzsec are not hacktivists, they are quite literally doing it "for the lulz").
While I doubt Lulzsec, Anonymous, etc. would DDoS mining projects, exchanges, and other Bitcoin-related sites, the concern is reasonable. That is why it would be better if there were a lot more smaller pools and exchanges, so that one hit cannot have so much influence on the entire economy. If someone really DDoSed the major mining pools for weeks on end, and these pools would not be able to get back up, it could indeed be an issue.

EDIT: Whoops, almost forgot. The value of BTC is based on supply and demand - and indirectly on what people think the value is. Just like gold, and fiat currencies, if noone sees value in Bitcoins, they wouldn't have value. However, I believe that because Bitcoin is technically sound and has certain advantages, that alone is enough reason for there to be a value for at least a small group of people.

And you are getting this idea where exactly?

Again, two factor auth using email would be incredibly easy to implement, and a huge improvement in security - mostly because you can't get around that by having database access through a vulnerability.

Bitcoin has had a lot of attention lately. Of course there will be attacks from every side. People who just want to earn a buck from it in less elegant ways, and people who want to see Bitcoin vanish off the earth.

How exactly would an XSS work in this case? I have never followed any links to Mt. Gox from external sites, and my account was broken into at a point where I couldn't even access Mt. Gox (probably due to the DDoS attacks).