Slashdot Mirror
Google Solves Sharing Bug In Google Docs
RichardDeVries writes "Three weeks ago, I contacted Google about a bug in Google Docs that shared documents without permission. The issue has been resolved and affected documents have had their collaborators removed. The documents' owners have been notified: 'To help remedy this issue, we have used an automated process to remove collaborators and viewers from the documents that we identified as being affected. Since the impacted documents are now accessible only to you, you will need to re-share the documents manually.' See my journal entry for details on my contact with Google. Although I think Google handled the issue admirably, this raises questions (again) about cloud computing, as well as Google's eternal beta-status for a lot of their services."
Prime reason to avoid online office suites and the like. Another good reason is that even these days Internet access is not a given 24x7 every place you want to be.
These posts express my own personal views, not those of my employer
It seems Google treats their beta products better than most treat their production stuff. Fitting, considering Google has more users of their beta stuff than other companies have paying users.
It raises more immediate questions about SAAS, which Google docs is, not cloud computing. (Google docs is software as a service, not a cloud computing service like Amazon ec2.) Someone else's custom app can have a bug, and leak your data.
So can your provider's closed-source proprietary cloud computing applications, user provisioning, storage, etc.
If, however, the provider uses an open-source hypervisor (like KVM), and open-source provisioning, management tools, and scripts (so the wrong user isn't given access to your storage), cloud computing should be much more secure than a SAAS platform like Google docs.
But yes, it does raise some question about services like ec2, because they're fairly opaque and using proprietary software, how can you possibly prove that their provisioning system is secure (in that YOUR elastic block store can't accidentally be provisioned onto someone else's ec instance)?
One possibility is to use full-drive encryption on all your volumes, and require interaction with custom software on your side to boot your instances.
Although I think Google handled the issue admirably, this raises questions (again) about cloud computing, as well as Google's eternal beta-status for a lot of their services.
Really? I don't use Google Apps but I don't think the act of fixing a bug in any way raises questions about the overall concept any more than Microsoft fixing a bug in Sharepoint would raise questions about closed source Windows services, or fixing a bug in KnowledgeTree would raise questions about similar open source services.
Software application has bug; bug gets fixed. Jesus people, why is this different from any other similar bug being fixed? Oh, it's Google, better get blogging.. Gotta get those ad impressions up.
When there's a bug in my internal doc collab and versioning service, it isn't exposed to the entire world.
I think that's the question raised.
My god, every time someone comes up with a solution in IT, we have this built in expectation that everyone should fall on board. Cloud computing is just the latest. Are we to now upgrade every system to use the "cloud". Are we to do web applications for everything? This isn't an engineering profession, its a fashion one. We're not like Mr. Spock from Star Trek. We're like the guy on America's Next Fashion Designer.
There are a bunch of good reasons web applications have become popular.
First, they're easy to deploy. Put up a web page, point the users to that web page, and you're done. No need for an installer. No need for an updater. No need to convince users to download and run an executable (which is a scary and complex undertaking for many of them).
Second, they're relatively safe for the user. Which puts the user at less risk, navigating to a web site, or downloading and running an executable which may or may not contain malware?
Third, they're cross platform. With a little effort, your web application will run on Windows, OSX, and Linux. This should make Linux users very happy, since it helps even the playing field between Windows and Linux!
Fourth, in many cases, web application providers can offer superior document management. For example, regular users aren't good at keeping backups, and in the old days, just plain said goodbye to their archived email if their hard drive crashed. Or, if they upgraded from a Pentium 3 computer to a Pentium 4 computer, they spent hours trying to configure their new email program, and then more hours trying to move their archived email from their old computer to their new computer. Compare that to web email, which Just Works.
Do web applications involve risks and tradeoffs? Yes, this article demonstrates that. But it's up to individuals to decide what risks and tradeoffs are worthwhile, and many, many people choose web applications because the advantages are worth it to them.
Claiming that web applications are successful only because they're fashionable to developers these days is, well, just plain stupid. The fact is, web applications are the best choice among the alternatives for many users, and plenty of developers recognize that fact and leverage it by building web applications instead of thick client applications.