Google Fail

So I live in College Park, Maryland; which is directly on the North Western side of the infamous Beltway…and let me tell you: traffic is a bitch right outside of DC.

For those of you unfamiliar with the DC area, there is a single, unifying highway that loops completely around the metropolitan area of Washington and it is absolutely plagued (daily) with traffic.  It is, with out a doubt, one of the most traffic-prone highways in the country, and it can take up to an hour and a half just to travel 15 miles to my hometown of Rockville, Maryland.

Anyways, if duty, family or friends call and I need to head back west towards Rockville, I have taken it upon myself to check Google Maps for the traffic.  Google recently started collecting location data from people who use smartphones and have Google Maps installed.  When this geo-location data is cross referenced with location change over time gives Google the unique opportunity to display live traffic on their maps…or so they say.

Whats that you say? That sounds awesome! Well it would be, if the damn data was correct.

Tonight I took a short trip back to the Rockville area for a meeting that was scheduled to begin at around 7:30pm.  I figured, given no traffic, I would be able to venture home in about 30min MAX.  So I open up my browser and head over to maps.google.com, only to discover at about 7:00pm (optimal time for departure from College Park under ideal circumstances) that the entire route back is shaded in red.  Red = slow (or sometimes danger or stop…but this is irrelevant to this discussion).

Since according to Google Maps the traffic was bad and I had no time to waste, I decided to head out west anyway. And guess what? NO TRAFFIC.

I was in a green situation, not a red catastrophe like I was expecting.  After about 5 minutes on the highway, traversing many more miles than Google Maps predicted, I checked Google Maps on my Palm Pre.  Still red! Even in the area I was currently in.

So no, Google Maps can’t possibly have “live” traffic reporting.  I even restarted my phone just to see if it was a caching problem…but no, Google was just flat out wrong.

Google, take off the “live traffic” label on your maps.  I really rely on that to planning my schedule in a hectic traffic are of the country. Could you (Google) or anyone else provide a legitimate excuse for a misrepresentation like this? I can’t see one…and I am part of the Google Generation.  My contemporaries, at the very least, need to feel like they can rely on Google…and usually they deliver.  But no on this one.

I’d really like more in depth idea of how Google’s Live Traffic feature truly works, or the “Live” aspect of the system needs to be revised or removed. Gracias.

</end rant>

So I believe I have found a security hole in Google’s login systems, and why they need to unlink usernames with YouTube accounts, or they need to at least address this issue.  I’ll be brief.

Say you have a Google Account.  Say your other friend has a Google Account.  These two accounts are independent of each other, as in they have different User Names and Passwords.  Now, say that one of you, lets say here that its you, who creates a YouTube account to share with me.

This is a very common circumstance.  I have about 3 different accounts in YouTube that I share with other people, it makes it convenient to maintain the account, especially if its a busy one.

So here is the security hole: if I log into this ‘Shared” YouTube account, all I need to do is head on over to Gmail and waalaa! I’m now in your Gmail.  I have full access, and I can poke around all you want, without you knowing except for that little IP log at the bottom of the Gmail window.  I could do some real damage  and snooping.

This is a serious issue, please unlink the accounts.  Email is rarely shared, YouTube accounts are.

UPDATE: This is the response from Google about this issue.  Apparently these accounts were linked, and this is just the nature of the Google Accounts system.  I agree with the security team, there is no large risk, but this is still a problem in my opinion.

From Google:

Thank you for the clarification, and for helping me figure out the
nature of the problem.

As you noted, the user in fact disclosed his Google account password
to you, along with an “alias” on YouTube. This alias serves simply as
a nickname for his canonical account with Google, and the password
could be readily used to access services such as Google Mail or Google
Docs without the need to rely on YouTube at all.

Although the fact you gained access to all Google services by logging
in via YouTube with this alias may sound somewhat counterituitive to
people less accustomed to a variety of Google services, I believe
there is no security risk. As noted, you could have used the same
password, and his canonical account name, to simply log in at:

https://www.google.com/accounts/Login

…to gain access to the same services. Canonical account names are
not a secret, and could be easily discovered, e.g. through the YouTube
UI itself.

Now, it goes without saying that sharing your password with other
parties is usually not a good idea, for a number of reasons; if this
can’t be avoided, we would recommend creating a separate Google
account for this purpose.

My Short Rebuttal:

Absolutely Sir, I will add to my blog post now.  I have to say though, that it would not be a bad idea to give Google Account’s access permissions.  This would also prevent something confusing like this from happening.  The basic fact is, while your correct there is no large security risk, this user (my friend) had no intention, nor any indication (according to him) that he was allowing me to access his Gmail without his explicit consent.  Just something to consider…

I would like to praise Google right now, for their quick response time, and detail centered approach.  This encounter with them was exciting in its short life span.  Onward…

I just want to do a really quick post about the recent Twitter changes with regard to how @reply’s are shown in your timeline.

If you don’t know already, Twitter used to have an option in the settings area that allowed you to choose whether or not you could see reply’s of a particular user you were following that were replying to another user. In short, if you don’t know the person in an @reply from a user, their tweet with that reply would not show up in your timeline feed. You could, before the controversial change, opt to see them or not.

Now you cannot see them, and that is the default setting that cannot be changed by the user anymore. The reasons for this change, again in short, is to limit the API calls (relieve stress on Twitter’s servers) and because only 3% of users enabled the option to see @reply’s to people they did not know.

The problem is that now I cannot start ANY tweet with a user name, even if its not a reply, I am just starting my tweet with their user name. For instance:

Laura sucks at running.

I would write this on twitter as:

@limerasp sucks at running.

This is not a reply to her, rather a statement that begins with her Twitter user name. With the recent change to how Twitter handles @reply’s, any user who does not follow @limerasp, will not see my publicly intended tweet.

This has happened to me about 4 to 5 times recently, and I think that its something that needs to be changed. I don’t want to add ‘Well, ” before a publicly intended tweet that starts with a user name; but thats exactly what I’m being forced to do now.

I didn’t really mind the whole @reply change at first, but I’m starting to see that it puts Twitter at both a semantically and socially disadvantaged position. I can’t say exactly what I want to say, or at least I have to add words to my tweets that I otherwise would not include, and honestly, everyone should be able to see an @reply, even if the user isn’t following that particular user. The reason is it promotes the conversation. When I follow someone, I follow them not just because I like what they are saying, but because I like THEM as a person (in-so-far as I know them). I do want to see who they are talking to, because of what I will call the Transitive Follower theory.

It’s this: It is of high likelihood that a frequent conversation between an unknown user and a user who you follow that the unknown user is a worthy follow.

Simple. Kind ofa obvious, but needs to be said.

If you don’t want your stuff public, send a DM. That’s what they are there for.

According to David Martin, Vice President of Primary Research at Nielsen Online, Twitter’s current visitor retention rate, that is the number of visitors that return to Twitter after signing up within one month is about 40%.  Now for some strange reason, Nielsen isn’t reporting how exactly they determined that number.  I’m going to go ahead and guess here, but I bet they calculated that based on visitors returning to the Twitter.com domain, which won’t give them correct results.

Now whats the problem with this?  Well, everyone, and I do mean pretty much everyone, wants to get off of the Twitter website ASAP and use a client that sits on their desktop.  If not a desktop client, they are using the Twitter API in a number of other ways on another site, normally another social network mash-up site. This is because while Twitter is a great service, their UI isn’t as seamless as Tweetie for Mac or Twhirl or as popular as TweetDeck.  It’s not that Twitter’s UI is bad, its just why would anyone rather be reloading the pages on a website to see their replies rather than just switch to an already loaded tab sitting on your desktop. So really, the Twitter.com domain isn’t being hit by folks like Facebook or Myspace (as they have graphed out), rather they are relying on Twitter’s API to deliver the service. The incredible amount of API call statistics, I would assume, is not accessible by Nielsen since Twitter is a centralized service running exclusively through themselves.

I call for a new study to be published with these variables addressed, or reveal that they accurately analyzed Twitter’s retention rate with respect to Twitter’s API.

After being on twitter for over a year now, I have finally started to follow mass amounts of people in search for some good inspiration, friendship, and conversation.  However, with all these new friends, I have to weed out the one’s who aren’t really contributing to the conversation-at-large.  The ‘tweeps’ I usually weed out are users that….

1. Only tweet links.
2. Talk about how their inbox has 1,000′s of unread emails that must be answered.
3. Constant @’s to my tweets that have nothing to do with what was said.
4. Only retweet other people’s tweets.  What, have you no thoughts of your own?
5. Celebrities or other famous-types who don’t interact with fans.
6. Celebrities or other famous-types who have someone else tweet for them.
7. Constant self promotion.
8. Ridiculous use of Adjix to gain money from links (it’s even worse if an ad is embedded into the tweet).
9. Tweet once in a blue moon. I will never see your tweets, so whats the point of following you?
10. Tweeting about completely useless information, like “I’m hungry”.  At least tell me what you are eating or going to eat; or even you inexplicable love for a particular cuisine.
11. Anyone who is rude in 140-characters or less.

What kinds of tweeple do you drop when weeding needs to be done?