Google's pissing me off again



  • Look at this shit:

    Just look at it.

    What is this shit? Leaving aside for a moment the gigantic "WTF" involved in Google even KNOWING who I'm following on Twitter (or even knowing that I'm using Twitter, for that matter), why did it just decide to COMPLETELY HIJACK MY SEARCH to ask me this/these question(s)? Hell, what question is it even asking? Is it asking "want to share your tweets with your friends on Google?" or is it asking "add blakeyrat to your public Google profile?" Because those are two different questions!! (The answer to both of which is "no". Or more accurately, "go away and never show me this again, fuck you".)

    Ah, but even better. I was looking for the source code of that amazing C# one-liner Alex posted a week or so ago, where you can redefine the "null" field of the string object template to be a space instead of null. That's the correct search result (I think) but the link is broken because Community Server of course it is. Since all I need from the search is a single line of text, once that comes directly after the blurb, it sure would be spiffy-nice if I could view Google's cached version of the page! Oh but where's the cache link? It's not there, I get this retarded Twitter shit instead!

    Google, seriously. Bing is kicking your ass in everything but customer numbers at this point. Your product has turned to crap, and every attempt you make to RIP OFF Bing's features make it worse. Figure. It. Fucking. Out. Right now the only thing you have going for you is Bing can't find that page, either.

    (Or just create a brand-new format we don't need, maybe that'll help. How about a video for--no wait you did that one. How about an image forma-- oh nope, you did that one too. Maybe you could replace HTTP with-- oh wait...)

    Edit: fixed image link, sorry email users



  • Are you currently logged into twitter because it may be guessing that you'd be interested based off the fact you have a twitter auth cookie... 

    What makes you think it knows who you're following?  It just asks you if you want to share your tweets, not "do you want to share your tweets with [list of your contacts]".

    Besides, Google === Borg.  Resistance is futile, etc.  This was going to happen eventually, might as well get used to our computer-based overlords.



  •  @blakeyrat said:

    I was looking for the source code of that amazing C# one-liner Alex posted a week or so ago, where you can redefine the "null" field of the string object template to be a space instead of null.

    typeof(string).GetField("Empty").SetValue(null, " ");

    Edit: I think this is what you meant though it changes string.Empty to a space, not some null field.



  •  http://thedailywtf.com/Comments/The-Disgruntled-Bomb.aspx?pg=2#340740

    As noted it sets String.Empty to a space instead of null.

    But still a phenomenal trick to play.



  • @locallunatic said:

    typeof(string).GetField("Empty").SetValue(null, " ");

    THANK YOU!!!

    I think a lot of my frustration stems from:
    1) This being the first in a long time I've failed to find something I've been searching for
    2) Google getting me SOO DAMN CLOSE to the answer, then failing utterly on the last inch

    @locallunatic said:

    Edit: I think this is what you meant though it changes string.Empty to a space, not some null field.

    It is. If my memory were good enough to remember that detail, I probably wouldn't have had to search for it in the first place.



  • @blakeyrat said:

    I think a lot of my frustration stems from:
    1) This being the first in a long time I've failed to find something I've been searching for
    2) Google getting me SOO DAMN CLOSE to the answer, then failing utterly on the last inch
     

    That's where my frustrations usually come from with them.  If you want to avoid some of the newer features (assuming the twitter junk is part of that) then you may want to try their SSL site, it's generally a few steps behind on new things getting added (but stuff that's used bunches comes through).



  • @C-Octothorpe said:

    Are you currently logged into twitter because it may be guessing that you'd be interested based off the fact you have a twitter auth cookie...

    Yeah, I am, but my Twitter auth cookie is on a completely different domain that Google shouldn't be touching to deliver me search results. That also doesn't excuse the awful copywriting on that question... I've read it a dozen times, and I still have no clue what would exactly happen when I hit "yes." (I figured it out by going to the More Info link, but... seriously, that's terrible copywriting.)

    The funny thing is, I actually wouldn't mind that feature in, say, Google Reader (which you see I already have open)... if Google would suggest new feeds based on my Twitter friends, I'd actually probably turn that on. But SEARCH RESULTS? Redonkulous.

    @C-Octothorpe said:

    What makes you think it knows who you're following?  It just asks you if you want to share your tweets, not "do you want to share your tweets with [list of your contacts]".

    At the very least it knows I'm following Alex. I didn't screenshot it, but it explains that in the "more info" link.



  • @C-Octothorpe said:

    it may be guessing that you'd be interested based off the fact you have a twitter auth cookie... 
     

    What?



  • I've tried my best to replicate this issue, but I can't. The only way that I know of, for google to know about a twitter account is to have the user connect the twitter account with the google account.



  • @delta534 said:

    I've tried my best to replicate this issue, but I can't. The only way that I know of, for google to know about a twitter account is to have the user connect the twitter account with the google account.

    Some variables to consider:

    1) it could be an A/B test, and I'm part of the test group (for example, 1% of users see this new feature)
    2) it could have something to do with me signing up for Twitter using a Gmail address
    3) it could involve my Gmail address and Twitter alias having the same name

    As far as I'm aware, I've never plugged my Twitter login into any Google sites.



  •  @sabbott64 said:

     http://thedailywtf.com/Comments/The-Disgruntled-Bomb.aspx?pg=2#340740

    As noted it sets String.Empty to a space instead of null.

    But still a phenomenal trick to play.

    Personally, I'd rather set String.Empty to "Empty".  Then when one of my coworkers figures out that his code isn't working because String.Empty is "Empty", I can say, "Well, what did you expect?"



  • Google always pisses me off.


  • Trolleybus Mechanic

    @zzo38 said:

    Google always pisses me off.
     

    Everything always pisses everyone off. So there.

     I've hated that fucking twitter hijack ever since it was put in. I haven't seen it since, thanks to a greasemonkey script called [url=http://userscripts.org/scripts/show/65892]Get twitter out of my google![/url]

     

    mod: fixed link. URL values don't need quotes. -dh



  • @ShatteredArm said:

     @sabbott64 said:

     http://thedailywtf.com/Comments/The-Disgruntled-Bomb.aspx?pg=2#340740

    As noted it sets String.Empty to a space instead of null.

    But still a phenomenal trick to play.

    Personally, I'd rather set String.Empty to "Empty".  Then when one of my coworkers figures out that his code isn't working because String.Empty is "Empty", I can say, "Well, what did you expect?"

     

     

    Ah, the Classic Noman technique! 



  • @blakeyrat said:

    @C-Octothorpe said:
    Are you currently logged into twitter because it may be guessing that you'd be interested based off the fact you have a twitter auth cookie...

    Yeah, I am, but my Twitter auth cookie is on a completely different domain that Google shouldn't be touching to deliver me search results. That also doesn't excuse the awful copywriting on that question... I've read it a dozen times, and I still have no clue what would exactly happen when I hit "yes." (I figured it out by going to the More Info link, but... seriously, that's terrible copywriting.)

    It's basically asking "Do you want to add the twitter account called Blakeyrat to your Public Google Profile[1] so it will share your tweets with your friends on Google"

    As for how it knows your twitter account: do you per chance have a Twitter client plugin/extension installed? That would be the only sane and non-violating answer if it did that to me.

    [1] I find it ironic how it abbreviates to PGP.



  • @Shondoit said:

    It's basically asking "Do you want to add the twitter account called Blakeyrat to your Public Google Profile[1] so it will share your tweets with your friends on Google"

    Well I know that now. I'm complaining about the crappiness of the copy, not asking, "HEY GUIZ COUD SOMEONE ESPLAIN THIS TO MEEEEE PLZZ?" Do you guys do this in real-life conversations? How do you avoid punches to the nose?

    "Hey Bob, don't you think the copy on this site is pretty confusing?"
    "No, see, what it's saying is that if you hit Yes it will add your Twitter account to your Google profile."
    "Well, yeah, but... the copy is really vague, I don't think someone who doesn't work here will be able to figure out what--"
    "You hit Yes, and it adds the Twitter account to your Google profile. Then you can share tweets on your Google profile."
    "Bob, I'm not asking how it works, I'm telling you the copy needs work, it's not very go--"
    "The Yes button will add the Twitter account to your Goo--"
    [PUNCH IN FACE]

    @Shondoit said:

    As for how it knows your twitter account: do you per chance have a Twitter client plugin/extension installed? That would be the only sane and non-violating answer if it did that to me.

    No, but now that you mention it, I do use the Twitter app on my Android phone that also has my Google account on it.



  • Sorry, I reread what I wrote and wasn't clear at all...  I feel wierd because you already know this but what I was saying that when blakey authed with twitter (wasn't sure if he did), the auth token is stored in a cookie, and that Google sniffed the cookie pasted back in the search request to see "hey, he's got a twitter account. Let's show him some shit he may be intereseted in". Doesn't make much sense though because it's from a different domain (as pointed out by blakey) and Google wouldn't get the cookie from twitter.



  • @C-Octothorpe said:

    Doesn't make much sense though because it's from a different domain (as pointed out by blakey) and Google wouldn't get the cookie from twitter.

    Well, Google could ping Twitter and ask if I'm logged-in-- Twitter uses OAuth so that's actually possible (and one of my many, many complaints about shitty-ass OAuth). The point isn't that they technically can't do it, the point is that they shouldn't be doing it. Especially on Search. (I could see them doing that in Gmail or Google Reader-- it would still annoy me but not nearly as much as interrupting my search and erasing the Cache link to give me that crap confusing copy.)

    What's more likely is that Google pinged Twitter and asked if anybody with my email alias existed on Twitter, and based it on that. Twitter wouldn't give away my email address without a proper login, so it couldn't be that... hm.

    You know how when Amazon makes a recommendation they have a link that says "why we recommend this" and it returns something like, "we think you will like The Fifth Element because you bought other movies starring Bruce Willis: Die Hard and Pulp Fiction". Google badly needs that.



  • @ShatteredArm said:

     

    Personally, I'd rather set String.Empty to "Empty".  Then when one of my coworkers figures out that his code isn't working because String.Empty is "Empty", I can say, "Well, what did you expect?"

    I'd rather set it to the end of file character.


  • @blakeyrat said:

    @C-Octothorpe said:
    Doesn't make much sense though because it's from a different domain (as pointed out by blakey) and Google wouldn't get the cookie from twitter.

    Well, Google could ping Twitter and ask if I'm logged-in-- Twitter uses OAuth so that's actually possible (and one of my many, many complaints about shitty-ass OAuth)

    That doesn't seem to make much sense either - isn't OAuth supposed to only pass on information about sessions if the user actively connected with your application?



    Then again, google being the evil overlord company they are, maybe they're using chrome to actively sniff for twitter cookes. I wouldn't put it past them.



    Or, the boring variant, they just made some kind a deal with twitter to get the sessions.



  • Forget the copy, this is potentially very sinister. I don't recognize the browser, as I only use good ones, but my suspicion is that it's Chrome and that the whole thing could be inserted client side. Which would be a whole new can of WTF. Chrome could essentially be an elaborate ploy to dupe the masses into sharing non-Google-domain cookie data with Google.

    But really, if "Bing is kicking [Google's] ass in everything but customer numbers at this point. [Google's] product has turned to crap", why are you still using Google rather than Bing to search today?

    Vote with your feet, or quit complaining.



  • Holy speculation, you guys!

    Yet none of you are seeing what's really going on! Through backroom deals with monitor suppliers, Google is inserting a special driver hook that activates hypnosis waves whenever Google-brand products are used. These are injecting advertisements right into your mind! Now they're doing the same with third-party products! WAKE UP, SHEEPLE!



  • @PSWorx said:

    That doesn't seem to make much sense either - isn't OAuth supposed to only pass on information about sessions if the user actively connected with your application?

    It's probably "supposed" to, yes.

    But consider the following:
    1) I have Twitter open and logged-in in another browser tab/window
    2) I visit a blog that optionally lets me authenticate using my Twitter login, via OAuth
    3) To authenticate, the blog provides me with an iframe lightbox containing a link to Twitter domain page
    4) Twitter then immediately redirects back to the blog "oh yah, I know this guy, he's got a cookie on my domain, he's cool"
    5) I'm now logged into the blog using my Twitter credentials

    So far so good, right?

    Now, imagine the iframe in step 3 was sized 1x1 pixel. Further, imagine that the URL of the iframe was set directly to the Twitter OAuth page. Twitter would do the auth with absolutely no interaction on my part. As long as the blog and Twitter are open at the same time in the same browser, BAM, the blog knows I'm logged-in to Twitter, and it knows my Twitter account name.

    (To be fair to Twitter, it does ask my permission if the blog wants to actually use my account for something other than authentication-- like sending tweets or adding followers-- but for basic authentication, Twitter doesn't ask squat, it just redirects back immediately. Also to be fair, this process probably wouldn't work in Safari, due to its weird cookie handling. Although I haven't tested that.)

    @bertram said:

    But really, if "Bing is kicking [Google's] ass in everything but customer numbers at this point. [Google's] product has turned to crap", why are you still using Google rather than Bing to search today? Vote with your feet, or quit complaining.

    Fair criticism, and I am going to go back to Bing. (I used it about 2 months or so previously; but eventually laziness got the better of me.)



  • Blakey's browser does indeed look like Chrome, but I hope for his sake that it's actually Iron. At least with Iron, you could be sure that the 'intrusion' is NOT coming from the client-side browser code.



  • @Cad Delworth said:

    Blakey's browser does indeed look like Chrome, but I hope for his sake that it's actually Iron. At least with Iron, you could be sure that the 'intrusion' is NOT coming from the client-side browser code.

    oooOOOooOOOooOOOoooOOOOoooOOO the calls are coming from inside the house!!

    Oh wait, I'm not a paranoid kook. Of course it's fucking Chrome. Christ.



  • @bertram said:

    Forget the copy, this is potentially very sinister. I don't recognize the browser, as I only use good ones, but my suspicion is that it's Chrome and that the whole thing could be inserted client side. Which would be a whole new can of WTF. Chrome could essentially be an elaborate ploy to dupe the masses into sharing non-Google-domain cookie data with Google.

    But really, if "Bing is kicking [Google's] ass in everything but customer numbers at this point. [Google's] product has turned to crap", why are you still using Google rather than Bing to search today?

    Vote with your feet, or quit complaining.

    Are you serious? Like, really? It is very easy, bordering on trivial to prove that Chrome is not sharing cookies with Google that it shouldn't be. See: [url=http://www.fiddler2.com/fiddler2/]Fiddler[/url].

    Never mind the legal liability they would be setting themselves up for if they did so. It would be like the Microsoft scandals, except a googol times worse.



  • Well, obviously Chrome does a browsing history analysis first and only starts sending cookies if the user has spent sufficiently long time on Craigslist and is thus evidently too stupid to operate Fiddler.



    Also, thinking about it, this wouldn't even work from a conspiracy theorist point of view. With all the google ads, analytics, youtube embeds and google-hosted JS libraries in use, Google has probably direct access to the JS/DOM of 95% of relevant sites on the web anyway. Hell, if they wanted, they could just include a keylogger script into every 10.000th adwords/analytics/jQuery call and collect the passwords directly. (Or, you know, just ask the user. It worked for Facebook.) No need to take the detour via Chrome.



    So, no, I don't like Chrome's "Silent Updating" either but I agree, aside from moral standpoints, there is probably not much to fear yet, since, should Google really try to make use of all the theoretical things they can do, it would probably be the most stupid decision they could ever do.

    @blakeyrat said:


    4) Twitter then immediately redirects back to the blog "oh yah, I know this guy, he's got a cookie on my domain, he's cool"

    5) I'm now logged into the blog using my Twitter credentials



    That would seem more like a short-lived bug in Twitter's OAuth implementation. Their own API doc says:

    1) If the user is logged into twitter.com and has already approved the calling application, the user will be immediately authenticated and returned to the callback URL.

    3) If the user is logged into twitter.com and has not already approved the calling application, the OAuth authorization prompt will be presented. Authorizing users will then be redirected to the callback URL.

    So you need to ask the user at least once before you get the auto-redirect feature. When did you grant access to google? Oh wait, you didn't.


  •  I really doubt this feature is to blame for the missing Cache link. I often find that certain results on Google will be missing the Cache link, even when there's nothing else different about it.



  • @Cad Delworth said:

    Blakey's browser does indeed look like Chrome, but I hope for his sake that it's actually Iron. At least with Iron, you could be sure that the 'intrusion' is NOT coming from the client-side browser code.
    Right.   Because if you don't trust Google you can surely trust some random guy who claims to have created a privacy-enhanced version, even though a diff of the Chrome and Iron source code shows only trivial changes that don't do anything that you can't do yourself with user-configurable settings.

     



  • @PSWorx said:

    That would seem more like a short-lived bug in Twitter's OAuth implementation. Their own API doc says:

    1) If the user is logged into twitter.com and has already approved the calling application, the user will be immediately authenticated and returned to the callback URL.

    3) If the user is logged into twitter.com and has not already approved the calling application, the OAuth authorization prompt will be presented. Authorizing users will then be redirected to the callback URL.

    Hm, you might be right. The blog I did that exact process with is ComicsAlliance, which if you read the fine-print is actually owned by AOL. And I notice AOL is in my list of Twitter-approved applications. Which is even more misleading, since whatever AOL site made it appear on that list was not ComicsAlliance... (Google is not on the list, BTW. Twitter For Android is.)

    So it looks like what happened is that some time in the nebulous past, I signed on to some AOL site using my Twitter handle, that login was stored under AOL's name and not the site's, then later when I sign on with another AOL-owned site, I don't see the login dialog.

    OAuth is still crap though.



  • Not having a twitter account keeps getting more and more advantages.



  • Sometimes I wish software would only do what it's supposed to do, rather than all these other little side effects and "features."  I'm trying to decide if the occasional "oh that's a cool feature I wouldn't have thought of!" is worth the pain of all this other stuff.

    Of course, the thing that first (and still) prevents me from installing any Google application is the unsolicited Google Update startup item.  I actually tend to eschew any program which installs such a thing without even telling me, let alone without my consent.  (I mean, really, is it that hard to have a dialog that says "the following things will be installed"? Then I can only blame myself for not reading the list.)



  •  And to think that I was pleased that Google no longer treats the hash in C# as meaningless filler.


Log in to reply