This topic is currently marked as "dormant"—the last message is more than 90 days old. You can revive it by posting a reply.
LibraryThing has replaced its Google reCaptcha with a LibraryThing-created "Are you a robot?" page. If you sign out and sign in again, you should get it.
The page asks you to identify cooking and other food books amongst other books.
If you dislike it, or find it too difficult, you can click "Traditional Captcha" to get the standard Google reCaptcha one. We figured clicking on cookbooks was more fun than clicking on cars and streetlights.
* We've tested the page on every browser and OS we have access to. If you have problems, we want to hear them.
* Because some members disliked Google being involved, the alternative, Google one does NOT load unless you click to see it.
* The new page will give you a number of tries. If you fail too many times, it will require you to answer the Google one.
* You should only have to deal with the new robot-check page once per month, if you use the same browser and account. If you switch browsers and users a lot, you may need to deal with it more often. Clarification: If you stay logged in, you won't get it at all.
* We developed the system with various "levels" of security. If we see an increase in attacks, for example, we will get harder on missed answers, obfuscate the data more, etc. We are watching.
I hope you find the new one fun. If you find it annoying, my apologies. But we NEED this to cut down on automated attacks. All-day, every-day LibraryThing is being attacked by bots, seeking to create new accounts, or compromise existing ones. (The attacks come from huge "botnets," and can't be stopped by IP restrictions.) We MUST do something about this, for your security and for the health of the site overall.
So you're going to start expiring login cookies once a month, then?
Added "Clarification: If you stay logged in, you won't get it at all."
We won't absolutely promise we won't expire cookies in the future; if we have to to fight bots, we will. But I don't foresee it.
Anyway, thanks for putting a stop to these spammers.
Only insofar as logging into the site requires it. The other Wiki captcha is separate, and part of the wiki software.
Tim was saying that you can sign out and in as many times as you want and you will only see the captcha once per month once you've captcha'd successfully.
Sorry. The Google one has an audio option, if you want to try it. We really need it—the bots just keep at it.
When I wrote #4, Tim had not yet edited #1 to make that obvious.
(As an aside, I really, really hate it when people do that, rather than clarifying in a response. It makes the responder who asked what was at the time an entirely reasonable question look like a total idiot, and then later responders treat them as such, like here.)
Googlejuice is another possibility. (LT has nofollow turned on, but the spammers don't necessarily know that.)
Lorax, I added the clarificaiton with the words "Clarification:" and then the clarification in bold. And then I posted a message with "Added "Clarification: If you stay logged in, you won't get it at all." What on earth could be more clear and explicit than that?!
Yipes. It's going to be really hard to support systems with as few users as Opera Mini for BlackBerry. Sending you a message by direct message.
It's mostly Googlejuice, I think.
The clarification is perfectly clear.
It was not clear prior to your edit, which was WHEN I POSTED #4.
Then conceptDawg came along in #9, AFTER YOUR EDIT, and responded to my #4 as though it had been written in the context of #1, in which case I would have been very stupid indeed to not have understood your very clear EDITED post.
I also really prefer the covers to the horrible horrible picture captcha, and I'm also happy not to be training google AI.
>7 bnielsen: HA-HA-HA.
>1 timspalding: Today when I logged in, the Google captcha was the default and it didn't work. Trying to remember all the way back to this morning, I think nothing actually came up for me to click on. I then clicked on the LT captcha, selected the cookbooks and it worked just fine. I'm on a Windows 10 PC with updated Chrome browser.
Some of the non-English LibraryThing's default to the Google Captcha. It has to do with how many local covers we have. Could that be what you're talking about?
re: Opera Mini
The problem with Opera Mini is that it's not really a browser in the real sense of the word. It adheres to very few web standards and nearly anything and everything breaks it. It's basically just taking screenshots of sites and serving those up to you. Yes, that's hyperbole, but it really is a tough browser to support effectively for any interactive content.
I came up with a solution for that user, CH. Talk to me about it if you don't know what I mean.
Having learned the lessons of PKD's parables, I have existential angst every time I check the re-Captcha box. It increases when I fail.
Well, I needed strongly non-overlapping genres. There are some other options, but you can understand why it's not science fiction vs. fantasy, memoir vs. history, science vs. nature, etc.
Also, who doesn't like food?
I think traffic lights and cookbooks is a good choice :-)
(the covers didn't display here).
I did think of diet books and thought there might be a single book with traffic lights on. So now I challenge you to find books about how to cook and serve real traffic lights :-)
The reCaptcha system often requires you to identify some image--usually pick out the cars, traffic lights or whatever in a street scene.
covers to decide which was actually food related and which might just have a metaphorical title
FWIW, we're currently allowing one error per page. The success rate is very very high.
Regarding the covers, I was only stating my reason for preferring a simpler captcha (the one-click checkbox) to log in, not implying anything is wrong with doing something more thematic to the site. It is human nature to choose the easier path when the alternative doesn't make a difference on the end result. :)
Right. Sometimes, if it's not so sure, it'll give you an image grid. That's the problem with the one-click checkbox—if Google doesn't like the smell of you, your browser or your IP, you're labelling crosswalks in Hell :)
ETA: the book-reading robot
>41 timspalding: Ah! I did miss one when I was signing in to my son's account but I had already messed up the password. However, the captcha let me through but I just had to re-enter the password (without having to redo the captcha after that).
Liking the cookbooks; they're nice and big and easy to see. As well as being book-related. My only problem is that my track pad seems to be sticking - but that's a personal issue.
Chances are you have your browser set to clear all cookies whenever you close the browser or session. If so, you will get it every time. We rely on a persistent session cookie to remember which users you have used to log into in the last 30 days.
I usually log out of sites as a security habit. I haven't had to redo the robot thing since a few days prior to the treasure hunt.
Also, now I really want to bake traffic light brownies!
I would like to translate the whole "robot-thing" into german. But in the translate page, can not find the yellow marked phrases in my librarything.de pages.
BTW: This occurs also on many other occasions.
Who can I find ALL not in german (or other languages) translated phrases?
If I log out and log back in the next day, do I get the check?
Ha. No AI involved. Just books and genres.
Then again, now that everything is AI, indeed, it's AI. Big-data AI. Machine learning AI. Strong AI!
World domination through superior recognition of cookbooks?
A note that's not as much fun, perhaps: Parts of book covers were cut off in the LT robot check for me today. On the worst if the batch, I could see the center of the cover but not all the edges, which mean the cover art, title, and author name were only partially visible.
Yeah, sorry. On the plus side, you can get it wrong multiple times and indeed you don't need to get them all correct at any time.
"Which of these covers contain Explicit Images?"
Right. We had hackers going through a giant database of old stolen credentials, trying their username/password combinations on LibraryThing. Something like a percent worked. We had been defending against such attacks by penalizing too many hits by IP address, or region, but the attackers in this case were willing to slow down their attack over weeks, and had many thousands of machines all around the world, presumably from a zombie bot-net. The Captcha cut this off completely.
They don't really care about the poles. They care about the pattern, and slow, meticulous overthinking looks a lot like being passed through an image recognition program. For that sort of traditional captcha, fast and slightly wrong is often better than slow and totally right.
This isn't the only site where I am seeing the Gotcha request. Thanks for working to keep things updated and fun at the same time.
According to English Wikipedia, a car is "a wheeled motor vehicle used for transportation. Most definitions of car say they run primarily on roads, seat one to eight people, have four tires, and mainly transport people rather than goods", which would rule out trucks.
Try going into your "Settings" (under the three dots in the top corner) and check the options under Privacy and under Passwords to make sure you have cookies and password-saving still set the way you want them to be. (And make sure that, under Manage Passwords, LibraryThing didn't somehow get set to Never Ask.)
If that doesn't work it might be something more complicated going wrong with the way your computer is set up, and it might be worth asking someone local to you who can help you with it.
I'm not unsympathetic. I get very frustrated at how hard life is made for those with mobility issues. I know. Life is hard enough without extra humps and steps and rough surfaces and obstacles being put in the way. The cookbook version makes life a lot easier for most of us. If some covers are hard for colour blind people to recognize, then I think you need to point out to the publishers that they are losing a group of customers who can't see titles of books they might otherwise want to buy.
Yes, although the system will remove a book that people misguess frequently. So I'd rather just let it work its own way out. At present, you can even get one wrong and still get in.