Topic: The mysterious bitcoin flower opens once a day and reveals a free bitcoin - page 5. (Read 18326 times)
August 14, 2011, 10:53:32 AM
August 14, 2011, 10:43:38 AM
I'd like if you could also show a small animated clock with the current CDT date & time or just have time left to end of day & if the flower had opened as yet in that 24 hr period
Also a different favicon than the forum's, say of your flower opened (or not):
http://www.favicon.cc/
Nice concept, I'm looking forward to seeing how it evolves
Also a different favicon than the forum's, say of your flower opened (or not):
http://www.favicon.cc/
Nice concept, I'm looking forward to seeing how it evolves
I, too, see it evolving, Otoh. The idea will diffidently evolve, now that evolve has commented on this thread.
legendary
Activity: 1862
Merit: 1011
Reverse engineer from time to time
August 14, 2011, 10:34:10 AM
If I understood correctly then one doesn't need to refresh the page (multiple times) to see the blooming flower, it will show up when it changes if the page is just left open?
If so good to say this on the page perhaps...
I failed to see any JS that would make it bloom without refresh, sorry. If so good to say this on the page perhaps...
August 14, 2011, 09:37:49 AM
If I understood correctly then one doesn't need to refresh the page (multiple times) to see the blooming flower, it will show up when it changes if the page is just left open?
If so good to say this on the page perhaps...
If so good to say this on the page perhaps...
August 14, 2011, 09:26:10 AM
I'd like if you could also show a small animated clock with the current CDT date & time or just have time left to end of day & if the flower had opened as yet in that 24 hr period
Also a different favicon than the forum's, say of your flower opened (or not):
http://www.favicon.cc/
Nice concept, I'm looking forward to seeing how it evolves
Also a different favicon than the forum's, say of your flower opened (or not):
http://www.favicon.cc/
Nice concept, I'm looking forward to seeing how it evolves
August 14, 2011, 09:18:36 AM
If the likelihood for a flower opening is the same every minute, and such that the average time interval between two successive openings is equal to one day, then we get the following statistical distribution for the time interval between two successive openings:
http://i55.tinypic.com/v2uvco.png
Probability that time interval is < 0.9 days/0.5 days/0.25 days/1 hour = 59.6/39.7/22.4/4.1 %
Probability that time interval is > 1.0 days/1.5 days/2.0 days/3.0 days = 36.6/22.3/ 13.5/5.0 %
This makes more sense than my previous post with the Rayleigh distribution.
Matlab/Octave code:
http://i55.tinypic.com/v2uvco.png
Probability that time interval is < 0.9 days/0.5 days/0.25 days/1 hour = 59.6/39.7/22.4/4.1 %
Probability that time interval is > 1.0 days/1.5 days/2.0 days/3.0 days = 36.6/22.3/ 13.5/5.0 %
This makes more sense than my previous post with the Rayleigh distribution.
Matlab/Octave code:
Code:
N=100000;% number of days to simulate
increment = 1/(24*60); % 1 min granularity
T=zeros(1,N); % init memory
for k=1:N,
T_inc = 0;
while rand(1,1)>increment,% An event occurs if random number in [0;1] is <= increment
T_inc = T_inc + increment;
end
T(k) = T_inc;
end
mu = mean(T),
value_below_090__050__025__004 = [100*length(T(T<0.9))/length(T), 100*length(T(T<0.5))/length(T), 100*length(T(T<0.25))/length(T), 100*length(T(T<1/24))/length(T)]
value_above_100__150__200__300 = [100*length(T(T>1.0))/length(T), 100*length(T(T>1.5))/length(T), 100*length(T(T>2.0))/length(T), 100*length(T(T>3.0))/length(T)]
figure
hist(T,50); title(['Histogram: Same probability for an event in each time instant']);
xlabel(['Time T between two successive events, mean=',num2str(mu)])
increment = 1/(24*60); % 1 min granularity
T=zeros(1,N); % init memory
for k=1:N,
T_inc = 0;
while rand(1,1)>increment,% An event occurs if random number in [0;1] is <= increment
T_inc = T_inc + increment;
end
T(k) = T_inc;
end
mu = mean(T),
value_below_090__050__025__004 = [100*length(T(T<0.9))/length(T), 100*length(T(T<0.5))/length(T), 100*length(T(T<0.25))/length(T), 100*length(T(T<1/24))/length(T)]
value_above_100__150__200__300 = [100*length(T(T>1.0))/length(T), 100*length(T(T>1.5))/length(T), 100*length(T(T>2.0))/length(T), 100*length(T(T>3.0))/length(T)]
figure
hist(T,50); title(['Histogram: Same probability for an event in each time instant']);
xlabel(['Time T between two successive events, mean=',num2str(mu)])
Great analysis
. This is definitely the way to do it and is what I suggested back in post 24. It's similar to how these things are handled in nature (radioactive decay). 
August 14, 2011, 07:12:43 AM
If the likelihood for a flower opening is the same every minute, and such that the average time interval between two successive openings is equal to one day, then we get the following statistical distribution for the time interval between two successive openings:
Probability that time interval is < 0.9 days/0.5 days/0.25 days/1 hour = 59.6/39.7/22.4/4.1 %
Probability that time interval is > 1.0 days/1.5 days/2.0 days/3.0 days = 36.6/22.3/ 13.5/5.0 %
This makes more sense than my previous post with the Rayleigh distribution.
Matlab/Octave code:
Probability that time interval is < 0.9 days/0.5 days/0.25 days/1 hour = 59.6/39.7/22.4/4.1 %
Probability that time interval is > 1.0 days/1.5 days/2.0 days/3.0 days = 36.6/22.3/ 13.5/5.0 %
This makes more sense than my previous post with the Rayleigh distribution.
Matlab/Octave code:
Code:
N=100000;% number of days to simulate
increment = 1/(24*60); % 1 min granularity
T=zeros(1,N); % init memory
for k=1:N,
T_inc = 0;
while rand(1,1)>increment,% An event occurs if random number in [0;1] is <= increment
T_inc = T_inc + increment;
end
T(k) = T_inc;
end
mu = mean(T),
value_below_090__050__025__004 = [100*length(T(T<0.9))/length(T), 100*length(T(T<0.5))/length(T), 100*length(T(T<0.25))/length(T), 100*length(T(T<1/24))/length(T)]
value_above_100__150__200__300 = [100*length(T(T>1.0))/length(T), 100*length(T(T>1.5))/length(T), 100*length(T(T>2.0))/length(T), 100*length(T(T>3.0))/length(T)]
figure
hist(T,50); title(['Histogram: Same probability for an event in each time instant']);
xlabel(['Time T between two successive events, mean=',num2str(mu)])
increment = 1/(24*60); % 1 min granularity
T=zeros(1,N); % init memory
for k=1:N,
T_inc = 0;
while rand(1,1)>increment,% An event occurs if random number in [0;1] is <= increment
T_inc = T_inc + increment;
end
T(k) = T_inc;
end
mu = mean(T),
value_below_090__050__025__004 = [100*length(T(T<0.9))/length(T), 100*length(T(T<0.5))/length(T), 100*length(T(T<0.25))/length(T), 100*length(T(T<1/24))/length(T)]
value_above_100__150__200__300 = [100*length(T(T>1.0))/length(T), 100*length(T(T>1.5))/length(T), 100*length(T(T>2.0))/length(T), 100*length(T(T>3.0))/length(T)]
figure
hist(T,50); title(['Histogram: Same probability for an event in each time instant']);
xlabel(['Time T between two successive events, mean=',num2str(mu)])
August 14, 2011, 06:44:40 AM
Quote
randomize the winning link name so it cannot be used twice.
My competing project has this protection, also CAPTCHA, IP blacklist and even statistical analyzer. Regardless this bots still trying to cheat me. I give away big number of relative small prizes, so <10% cheat level is tolerable, but you pay only one big bounty per day, which gives big incentive to the bot makers and they can easily destroy your business even with few successful cheat attempts.Another problem is that you give away only one prize per day and all people except one lucky will get nothing, although they have spent much time on your website catching bounty. This fact can disappoint and many of them could lose hope to win and never return again. I would suggest to pay some "booby prize".
August 14, 2011, 05:20:10 AM
Sorry to start trouble here with the bot thing...I wouldn't write a bot for this myself either, but I did want to bring it to the OP's attention that sooner rather than later, someone world.I think the only way to defeat bot scraping here would be to have to type in the captcha before you can see if the flower is open or not.
I do like her idea and admire her for her generosity and hope she has fun with it.
If it was me though, I would have 10 x .1 BTC prizes a day...or even 20 x.05 BTC....probably get more frequent visitors like that...idk.
But anyway, keep up the good work!
August 14, 2011, 03:41:08 AM
This is why good will doesn't work, except in the form of small faucets.
Unknown people will instantly come up with ways to monetize your 'charity' or even plain out steal from it, or make it their regular income through proxies and decaptchas.
It's a generous idea but all it does is attract scumbags. People who genuinely visit the site every now and then will never get anything.
Unknown people will instantly come up with ways to monetize your 'charity' or even plain out steal from it, or make it their regular income through proxies and decaptchas.
It's a generous idea but all it does is attract scumbags. People who genuinely visit the site every now and then will never get anything.
this is pretty much true. by the way I'm not actually doing what I said was possible, I'm just disclosing what is possible. I don't care to write a bot for this, I'm above this sort of thing. I wrote bots for people who wanted to auto-play perfect strategy at casino games, not to rip off 1BTC gifts
August 14, 2011, 02:39:57 AM
This is why good will doesn't work, except in the form of small faucets.
Unknown people will instantly come up with ways to monetize your 'charity' or even plain out steal from it, or make it their regular income through proxies and decaptchas.
It's a generous idea but all it does is attract scumbags. People who genuinely visit the site every now and then will never get anything.
Unknown people will instantly come up with ways to monetize your 'charity' or even plain out steal from it, or make it their regular income through proxies and decaptchas.
It's a generous idea but all it does is attract scumbags. People who genuinely visit the site every now and then will never get anything.
August 14, 2011, 02:11:21 AM
but lets be honest.. if you can code up a bot quickly to analyze pictures that well.. maybe that person should be submitting a resume to
several companies to make a bit more then 1 btc a day ;-)
several companies to make a bit more then 1 btc a day ;-)
I'd love to but I don't know any companies that want an ex-blackhat script kiddie. They are probably out there though but who's hiring these days...
August 14, 2011, 02:02:06 AM
and thus the bot really cannot screen scrape the html/jpg for anything interesting...
sure it can, a quick md5 hash check will reveal that the jpg has changed, regardless of the filename
even if you did it all in flash or java or something invisible to a bot, I can load it in the background with a bot and take a screenshot of the rendered flower and take a checksum of this screenshot and wait for it to change.
some kind of randomized animated thing with changing colors and patterns and such would make it more difficult, but as long as the "open" flower looks significantly different than the "closed" flower to a human, it might be possible for a bot's algorithms to determine the same thing
crap. good point.
i like the idea of randomizing the picture in some way also each time you need to change it.. use ?imagemagick? or whatever
to change it ever so slightly to screw the checksum up each time.
but lets be honest.. if you can code up a bot quickly to analyze pictures that well.. maybe that person should be submitting a resume to
several companies to make a bit more then 1 btc a day ;-)
to the OP.. i like your webpage. very cool. you should put up a banner though to offset costs. no one would care i think.
August 14, 2011, 01:51:25 AM
and thus the bot really cannot screen scrape the html/jpg for anything interesting...
sure it can, a quick md5 hash check will reveal that the jpg has changed, regardless of the filename
even if you did it all in flash or java or something invisible to a script fetching the page, I can load it in the background with a bot and take a screenshot of the rendered flower and take a checksum of this screenshot and wait for it to change.
some kind of randomized animated thing with changing colors and patterns and such would make it more difficult, but as long as the "open" flower looks significantly different than the "closed" flower to a human, it might be possible for a bot's algorithms to determine the same thing
I've written many casino bots, and have defeated all the anti-bot tech put out by all the casino software I've encountered and needed to write a bot for. The hardest was a video poker game which each time a card was shown in a particular spot, it was slightly (randomly) different in card location (by a few pixels) and in the font edge details like angle and anti-aliasing. I had to resort to custom training OCR software instead of my tried and true screenshot checksum hash lookup method
August 14, 2011, 01:49:57 AM
about the bot thing.. cannot the OP simply move the closed.jpg to closed.jpg.bak when he is ready to release the coin...
and then move open.jpg.bak to closed.jpg for 15 minutes...
make sure the file sizes are the same and any other relevant info inside the jpg description (whatever that is called).
and thus the bot really cannot screen scrape the html/jpg for anything interesting...
but then one has to be careful they do not learn the link url for claiming the coin.. but there is a catchpa.. which is nice..
but a quick text message to the bot runner would alert then in about 15-20 seconds to hop on the PC and finish the task...
not totally reliable but i am sure they could claim the coin a few times a week...
in summary.. the real solution in my view to defeat the bot is to not allow them to scrape anything useful and randomize
the winning link name so it cannot be used twice.
and then move open.jpg.bak to closed.jpg for 15 minutes...
make sure the file sizes are the same and any other relevant info inside the jpg description (whatever that is called).
and thus the bot really cannot screen scrape the html/jpg for anything interesting...
but then one has to be careful they do not learn the link url for claiming the coin.. but there is a catchpa.. which is nice..
but a quick text message to the bot runner would alert then in about 15-20 seconds to hop on the PC and finish the task...
not totally reliable but i am sure they could claim the coin a few times a week...
in summary.. the real solution in my view to defeat the bot is to not allow them to scrape anything useful and randomize
the winning link name so it cannot be used twice.
August 14, 2011, 01:42:07 AM
hahahahahahahahahahahahahahaha!!!!!!!!!!!!
August 14, 2011, 01:35:59 AM
Inglourious Basterds
another great movie!!! tarantino was on top of his game on that one....amazing dialogue broken up with scenes of brutal violence.... gotta love it!
This reminds of this picture that had me laughing my drink out of my nose when I first saw it.
August 14, 2011, 12:59:09 AM
Inglourious Basterds
another great movie!!! tarantino was on top of his game on that one....amazing dialogue broken up with scenes of brutal violence.... gotta love it!
"gots a german here 'at wants to die for country....oblige him"
http://www.youtube.com/watch?v=yIja0yValiE&feature=related
really cool interview with eli roth about inglorious basterds (he played donnie aka "the bear jew" ...for those that dont know, he also directed hostel, cabin fever, and few other movies)
http://www.youtube.com/watch?v=SwMoq9ZHxpw&feature=relmfu
August 14, 2011, 12:48:23 AM
This idea has great potential but, once again, only if it is animated. You could even have an embedded video to get around the bots.
Discuss?
Discuss?
Actually I wanted to make the whole thing animated (using jquery, leaves flying around etc). But in order to keep it simple I decided not do go this way in the beginning. I might add this late on. And your idead of using a video could really be a big step in preventing bots from scraping the site.
Post #25 really started a lot of things in my head rolling, some great ideas there! Have to go back to my thinking box and …
August 13, 2011, 09:32:26 PM
I just gotta say this thread is awesome. Where else can you get a discussion of a cool new and artistic bitcoin site, website monetization, bot arms races, and Tarantino films all in one shot (and for the record, I really liked Death Proof...the other one not so much).
...and now add magic to the mashup, Steve:
Jump to: