turkinthailand
Member
- Joined
- Aug 4, 2017
- Messages
- 381
- Reaction score
- 291
- Points
- 313
- Location
- Bangkok, Turkmenistan
- Gender
- Male
Mturk decides when to give captchas or not. Panda Crazy just tries to accept hits so I don't see how it would cause more Captchas. Just keep the timer at around 1000ms and that should be fine.Since yesterday I am getting huge numbers of captchas in PC, like, more than one a minute. Any clues? I am still running 0.3.4 so it's not that.
Leaving it on for a week is a long time. I recommend restarting the browser and the script every day. Sometimes the browser could have things still running in the background or memory leaks that could cause problems. A restart of the computer can fix many problems too but usually you can go for weeks without a restart.Seems I just needed to restart the script. It's been running for a week or more, oops!
Sounds like they might have changed something on there pages that PC is detecting as a captcha. I don't get captchas so I have no way of testing it. You can disable the captcha detection in the options general menu. Disable Captcha Alert should be set to true and it shouldn't give any captcha alerts. You can also try clearing out your cache and cookies from mturk to see if that might help.Hello, since yesterday I get enormous amounts of captchas in panda crazy.l Sometimes I fill them in. Then they come in the next hit again, even though was accepted.
I believe so if you use this link:Will Panda Crazy work with Accept_Random links on the worker site?
yes it will. There are two urls to make it work. You should use the url on worker.mturk.com rather than the old one so you are ready when they switch off the old site.Will Panda Crazy work with Accept_Random links on the worker site?
When no HITs are available, the Accept_Random links forward to a search page. Does this mean your script properly deals with this?yes it will. There are two urls to make it work. You should use the url on worker.mturk.com rather than the old one so you are ready when they switch off the old site.
Old url: https://www.mturk.com/mturk/welcome?pandacrazy=on&usenewsite=yes
New url: https://worker.mturk.com/?filters[search_term]=pandacrazy=on
Yes the script knows when it's looking for an accept page or a search page. So when it receives a search page from a panda it just ignores it and labels it as not found. It's not able to figure out if there really are no hits or mturk is just throttling it. Even in search mode it knows when to ignore and not to ignore a search page.When no HITs are available, the Accept_Random links forward to a search page. Does this mean your script properly deals with this?
Does it reload the Accept_Random page when a search page is found?Yes the script knows when it's looking for an accept page or a search page. So when it receives a search page from a panda it just ignores it and labels it as not found. It's not able to figure out if there really are no hits or mturk is just throttling it. Even in search mode it knows when to ignore and not to ignore a search page.
No the script does everything in the background so it doesn't reload pages. If it didn't find any hits it will go to the next job that needs to be collected. It can't know if there is no other hits or being throttled so if you know a hit is being throttled then just click on the goham button and it will constantly go to the panda url grabbing as many hits it can.Does it reload the Accept_Random page when a search page is found?
At this time it is not possible. The script does it's calculations by what it collects. If using the queue helper script then it will know when a hit is submitted or returned but doesn't recalculate the PE. That might change in the future but it still might not be totally accurate. It grabs a more accurate PE from other scripts so I was waiting for those scripts to work on the new site before I make a decision how to get it working. So it's planned to be added in the future but just need to wait.With PC on the old site, when looking at the dashboard it'll update your PE on PC if it's off in comparison. is it not possible to do at the moment with new site PC and the new dashboard?
No. Panda Crazy is useful in organizing your favorite hits, collecting hits with custom timers to lower pre chances and doing a search for hits from a requester so it can collect any new hits from them. To look for newer hits or unqualified hits you should be using something like Hit Finder with Mturk Suite extension or Overwatch script.Is Panda Crazy useful to find hits I'm not qualified for, or does it only ever show qualified hits?
Go into the details and make sure once is equal to false. Make sure you click on panda button for all batch hits. The once button is for hits you only want to collect once.What's the story on some hits where I press 'collect', job goes yellow, but only one hit is collected, and the collect button de-highlights itself? Ta.
If you have the Panda Crazy Queue Helper installed then you can go to https://worker.mturk.com/tasks?JRPC=nexthit and it will go to the next hit in your queue that isn't opened in another tab. The JRPC=nexthit also works for the old site too. I just updated that script last week with all the instructions and features for it here: http://www.mturkcrowd.com/posts/1031902/.And 2nd question... is there a URL I can bookmark that would "work on 1st job in queue?"
Sorry, I think we're talking about different scripts. Were you referring to https://greasyfork.org/en/scripts/23239-jr-mturk-panda-crazy-queue-helper ? I'm inquiring about https://greasyfork.org/en/scripts/19168-jr-mturk-panda-crazy which hasn't been updated in months.No the script does everything in the background so it doesn't reload pages. If it didn't find any hits it will go to the next job that needs to be collected. It can't know if there is no other hits or being throttled so if you know a hit is being throttled then just click on the goham button and it will constantly go to the panda url grabbing as many hits it can.