Panda Crazy Script for Panda's

[ScrapingForQuarters]

Thanks for the update, and I look forward to the new release. Thanks again for your responsiveness and all the effort you put into this excellent tool.

 

[Johnnyrs]

Today I released a new update (0.4.0) to greasyfork. Be sure to update it if greasemonkey/tampermonkey doesn't update it automatically. Should update it automatically sometime today. Be sure to backup your data by exporting it from the jobs menu at top so you won't lose anything if something goes wrong.

New Features:

• Added a captcha counter on the bottom after the queue watcher tab. The script will count each hit it accepts and will reset the counter when it gets to a captcha. There is a new option in the generals menu for the number of hits before a captcha appears. It's at 35 by default.
• Added a new option in the generals menu to disable the captcha counter for people who have masters and don't get captcha's.
• Added a new option for the number of hits before a captcha appears. This number will be used for future updates but now it's only used to make the captcha counter disappear after (this number+5) hits accepted without a captcha.
• The Search Jobs and Search Mturk window is now redesigned. It should be faster for large amount of jobs instead of taking around 10 seconds to load. The window looks a bit different too. The cancel or close button is at the top right of the window.
• The captcha alert window now shows more details of the hit it was trying to grab and has a link to the panda it tried.

Bugs Fixed:

• Fixed a bug where you couldn't delete multiple jobs in the search jobs window.
  
• Fixed a bug where the search jobs didn't show all of the jobs you have. This happened because some jobs became an orphan and they had no position in any of the tabs. This might have happened due to moving them to different tabs or an error when deleting it. Now when you search jobs it will actually delete the orphans to make sure those jobs won't cause a problem.

This update is a minor update for bugs and a redesign. I have tested it on more than 400 jobs saved and it it searches through the jobs in 3 seconds instead of around 10 seconds. There is still a long delay at the beginning when creating all the job cells. If you are using the queue helper then be sure to update it or it will stop working due to a bug in the version check. The queue helper also fixes the url on the last hit of your queue so it doesn't give a silly error even if you don't use PC.

 

[ScrapingForQuarters]

Things started okay once I realized the printed counter was going to stay at 0 till I had a Captcha. (I restart almost every night and use Vivaldi in the morning, and Chrome for the rest of the day, so this unknown state for starting again will be frequent for my usuage.) It started incrementing after the first captcha as expected and I was a happy camper. But then it disappeared as if I had changed the options on it. I don't recall for sure if I went to the general options just to look before this happened or not, but I certainly didn't turn it off on purpose. After that I went back, and they had been turned off for and the # HITs to captcha was set to 0. I turned them back on and reset to 35, but the counter did not reappear.

One possible factor is that it hit a Stop on queue limt (I had Zoltars to pause collection after I had collected 5 of them)

I did see improved output on the Captcha that I appreciated and I clicked on the link at least once and maybe several times and was able to enter the required string and continue as expected.

But later, I did that procedure, entered the string, came back to PC tab, but the unpause button would not do anything. I had to eventually close that window and start it up again. It still was not displaying the Captcha count once I restarted the PC tab ... not even the 0 it was displaying the first time I tried it.

It's not a big deal, but I wanted to turn collection off while I was dealing with the captcha (I didn't want to get more collected as soon as I unpaused) I can't change (in particular stop) a collection when it's in the captcha paused state. Why you might ask? I was using Subply .01 hits to push the counter up to force my first Captcha (so I knew where I was w.r.t. the count) and they collect very quickly if they're active.

 

[ScrapingForQuarters]

you weren't kidding about trying to discourage searches ...

can't do the fake a GID with a RID ... not atllowed.

So I have to copy over the RID I had scarfed for turkopticon with a GID from some other entry, re-enter that in an Add Job, erase it, and then recopy and paste the RID I had originally. Not exactly tons of fun.

Is there consideration for having a separate add search?

On the display, even though I cleared out the GID field (and went back and double checked it was clear) it is still displaying that GID I used (which is from a different requester) as the 3rd line of text, though at least it is showing [] in the single/double click area. So it Friendly Requester name on line 1,0.00 on line 2, and then the bogus GID on line 3.

 

[ScrapingForQuarters]

Okay, figured out one issue. The option is called
Disable Captcha Counter, but it works opposite of what it should.

When set to true, it turns the counter off.
When set to false, it turns it on.

 

[ScrapingForQuarters]

Rats. got that backwards, but it still works opposite of what it should.

when set to true, it's displaying the count.

when set to false, it hides it.

Personally, I think it would be clearer if it said

Print Captcha Counter
and you set it true to print it and false to not.

 

[Johnnyrs]

After that I went back, and they had been turned off for and the # HITs to captcha was set to 0. But later, I did that procedure, entered the string, came back to PC tab, but the unpause button would not do anything. I had to eventually close that window and start it up again.

Yes I was trying something that didn't work and forgot to take out that code. It should stay at the 35 and count up at the beginning. The 35 is just used to turn it off if there is no captcha received after that number + 5.

can't do the fake a GID with a RID ... not atllowed.
On the display, even though I cleared out the GID field (and went back and double checked it was clear) it is still displaying that GID I used (which is from a different requester) as the 3rd line of text, though at least it is showing [] in the single/double click area.

I didn't change anything about adding a job so that trick still works but not recommended. Just add any panda link and change that. There is no need to delete the Group ID. In search mode the script only uses the requester ID. When it finds a hit it fills in the group id and does a regular panda. No I don't want to have an add search option. Search mode should be used after finding a good requester by using a panda in the first place. Works better with the panda crazy helper or search mturk button because it fills in the requester ID if you use the panda or once buttons. Mturk doesn't always provide the requester ID in a preview or panda page so the script can't fill it in.

Print Captcha Counter and you set it true to print it and false to not.

Yes this was a mistake. I'm just going to label it differently.

 

[Johnnyrs]

Today I released a new update (0.4.1) to greasyfork. Be sure to update it if greasemonkey/tampermonkey doesn't update it automatically. Should update it automatically sometime today. Be sure to backup your data by exporting it from the jobs menu at top so you won't lose anything if something goes wrong.

Bugs Fixed:

• Fixed a bug where the new captcha counter would change captcha number to 0 and not count back up.
  
• Changed how the script displays the jobs to make it faster at load up. Tested it with 400 jobs and the delay is better now. I still recommend to have under 400 jobs but it shouldn't cause a lot of slowness now.
  
• Changed the text to disable captcha counter to: Show Captcha Counter Text
• Will now show unknown if groupid is blank.
• Will stop a panda if groupid is blank.
• Disabled moving of tabs due to a bug when moving jobs to the moved tab.

Mostly this update was a bug fix but a major change to how it displays the jobs. Before I had it add the job cells to the tabs which made it slow because Jquery UI had to do a lot of sorting calculations for each job. Now I add the job cells to a div and then append it to the tab which makes Jquery UI only do 1 sorting calculation. I might change lot of Jquery stuff into plain javascript to make things go faster in future updates. Next update I plan to look at groupings and make sure it's working correctly. I might rethink of a way to create groupings too because it might cause problems the way I have it now.

 

[Kat8mouse]

Today I released a new update (0.4.1) to greasyfork. Be sure to update it if greasemonkey/tampermonkey doesn't update it automatically. Should update it automatically sometime today. Be sure to backup your data by exporting it from the jobs menu at top so you won't lose anything if something goes wrong.

Bugs Fixed:

• Fixed a bug where the new captcha counter would change captcha number to 0 and not count back up.
  
• Changed how the script displays the jobs to make it faster at load up. Tested it with 400 jobs and the delay is better now. I still recommend to have under 400 jobs but it shouldn't cause a lot of slowness now.
  
• Changed the text to disable captcha counter to: Show Captcha Counter Text
• Will now show unknown if groupid is blank.
• Will stop a panda if groupid is blank.
• Disabled moving of tabs due to a bug when moving jobs to the moved tab.

Mostly this update was a bug fix but a major change to how it displays the jobs. Before I had it add the job cells to the tabs which made it slow because Jquery UI had to do a lot of sorting calculations for each job. Now I add the job cells to a div and then append it to the tab which makes Jquery UI only do 1 sorting calculation. I might change lot of Jquery stuff into plain javascript to make things go faster in future updates. Next update I plan to look at groupings and make sure it's working correctly. I might rethink of a way to create groupings too because it might cause problems the way I have it now.

Thanks for all you do -- this really is an awesome program!!!

 

[ScrapingForQuarters]

Low prioirty issue, but maybe fairly easy to implement since you already must be comparing all the pay rates to pick the highest when there's more than one.

A search I can't do with Turkmaster ...

I have a requester that posts 2 styles of HITs at the same time ...
the good paying one, and then a .01 "trouble ticket" HIT (the second of which stays up much longer than the "good one.")

So if I had a PC option to restrict on how much any accepted HITs must pay,
I could wait for the one I really want, even while the $.01 HIT is still up ...
just in case they release another batch of the big payers.

Otherwise, if only the .01 one is up, I can't initiate the search because it immediately accepts a HIT I don't want.

This of course is a pretty rare situation, but there may well be other requesters,
e.g., Rekognition, where they put out various HITs that pay anywhere from .20 to .40, but I only want to do say ones that pay .35 or more, and I believe they do change the GIDs from time to time, so I can't depend on PandAs to catch them. I might run this in the evening when a) I don't have any (or many) big money PandAs running, and b) I'm tired, but I could do things that don't require a lot of focus, should they happen to come up, and they've made it worth my while to drag myself back to the computer.

 

[catnapped]

I don't know if it's workable (or if for whatever reason you'd want to include it), but is there any way to put a ceiling on the auto slowdown? I know when I don't watch it sometimes it gets into the high 2000s. I'd rather not have PREs but I don't want it limiting refreshes quite that high either!

 

[Johnnyrs]

I don't know if it's workable (or if for whatever reason you'd want to include it), but is there any way to put a ceiling on the auto slowdown? I know when I don't watch it sometimes it gets into the high 2000s. I'd rather not have PREs but I don't want it limiting refreshes quite that high either!

Right now the best way is just press the reset timer and it will go back down to normal but that means you have to pay close attention to the elapsed time. I've been wanting to put a way for it to go faster after not getting pre's for 100 or so pages. I actually was testing it in the Pantha Collector Script but haven't gotten around to really test it and put it in this script yet. Maybe I'll add it later and put a button or option to toggle that feature on or off. It's not hard at all I just worry about it going faster while I'm still working on my queue. Have to figure out how to do it without causing more pre's while working.

 

[Johnnyrs]

So if I had a PC option to restrict on how much any accepted HITs must pay,
I could wait for the one I really want, even while the $.01 HIT is still up ...
just in case they release another batch of the big payers.

You can try the search option for minimum pay to only search for hits higher than .01. In the future I want to add an exclude and include string for the title of the hit so it only accepts specific hits with the string in the title or skip over hits with a string in the title. Haven't really figured out how to make that happen yet but it will be in the search option window also.

 

[arg]

I'm having an issue with the tabs. It started out fine and I could add new tabs and rearrange them. now the little + that makes new tabs disappeared and I cannot rearrange the tabs. Anyone know how I might fix this?

 

[Johnnyrs]

I'm having an issue with the tabs. It started out fine and I could add new tabs and rearrange them. now the little + that makes new tabs disappeared and I cannot rearrange the tabs. Anyone know how I might fix this?

The add button disappeared in the last update by mistake. Have to wait for a new update for this to be fixed, Sorry. You can't move tabs anymore because it never worked anyway and the fix would be lot of work so I'm going to wait for some other stuff first before I make it work again.

 

[arg]

The add button disappeared in the last update by mistake. Have to wait for a new update for this to be fixed, Sorry. You can't move tabs anymore because it never worked anyway and the fix would be lot of work so I'm going to wait for some other stuff first before I make it work again.

Ok, I thought I did something to screw it up. Your script is awesome, thanks!!

 

[Maxzine]

The buttons are not working on mturk for me with the helper. I have the panda crazy script up and running. I turned every other script off to see if something was interfering. Any idea why it's not working?

 

[Johnnyrs]

The buttons are not working on mturk for me with the helper. I have the panda crazy script up and running. I turned every other script off to see if something was interfering. Any idea why it's not working?

I tested everything right now and it still works for me. Make sure you update the main Panda Crazy script and the Panda Crazy Helper script. You could even try removing those scripts and reinstalling them. As long as Panda Crazy is running in the same browser you are trying to add with buttons it should work. I tried it both in Chrome and Firefox. Make sure those are updated as well.

 

[Johnnyrs]

Today I released a new update (0.4.2) to greasyfork. Be sure to update it if greasemonkey/tampermonkey doesn't update it automatically. Should update it automatically sometime today. Be sure to backup your data by exporting it from the jobs menu at top so you won't lose anything if something goes wrong.

New Features:

• Will now show release notes for current updated version at first start after an update.
• Can get a lot more help for the script at http://pandacrazy.allbyjohn.com/
• Added an option to maybe speed up pantha mode but could cause more pre's.
• Time accepted is now in a different color so it's easier to see in the accepted status window.
• When exporting alarm sounds the file will end with "_w_alarms" to distinguish it from others with no alarms.
• When importing a file the browse button will only show files ending with .json so data files are easier to find.
• When searching through jobs you can now sort by jobs that have the once option enabled to make it easier to find disposable jobs.

Bugs Fixed:

• Fixed a bug where the add tab button disappeared by mistake.
  
• Fixed a bug where the captcha counter may not disappear after not getting any captchas after the limit.
• May have fixed a bug when a hit is collected in between grabbing the queue hits and not having the newest hit in queue watch.
• Fixed a bug when collection is stopped because once is enabled but the job cell stays red.
• Fixed a bug where trying to change once in the details window will not work correctly.
• Fixed a bug when changing options and it would change it before you pushed the save button.
• Fixed a bug where you couldn't put a job into Save Search Group ID Mode.

This is mostly a bug fix but some new features could be useful to some. I am still working on the help web site so it's not complete but the important parts are there.

 

[mordea]

I have to admit that the work you've put into this for both yourself and other workers is pretty admirable. Awesome stuff.