02/03 - Workity Workity Work Wednesday

glad you guys are making headway on finding the bugs... I still see posts everyday on the FB mturk group of people asking how to revert to the old MTS, or people who have stopped using it completely. Works great for me, but searching for a few hundred survery hits is different than multiple thounsands of batch hits.

 

early sleep tonight
gnite
except @AlexZ
bnite.

 

I only did like 1200 total over those days, lol. I can search 01/12 through today and that works fine.

Cursed dates, imo.

 

If I look over the past month, I get nothing. But setting the start date at 1/13, I was able to produce a contain result for the first time. So yay!

But for me, not just those days. I've picked several months here and there to just do a search for whatever. So just now, I checked 3/1/2020 to 7/1/2020. I can easily pull up all the HITs between those dates. But when I add in the term GeoHive, I can't get a result. But I chose to search for GeoHive because I saw that in the results, so I know that's something from that period. Did the same thing again between 9/30/2019 to 12/18/2019 and got all those HITs to come up. Search for Peter Burke which I saw among the results for the same period, just a spinning wheel and no results.

But I want to stress again @ChrisTurk , I'm good. Thrilled to be able to search in any way at all. So work on this if others need you to keep at it, and I'm happy to provide info if you're looking to nail it down, but I'm good.

 

Okay. You're right, it didn't work. It's a memory issue? I have plenty of RAM on my system, so I don't think that is a problem. I have almost 300,000 HITs. Is that the issue?

 

MASC Lab [A4DQ8Q2C0VB3F] Respond to different faces(~ 7 minutes) - $1.00 | PANDA
[div] style="border: 1px solid transparent; padding: 10px 0; display: flex; align-items: center; border-color: #ecf0f1; box-shadow: 0 1px 1px rgba(0,0,0,0.05); text-align: center;" |
[div] style="float: left; width: 25%;" |
Unrated[/div]
[div] style="float: left; width: 41%;" |[div] style="float: left; width: 50%;" |
Unrated[/div]
[div] style="float: left; width: 50%;" |
Unrated[/div][/div]
[div] style="float: left; width: 33%;" | $13.85 / hour
00:04:20 / completion time[/div]
[/div]

​

[div] style="width: 100%; display: flex; flex-items: center; margin-top: 5px;" |[div] style="width: 50%; display: inline-block; margin: 0 5px 0 0;" | Pros:
[/div]
[div] style="width: 50%; display: inline-block; margin: 0 0 0 5px;" | Cons:
[/div]
[/div]


Catchability: Great (HIT Last Seen: 23 seconds ago)
Qualifications: Total approved HITs >= 100; SurveyGroup [4787] DoesNotExist ; Exc: [-1231424536-273412] DoesNotExist ; 48df0769 DoesNotExist ; 26cb2be6 DoesNotExist ; HIT approval rate (%) >= 95; CR Research Group #1 >= 100; Location In US;
[div] style="margin-top: 5px;" |

To read MASC Lab's full profile check out TurkerView!​

[/div]
[div] style="display: none;" data-type="tv-review" data-revId="992712" data-reviewerId="24847" |[/div]

 

ETH CogSci MTurk [A2M5CRQ7MREP8G] Decision study - $0.5 base, ~$2.5 bonus for completion(~ 18 minutes) - $0.50 $0.01 bonus | PANDA
[div] style="border: 1px solid transparent; padding: 10px 0; display: flex; align-items: center; border-color: #ecf0f1; box-shadow: 0 1px 1px rgba(0,0,0,0.05); text-align: center;" |
[div] style="float: left; width: 25%;" |
Unrated[/div]
[div] style="float: left; width: 41%;" |[div] style="float: left; width: 50%;" |
Unrated[/div]
[div] style="float: left; width: 50%;" |
Unrated[/div][/div]
[div] style="float: left; width: 33%;" | $2.92 / hour
00:10:28 / completion time[/div]
[/div]

​

[div] style="width: 100%; display: flex; flex-items: center; margin-top: 5px;" |[div] style="width: 50%; display: inline-block; margin: 0 5px 0 0;" | Pros:
[/div]
[div] style="width: 50%; display: inline-block; margin: 0 0 0 5px;" | Cons:
[/div]
[/div]


Catchability: Okay (HIT Last Seen: 18 minutes ago)
Qualifications: Exc: [2141341462-273298] DoesNotExist ; Total approved HITs >= 1000; CR Research Group #2 DoesNotExist ; HIT approval rate (%) >= 95; CR Research Group #1 >= 100; Location In US;
[div] style="margin-top: 5px;" |

To read ETH CogSci MTurk's full profile check out TurkerView!​

[/div]
[div] style="display: none;" data-type="tv-review" data-revId="992713" data-reviewerId="2326" |[/div]

 

I guess I will just have to rely on QBC for everything. I just cannot make MTS work for me anymore. : ( Pretty disappointed.

 

CT and Kada will fix it. It was a complete re-write from the previous version that was stable for months. It's just taking a bit to find and fix bugs that only affect some users, but not others.

 

Night All.

 

If you don't care about exposing your 'sekrit batches' it'd be helpful if you exported the db & emailed it to me (hello@turkerview.com) / filedropped & DM it my way. Then I can just to a straight 1:1 test. FWIW, idc about your sekrit batches & will nuke it as soon as I'm done but that's odd lol


On that note if anyone has a true MTS tracker export w/ >100,000 hits it'd save me from hours trying to mock data lol

I've only cursory looked at the code, AFAIK if you do a contains search it's doing an absolute nightmare amount of computational work to try and make sure it finds what you want and that'll be exponential to the # of records you're pulling in the search

It's a similar issue to how TV used to die if you tried to search for HITs on the website, but req search was relatively fast.

Worse off b/c it's a client side db, so even a gif on the daily thread is competing with the db for resources (..in a very abstract way, that's not actually that impactful).

 

Thank you. I know they are doing their best. I really do appreciate all their help. I relied on the tracker a LOT, so I hope that can be fixed in future updates. For now, I will just have to disable it since it causes too many issues for me.

 

SeJeong [A2HY4W8SF02YKQ] Understanding an individual's thoughts and opinions - $0.15 | PANDA
[div] style="border: 1px solid transparent; padding: 10px 0; display: flex; align-items: center; border-color: #ecf0f1; box-shadow: 0 1px 1px rgba(0,0,0,0.05); text-align: center;" |
[div] style="float: left; width: 25%;" |
Unrated[/div]
[div] style="float: left; width: 41%;" |[div] style="float: left; width: 50%;" |
Unrated[/div]
[div] style="float: left; width: 50%;" |
Unrated[/div][/div]
[div] style="float: left; width: 33%;" | $16.88 / hour
00:00:32 / completion time[/div]
[/div]

​

[div] style="width: 100%; display: flex; flex-items: center; margin-top: 5px;" |[div] style="width: 50%; display: inline-block; margin: 0 5px 0 0;" | Pros:
[/div]
[div] style="width: 50%; display: inline-block; margin: 0 0 0 5px;" | Cons:
[/div]
[/div]


Catchability: Great (HIT Last Seen: a second ago)
Qualifications: HIT approval rate (%) >= 95; Location EqualTo US;
[div] style="margin-top: 5px;" |

To read SeJeong's full profile check out TurkerView!​

[/div]
[div] style="display: none;" data-type="tv-review" data-revId="992716" data-reviewerId="24847" |[/div]

 

New Requester Profile!

Blake Nichols [AM2RMJXFCD496] Answer a survey about diet preference - $1.00 | PANDA
[div] style="border: 1px solid transparent; padding: 10px 0; display: flex; align-items: center; border-color: #ecf0f1; box-shadow: 0 1px 1px rgba(0,0,0,0.05); text-align: center;" |
[div] style="float: left; width: 25%;" |
Fair[/div]
[div] style="float: left; width: 41%;" |[div] style="float: left; width: 50%;" |
Unrated[/div]
[div] style="float: left; width: 50%;" |
Unrated[/div][/div]
[div] style="float: left; width: 33%;" | $16.29 / hour
00:03:41 / completion time[/div]
[/div]

​

[div] style="width: 100%; display: flex; flex-items: center; margin-top: 5px;" |[div] style="width: 50%; display: inline-block; margin: 0 5px 0 0;" | Pros:
[/div]
[div] style="width: 50%; display: inline-block; margin: 0 0 0 5px;" | Cons:
[/div]
[/div]


Catchability: Okay (HIT Last Seen: 17 minutes ago)
Qualifications: HIT approval rate (%) > 97; Location EqualTo US;
[div] style="margin-top: 5px;" |

To read Blake Nichols's full profile check out TurkerView!​

[/div]
[div] style="display: none;" data-type="tv-review" data-revId="992717" data-reviewerId="11151" |[/div]

 

If you have 3 minutes and are curious, you can run Speedometer 2 benchmark ( https://browserbench.org/Speedometer2.0/ ) and see what your score is. That will tell you how good your CPU is for web stuff specifically (anything that you do in a browser basically). You probably want to score at least 60+ in 2021.

Here's a few charts to compare your score to

 

Okay, thank you for your help. I hope the problem(s) can be fixed in future updates. I know you are swamped with questions and issues about it, so I appreciate your help!

 

if that’s the way you talk about react which I feel you actually like, I’d hate to hear your real thoughts on ruby

 

I still have a galaxy s5.. if mts was a mobile extension I'd have found so many more bugs

 

The db engine isn't related to react

Also me talking about ruby would be less intelligent than a lobotomized monkey talking about newton's 16th law of interdimensional stock trading.