Page Index Toggle Pages: 1 [2]  Send TopicPrint
Locked Topic [CLOSED] Preventing Biases (Read 8308 times)
Administrator
Forum Administrator
*****
Offline


Yummm

Posts: 7
Location: Modders Rile
Joined: Oct 7th, 2014
Gender: Male
Re: Preventing Biases
Reply #15 - Feb 17th, 2002 at 12:57am
 
KyleYankan wrote on Feb 17th, 2002 at 12:38am:
I like this, but  ithink the first release should be marked somehow, like my suggest, with a [BETA] tag or something of the nature

Why that? Wouldn't that confuse people unnecessarily?
  

The Administrator.
Back to top
WWW  
IP Logged
 
KyleYankan
Senior Member
****
Offline


The KyleYCam!

Posts: 341
Joined: Dec 25th, 2001
Gender: Male
Re: Preventing Biases
Reply #16 - Feb 17th, 2002 at 1:34am
 
Quote:
Why that? Wouldn't that confuse people unnecessarily?


Yes, some but if there using YaBB and BoardMod they must be used to confusion. But like Mike T. said, time is a concern here. This would make it available to the public, and hopefully some of the bugs will be uncovered more quickly this way, before more people downlaod it. Besides i think a big ol [BETA] stamp wont confuse most people anyways.
  


//-KyleY-\\

Steal his bandwidth! He doesn't want it!
Back to top
IP Logged
 
Michael
God Member
*****
Offline


Recursion \Re*cur"sion\,
n. - See recursion.

Posts: 1003
Joined: Oct 23rd, 2001
Gender: Male
Re: Preventing Biases
Reply #17 - Feb 17th, 2002 at 1:44am
 
KyleYankan wrote on Feb 17th, 2002 at 12:38am:
ithink the first release should be marked somehow, like my suggest, with a [BETA] tag or something of the nature

I don't think there should there should be any "first release". I think that if we do this 3x tested system, then mods should stay in the queue until they're completely tested, not be put into the Approved section until they really HAVE been approved. It would be defeating the purpose.
  

~ Michael ~
-------------
The MikeCam
A truly wise man never plays leapfrog with a unicorn.
Back to top
IP Logged
 
KyleYankan
Senior Member
****
Offline


The KyleYCam!

Posts: 341
Joined: Dec 25th, 2001
Gender: Male
Re: Preventing Biases
Reply #18 - Feb 17th, 2002 at 1:47am
 
True, True, So the current summary so far is, Have 3 people test it, discuss it, and then upload it?
  


//-KyleY-\\

Steal his bandwidth! He doesn't want it!
Back to top
IP Logged
 
Michael
God Member
*****
Offline


Recursion \Re*cur"sion\,
n. - See recursion.

Posts: 1003
Joined: Oct 23rd, 2001
Gender: Male
Re: Preventing Biases
Reply #19 - Feb 17th, 2002 at 1:54am
 
*eeeennnnhhh* Wrong. Tongue

The file gets uploaded to the server when the modwriter submits it. It gets listed (with a download link) in the "Queue" page. 3 people test it. There is no discussion unless there are major differences in scoring.
  

~ Michael ~
-------------
The MikeCam
A truly wise man never plays leapfrog with a unicorn.
Back to top
IP Logged
 
throesch
God Member
*****
Offline


I will own you all. will
you be on the good list?

Posts: 561
Location: Arab
Joined: Nov 9th, 2001
Gender: Male
Re: Preventing Biases
Reply #20 - Feb 17th, 2002 at 2:01am
 
ok i like that idea but who is the testers.

im always ready i have both unix and win32 servers i can test on
  
Back to top
ICQYIM  
IP Logged
 
] hirnsub [
New Member
*
Offline


If something can go wrong
- it will!

Posts: 38
Joined: Feb 6th, 2002
Re: Preventing Biases
Reply #21 - Feb 19th, 2002 at 12:51am
 
wow, quite a lot of ideas
mine is:
-as i was the one who wanted to test mods hardly and safe i wouldn't like the idea of mods gettin reviewed and re-reviewed again. but i think the approved database should not be in hurry to list mods, they get listed from the moment of submitting, that should be okay for the mod writer!!! the first review should take place on the same day, of course, and feedback will be given, so if the mod writer did some changes and the first reviewer is ok with it, then the second AND third reviewer get involved and test the mod too, after all three rated and agree it gets listed in the approved database - or not Grin     -     ? ? ?
  

-=oX] hirnsub [xo=-
If something CAN go wrong - IT WILL !
Back to top
ICQ  
IP Logged
 
Michael
God Member
*****
Offline


Recursion \Re*cur"sion\,
n. - See recursion.

Posts: 1003
Joined: Oct 23rd, 2001
Gender: Male
Re: Preventing Biases
Reply #22 - Feb 19th, 2002 at 1:16am
 
Quote:
as i was the one who wanted to test mods hardly and safe i wouldn't like the idea of mods gettin reviewed and re-reviewed again.

You're supposed to give a reason for disliking that. It doesn't help anything if you just say you don't like it.

Quote:
the first review should take place on the same day, of course

No, it should be given whenever a mod reviewer gets to it. If the queue is pretty empty, it may get reviewed on the first day. If it's pretty full, it may be a bit of time.

Quote:
and feedback will be given, so if the mod writer did some changes and the first reviewer is ok with it, then the second AND third reviewer get involved and test the mod too, after all three rated and agree it gets listed in the approved database - or not Grin     -     ? ? ?

This doesn't make any sense. The point of having three reviewers is so that the initial evaluation is accurate and then the mod writer will know they have changes to make. Having one person rate it, send it back for fixing, then have it resubmitted for reviews #2 and 3 is inefficient, insanely hard to code, and totally pointless.
  

~ Michael ~
-------------
The MikeCam
A truly wise man never plays leapfrog with a unicorn.
Back to top
IP Logged
 
] hirnsub [
New Member
*
Offline


If something can go wrong
- it will!

Posts: 38
Joined: Feb 6th, 2002
Re: Preventing Biases
Reply #23 - Feb 19th, 2002 at 1:32am
 
first if ya read my post ya should have read why i don't want a mod to be reviewed and re-reviewed again, cos it's the point of not gettin approvement to hard and long awaitin. (something you always suggested) and if there is a queue for mods- okay- what i wanted to say was that i see the need of gettin mods tested in a short time - !

and i do think i does make sense that the first reviewer gives feedback to modauthors quickly and if there are big bugs there is no need for the other two reviewers to test the same bugs, but if the code was corrected the first reviewer is okay with it, then there is sence in the second and third testers work after that, so when all three are okay with the mods it gets approved or rejected...!
there should be nothing hard to code for this point, just add a field where the first reviewer can set the status to "in work", so other testers see someone is already testing it, and after he finnished - status can be unlogged so others see the need of reviewing and testing, three tests finnished ---> decission
  

-=oX] hirnsub [xo=-
If something CAN go wrong - IT WILL !
Back to top
ICQ  
IP Logged
 
KyleYankan
Senior Member
****
Offline


The KyleYCam!

Posts: 341
Joined: Dec 25th, 2001
Gender: Male
Re: Preventing Biases
Reply #24 - Feb 19th, 2002 at 1:58am
 
Time for a fairytale, once there was a rabit, who was fast and speedy....



ah screw it, you know the tale. Why not take our time? A day or two isn't going to kill anyone unless your name is HAL. But still he can live though it
  


//-KyleY-\\

Steal his bandwidth! He doesn't want it!
Back to top
IP Logged
 
Administrator
Forum Administrator
*****
Offline


Yummm

Posts: 7
Location: Modders Rile
Joined: Oct 7th, 2014
Gender: Male
Re: Preventing Biases
Reply #25 - Feb 19th, 2002 at 6:09pm
 
Well there is no reason to wait with mod reviewing. However I suggest not to put the testers under pressure. You have to consider that we all have a real life too. There might be days when most or even all of the testers don't have time to review. When that is the case then there is no problem with waiting with a review one or two more days. I suggest not to force people to test mods within one day. I think we will have anougth testers anyway to keep testing time short.

The main reason for let more than one person test the mod is to prevent unfair ratings! In addition, two testers might find more bugs than one.
  

The Administrator.
Back to top
WWW  
IP Logged
 
Michael
God Member
*****
Offline


Recursion \Re*cur"sion\,
n. - See recursion.

Posts: 1003
Joined: Oct 23rd, 2001
Gender: Male
Re: Preventing Biases
Reply #26 - Feb 19th, 2002 at 6:22pm
 
Quote:
first if ya read my post ya should have read why i don't want a mod to be reviewed and re-reviewed again, cos it's the point of not gettin approvement to hard and long awaitin.

Now, if you'd actually SAID that last bit in your post, maybe I might've had the opportunity to read it...

Quote:
and i do think i does make sense that the first reviewer gives feedback to modauthors quickly and if there are big bugs there is no need for the other two reviewers to test the same bugs

Well, you've got a point about testing the same bugs over, but the whole idea of more than one tester is the repetitiveness! You're missing the point: we're trying to make it so that one unfair, biased person can't just go and reject someone's mod. Having two or three people review the mod makes sure that a mod doesn't get rejected (or accepted) without reason. As XXL said:
Quote:
The main reason for let more than one person test the mod is to prevent unfair ratings! In addition, two testers might find more bugs than one.


Quote:
but if the code was corrected the first reviewer is okay with it, then there is sence in the second and third testers work after that

And if the first tester made a mistake and/or is stupid and a bad mod gets accepted? No thanks, I think I'd like double or triple assurance.

Quote:
there should be nothing hard to code for this point, just add a field where the first reviewer can set the status to "in work", so other testers see someone is already testing it

Two or three people should be able to review the mod at the same time, to make the process as quick as possible. There's no reason to wait until one person is done to let the other person have a crack at it.

Quote:
Well there is no reason to wait with mod reviewing. However I suggest not to put the testers under pressure.

Exactly, that's why I planned on having a queue. A mod would get assigned to a tester for review only whenever the tester logged in to get a mod for review. Otherwise, if mods were assigned automatically, testers would feel pressured, reviewers would become inactive and leave mods hanging, and things would be a total mess.

Quote:
You have to consider that we all have a real life too.

No! Shocked Ahhhhh! Tongue
  

~ Michael ~
-------------
The MikeCam
A truly wise man never plays leapfrog with a unicorn.
Back to top
IP Logged
 
Shoeb Omar
God Member
*****
Offline


Mod Writer

Posts: 5665
Location: San Diego
Joined: Jun 29th, 2001
Gender: Male
Re: Preventing Biases
Reply #27 - Feb 19th, 2002 at 6:29pm
 
*darns the quote function*
  

YaBB SP2 BETA!
Now taking pay jobs in PHP or Perl.  Contact me for details.
Back to top
IP Logged
 
Michael
God Member
*****
Offline


Recursion \Re*cur"sion\,
n. - See recursion.

Posts: 1003
Joined: Oct 23rd, 2001
Gender: Male
Re: Preventing Biases
Reply #28 - Feb 21st, 2002 at 2:36am
 
Okay, I think the middle ground that most people can agree on is to have three people test each mod before approval.

If you don't agree, say something now or forever hold your peace, because I'm locking this thread sometime tomorrow and moving ahead.
  

~ Michael ~
-------------
The MikeCam
A truly wise man never plays leapfrog with a unicorn.
Back to top
IP Logged
 
Michael
God Member
*****
Offline


Recursion \Re*cur"sion\,
n. - See recursion.

Posts: 1003
Joined: Oct 23rd, 2001
Gender: Male
Re: Preventing Biases
Reply #29 - Mar 17th, 2002 at 5:09pm
 
Okay, I said I'd lock this 3 weeks ago. Tongue Consider this issue closed, with this one addition: Must add a check to the code so that a reviewer cannot review a mod that they themselves have submitted.
  

~ Michael ~
-------------
The MikeCam
A truly wise man never plays leapfrog with a unicorn.
Back to top
IP Logged
 
Page Index Toggle Pages: 1 [2] 
Send TopicPrint