View Issue Details

Jump to NotesJump to History
IDProjectCategoryView StatusDate SubmittedLast Update
0001890HTML & PERLBug Reportpublic2010-11-25 19:282014-05-09 04:00
ReporterHinoe Assigned ToDerIdiot  
PrioritynormalSeverityminorReproducibilityhave not tried
Status resolvedResolutionfixed 
Summary0001890: Users commenting on others' creqs will change the creq state every time
DescriptionCreq id 3021293 was on HOLD (mod replied, awaiting for user). Upon me commenting on it, it changed to RETURN (user replied, awaiting for mod). I'm not the one who issued the creq, so I though it was very weird. I then thought it might be because I am flagged as user and the system didn't bother to check whether I was the creq issuer simply because users didn't comment on others' creqs before.

So it happens that I expressed these thoughts on another comment, and were I surprised to find out it went back to HOLD. Supposedly, only a mod's comment can change the state to HOLD. I sure as hell am not a mod, so that struck me as totally wrong. As I caught myself staring at the screen, I understood what the system was doing, and thought "evidently, that can't be right".

Being completely sure the next comment would send the creq back to RETURN, and telling myself "this is a very funny way to behave", I complained about the weirdness of it while expressing that I was going to talk to the bug tracker. Just according to keikaku (TM), the creq changed state again as I had thought.
Steps To ReproducePick up a regular user account, talk on other people's creqs. I'm not sure what states this "works" with, but it apparently works with HOLD and RETURN.
Additional InformationI think the system doesn't like having to swallow the comment of someone who is neither the creq issuer nor a mod. I can even suggest a possible way for the system to check whether it should change the state:

Current state: HOLD.
State tells us that the latest comment is a mod's. Someone comments.
Another mod? Yes -> Keep current state. No -> It can only be the issuer. Change state to RETURN.

Current state: RETURN.
State tells us that the latest comment is the issuer's. Someone comments.
The issuer again? Yes -> Keep current state. No -> It can only be a mod. Change state to HOLD.

If I'm right, the most obvious fix is:

Current state: HOLD.
State tells us that the latest comment is a mod's. Someone comments.
The issuer? Yes -> Change state to RETURN. No -> A mod or another user. HOLD is the proper state. Keep current state.

Current state: RETURN.
State tells us that the latest comment is the issuer's. Someone comments.
A mod? Yes -> Change state to HOLD. No -> The issuer or another user. RETURN is the proper state. Keep current state.
TagsNo tags attached.

Activities

Hinoe

2010-11-25 19:28

reporter  

state.png (18,715 bytes)   
state.png (18,715 bytes)   

Hinoe

2011-03-01 01:15

reporter   ~0003211

Mmmm, it just occurred to me there can be a 3rd state, for when someone who is neither the issuer nor a mod leaves a comment; that would end the state madness in the cleanest of ways.

Hinoe

2014-04-13 17:25

reporter   ~0003342

This is probably related to the fact PMs with the wrong content get sent when users interfere in others' creqs. (i.e. the "a mod replied blah blah" PM is sent instead of the "someone replied blah blah" one)

That's rather problematic because RANDOM_USER comments and CREQ_ISSUER acts on whatever RANDOM_USER says believing RANDOM_USER is a mod. That happens all the time.

Issue History

Date Modified Username Field Change
2010-11-25 19:28 Hinoe New Issue
2010-11-25 19:28 Hinoe File Added: state.png
2011-03-01 01:15 Hinoe Note Added: 0003211
2014-04-13 17:25 Hinoe Note Added: 0003342
2014-04-20 18:37 DerIdiot Status new => resolved
2014-04-20 18:37 DerIdiot Resolution open => fixed
2014-04-20 18:37 DerIdiot Assigned To => DerIdiot