please fix the chat bug


3dent

 

Posted

please fix the chat bug


 

Posted

You know, I bet until you started this thread that hadn't even occurred to them.


 

Posted

The bug is being worked on, and a fix will be released as soon as one is available and we're confident that it resolves the problem without creating any new bugs.

Please be aware that in all cases of a bug being present, there are several stages involved in fixing the bug starting with identifying the bug, being able to reliably reproduce the bug on demand and therefore gaining an understanding of possible causes, through isolating the specific code causing the issue, verifying that any change to that code will fix the bug without creating a new bug, testing the fix internally and on the training room ... and more.

Some bugs are easily identified, reliably reproduced on demand and therefore quick and easy to fix. Others aren't, for a variety of possible reasons, and that can elongate how long it takes for the cause of the bug to be isolated and fixed in the code.


 

Posted

GhostRaptor: thank you for the answer


 

Posted

Also there are some work arounds for that bug.

1- If you recieve a broken tell, you can reply with typing /r <your message>

2- Using global handle seems to overcome that bug as far as i know. Never had a trouble when using global handle. (Thanks to Bjorkiz for telling me that )


 

Posted

[/color]<blockquote><font class="small">Antwort auf:[/color]<hr />
The bug is being worked on, and a fix will be released as soon as one is available and we're confident that it resolves the problem without creating any new bugs.

[/ QUOTE ]
Isn't it quite significant that it gets mentioned explicitly that dev is quite confident not to create new bugs while fixing the original bug?
I was almost looking for a "..this time" in the sentence...

[/color]<blockquote><font class="small">Antwort auf:[/color]<hr />

... being able to reliably reproduce the bug on demand ...

[/ QUOTE ]
...like playing the game for about an hour to experience the problem several times?

Sorry for being a bit sarcastic here, but it's very hard to understand that every time a bug is fixed, a new one is introduced. Even more that the newly introduced bugs don't seem to be even remotely connected to the fixed issue.

It's even more frustrating that NC apparently put a lot of effort in aligning this and that powerset, changing this radius, changing that recharge rate instead of sitting down and dedicate the time on fixing a bunch of existing bugs first.

So. I feel better now.


 

Posted

[ QUOTE ]
Also there are some work arounds for that bug.

1- If you recieve a broken tell, you can reply with typing /r &lt;your message&gt;

[/ QUOTE ]

Actually, that has never worked for me. I've never been able to reply to broken tells.


 

Posted

did you try with 'backspace' or with /r?

Backspace never worked for me either, only /r does.


 

Posted

[ QUOTE ]
[ QUOTE ]

... being able to reliably reproduce the bug on demand ...

[/ QUOTE ]
...like playing the game for about an hour to experience the problem several times?

[/ QUOTE ]

The point you miss here is that what you experience is not what a QA team refers to as a reliable method of reproduction. Can you tell me, with confidence, that you know what specific set of circumstances causes the bug to manifest?

While the bug will generally manifest on multiple occasions during a prolonged play session, that alone is not useful to the QA team or the people who are assigned to fix the bug.

What we mean by reliable, on demand, reproduction of a bug is that a member of the QA team can sit down, do X, and see the bug immediately. And then be able to do it again. And again. And again. Immediately. Having it crop up at random over multiple hours of play is not a reliable, on demand, method of reproduction and doesn't help in terms of identifying the bug.

It is far, far more difficult (and actually quite unlikely) for a bug to be fixed without the team knowing under exactly what circumstances that bug occurs. Think of it as being like trying to find a needle (the buggy code) in a haystack (the entire code). If you don't know what specifically causes the bug, then it's like hunting for that needle using nothing but your hands and eyes. If you do know exactly what causes the bug, though, it's like having a metal detector: suddenly you know where to look in the code.

[ QUOTE ]
It's even more frustrating that NC apparently put a lot of effort in aligning this and that powerset, changing this radius, changing that recharge rate instead of sitting down and dedicate the time on fixing a bunch of existing bugs first.

[/ QUOTE ]

The people who work with powers, powersets etc are not the same people who work with chat code. Telling Castle and his team to stop working on power balance and instead help the chat programmer fix a bug won't get that chat bug fixed any sooner, however will certainly cause delays in work involving powers.


 

Posted

[ QUOTE ]
Also there are some work arounds for that bug.

1- If you recieve a broken tell, you can reply with typing /r &lt;your message&gt;

2- Using global handle seems to overcome that bug as far as i know. Never had a trouble when using global handle. (Thanks to Bjorkiz for telling me that )

[/ QUOTE ]

I think that must be right as we have never had any trouble chatting across servers


Happy to be on Defiant.
Global name @mereman
Member of P.E.R.C. Representing Defiant
Alts http://cit.cohtitan.com/profile/4488
CoH faces http://faces.cohtitan.com/profile/mereman

 

Posted

[ QUOTE ]
[ QUOTE ]
Also there are some work arounds for that bug.

1- If you recieve a broken tell, you can reply with typing /r &lt;your message&gt;

[/ QUOTE ]

Actually, that has never worked for me. I've never been able to reply to broken tells.

[/ QUOTE ]

Using /r has worked every time for me since this bug appeared. It works in the same was as the backspace though in that it will only respond to the last /tell you recieved.

If you got a bugged /tell then another complete /tell using /r will respond to your most revently recieved complete message, not the broken one.


Tidgy

---------------

 

Posted

Sometimes the /r hasnt worked but when i've done /r, then it seems to work, its only afew times I have had to change the way of doing it.


 

Posted

When it happens to me, I send /r sorry that was a broken /t
the reply to that usually gets through without a hitch.


 

Posted

i blame long character names and trying to talk to people out of your zone being the cause of the issue here (and not using the @GLOBAL name)

Short Character names = no issue in any sense.
Long Character names = issues when talking cross zone.

Who would have thought that having a LONG character name was the issue....

Now, if THAT is the problem, I *personally* would have never thought about it being the cause (then again I would have never thought that the language in a windows install would stop a program from installing, but it has done, ie using windows in Turkish prevented the beta client of Tabula Rasa prevented it from installing; dont ask me *why* it did... it just did. I know this because i was keeping an eye on the thread at the time)

So as GR has said... an intermittant issue (or even a one off issue) that only affects a *certain* portion of the player base with no immediately obvious common links does indeed make an issue harder to pin down to the wall.

For me, i doubt if *any* player has recieved a broken tell from me... if they have, then i am sorry (then again i normally try to keep my character names short and sweet)

Of course, i could be wildly wrong over the cause for the issue; so everything else is pure conjecture.


 

Posted

[/color]<blockquote><font class="small">Antwort auf:[/color]<hr />
Think of it as being like trying to find a needle (the buggy code) in a haystack (the entire code).

[/ QUOTE ]

I hope you didn't spread your chat and chat coloring routines (a pitchfork full of hay) throughout your entire code (the haystack)
Under these circumstances, I agree, it will be very difficult to find the bug (the needle).


Sorry, but I couldn't resist that one


 

Posted

[ QUOTE ]
[ QUOTE ]
Think of it as being like trying to find a needle (the buggy code) in a haystack (the entire code).

[/ QUOTE ]

I hope you didn't spread your chat and chat coloring routines (a pitchfork full of hay) throughout your entire code (the haystack)
Under these circumstances, I agree, it will be very difficult to find the bug (the needle).


Sorry, but I couldn't resist that one

[/ QUOTE ]
You may write sooper-modular code, so good that GoF asks you for design tips, but bugs whose causes and reproduction methods are more like conspiracy theories would STILL appear from time to time. (Like at least three documented bugs that were, in fact, caused by the Moon being in certain phase. One of them was in an e-mail client. I'm not making this up. Or, even better, infamous Bush bug in Notepad. (Interesting enough, it doesn't work now with the original string, but some others still produce the bug.))


 

Posted

Or like the "Database is slow when it rains on Monday morning" oracle problem ;-)
Yep. these problems do exist, no question about that.


 

Posted

Some bugs are just plain stubborn. Many years ago I played a game called Darklands. It was kind of like Oblivion but more awesome and with 256 colours!

The Microprose dev's were very keen to fix bugs in Darklands but they were defeated by the "green hair bug". Every now and then, a character's hair would turn green. Re-loading the game wouldn't fix it. There was no work around. Worse than that, nobody could reproduce it. The months went by and the dev's got increasingly desperate. They would not be defeated by this bug.

In the end, they actually offered money to anyone who could reproduce it.

Nobody claimed the money. They never fixed the bug.


 

Posted

/r doesn't work if somebody sends you a tell between your reciept of the broken tell and your reply.

Anecdotally - this seems to be actually getting worse, not better. Is there any news?



"You got to dig it to dig it, you dig?"
Thelonious Monk

 

Posted

[ QUOTE ]
/r doesn't work if somebody sends you a tell between your reciept of the broken tell and your reply.

Anecdotally - this seems to be actually getting worse, not better. Is there any news?

[/ QUOTE ]

Yes, there's news. Try looking in the training room forum.


@FloatingFatMan

Do not go gentle into that good night.
Rage, rage against the dying of the light.

 

Posted

[ QUOTE ]
Is there any news?

[/ QUOTE ]Yes, in the Test Server section. It's fixed.


 

Posted

please fix the chat bug


 

Posted

The bug referred to when this thread was created is already fixed. Please do not post new bugs in a pre-existing thread.

Further, when posting about bugs please a) search the forums for any existing reports that match yours, and b) provide a description of the bug. "Please fix the chat bug" is not a good bug report.

For reference, there's a report on the I14 chat bug here (although, unfortunately, it's still lacking full details for the devs - I've seen better reports elsewhere on the forums, however).