I don’t always comment but when I do its to say well done CCP 
But seriously the chat wasn’t working right on the sisi server over the last few weeks myself and a buddie presuming you where testing it out great makes seance , but then to see you implement something that clearly was broken as hell baffles me
Well done ccp , hopefully you fix it soon
You must be new to Eve. That, well lets be charitable shall we, that used to be the norm when it came to big patches.
Stuff broken?
Yup
Ship it anyway its still awesome!
Carebears still crying that they cant afk rat or super rat. Grow up, you a Wormholer now!
But in all honesty local seems to stabilize in systems with low population count, that being said they could all be cloaky camping too.
WH Space doesn’t have reliable gates that don’t need scouting to get into your space by?
Tinfoil time, this is a test to see how people cope with what is basically no local, before CCP remove it 
Yea this kinda feels like a test to see how viable it is, seems fine tbh lols
Re-Patch the old chat system into a patch and forget playing around in the main server till the sisi can run it correctly lost ships to reds and dec’s like madness across the cluster due to CCP jumping the gun. Get us back to normal and play with chat over the next year to fix it please… Game is losing numbers faster then ever when it’s not stable. Grrr CCP lol revert revert lol…
Sounds exciting. I guess it is for technical reasons that the old chat system was replaced with some undeveloped student project while the new chat system is being tested? Could you please let us know if the new chat system will be finalized this month?
Where is the counter of the listed avatars from the upper right corner of the chat window?
It was the most useful thing in “Local” - who cares about names, numbers are important!
In 90% of cases, the “Local” window is overlapped by others, leaving visible only the column of the list of avatars with a total counter.
I am loving all of this salt of lack of local. Not like most of you really use local other than be alerted that someone new is in system.
Now you have to gasp, use your D-scan and pay attention!!!
Please could we have delayed member lists back on chat channels, where names don’t show up until someone speaks?
Aside from helping with name list clutter (busy channel name lists are almost useless without this), it was quite an important bit of gameplay that you could lurk in some channels without people knowing you were there.
(By the way, deleting Local is almost, but not quite, as bold as deleting hisec. Congratulations are in order? 
)
If EVE becomes Wormholes Online you’ll be playing alone. CCP follows the money, as does every business, successful or not. Wormholes Online would be the final update before EVE shut down for good.
Do i need to resize the chat window everytime i login? Also i have more lag now.
As much as I used to be on board with the idea…
I think the bigger problem is the inconsistency. Some people say local is working, others say nay. Based on the testimony in the plethora of threads on this, each client seems to get different results. If everyone lacked local all the time, that’d be one thing. Still a bad thing for most, but you could at least call that even. But with some people having local and some not, that’s creating a huge difference in the amount of intel each side is potentially getting.
I don’t wish to dogpile on here. It seems CCP had a generally good idea in updating chat, but obviously they didn’t work out the problems on SISI first. I would have hoped they would do another mass test or two there before launching on TQ. If the people who did test on SISI are correct in that it wasn’t functional there either, then I really fail to understand why CCP launched it onto TQ in such a state. Usually, such haste is warranted when a current system is failing and time is of the essence. I have heard of no such justification here.
The issue is TQ is a large (very large) multi-clustered server that they recently moved to AWS while Sisi is a single server that they run internally. So it’s literally impossible for them to have an exact parity test between the two servers because they are implemented differently. The new update was obviously not good but when dealing with that amount of scale and that many application layers these things can sometimes happen.
Okay. I hear you. But reading that makes it sound so, so much worse. You are, in essence, saying that a central part of EVE could not be QA tested or debugged prior to release, and that the devs simply threw it in willy-nilly and hoped that an untested and untestable overhaul of a core mechanic would just come out for the best with no backup plan in case something very likely went wrong.
You may be entirely right but that makes them sound so much more irresponsible when you point that out.
Judge, I don’t deny my client crashed into several parked cars that night. But revoking her driver’s license over this once incident feels extreme. She’s normally a very good driver, with a near-perfect driving record. Her driving that night was just the result of the excessive drinking she was doing.
Could not be fully tested. Scale is hard to test.
An educated guess at to what happened:
They tested, with a server on their local network, with 50-100 people or so.
With it being local, the turnaround on the updates for local ran a lot faster, so didn’t overlap as much, leading to a far reduced load on the chat server, so they didn’t see any problems.
What you’ll commonly see in this kind of situation is there’s a tipping point, where just one additional person, being a little slower than the rest, leads to everyone being slower, which acts as a multiplier on the load, making things far worse than you’d think, from your projections. Instead of things slowly rising, in rough synchronicity with the number of people involved, you hit a point where it goes exponential instead.
It’s not irresponsible it’s only what they can do practically. It would literally cost them millions and million of dollars to replicate TQ. They are confined to economical and real pressures and scalability just like anything else. When things went wrong, they were right on it working day and night to fix it. If they hadn’t done that then I would agree with you but their tireless effort to correct it shows me how committed they are to making it right.
Edit: Also to note, it is also impossible for them to test the interaction on TQ due to the amount of players that login even if they could build a replicated version of it. So they test the best they can with things like Sisi. (Which Sisi is not the FIRST test. They also have internal DEV servers such as DUALITY which are off limits to us.)
Knowing eJabberD, this shouldn’t be a problem for it to begin with. More likely they did something that weighted on the bridge between EVE and chat server.
You know Eve is multithreaded, right? It’s only the simulation which isn’t?
