Ok Right, Wanted to let the dust settle a bit before I made any comments. I worked in I.T. twenty years for a 'Major" international computer software/hardware/support corporation (Unisys). Started in the computer room, then computer programmer, then programmer analyst, then systems analyst and finished as a project manager. If any of my projects had ever had the roll-out results of build 494.... I would have found myself unemployed, fired for cause. 1) Poor planning. 2) Poor design. 3) Poor testing procedures. 4) Poor implementation. 5) Poor notification. 6) Poor reaction policy. 7) Poor escalation policy. 8) Poor customer service. Pick any one of the above and I would have had to "hit the bricks". ISI seems to have hit the octo-fectra with build 494. Sorry to be so harsh but I tend to be very frank about my observations. p.s. Oh Yeah, don't forget to mention that it gets released right before the weekend and then everybody get the weekend off ???
LOL, Just wanted peeps to know I wasn't some pimple face who didn't know what they were taking about. Funny that is more important than the points of issue!!!
Well, if your points had not been mentioned a million times before they would have had some weight. And in case you are not aware, elite credentials on the interweb are usually just made up and/or fake. Not saying that's the case with you, but they don't do much for the majority of the public as we have no idea if your are truthful or just trolling. And on top of that, I doubt ISI would be impressed with the list you made, pretty sure they know they messed up and will be fixing stuff full force tomorrow.
I remember another post by you saying you are in the IT for 15 years. Now they're 20 after few weeks. Isn't the approximation another enemy of programmers?. J/k.... This is by you. Pity you changed your mind for whatever reason: BTW thanks for sharing your impressions with us but just to give you a late sunday support message, I can say we're talking about something it's already under fixing and testing so, don't worry to much. Cheers and thanks again.
I think it's quite understandable that every now and then a new build brings issues, there is always something that will be missed in testing. We see it all the time with modern games that are on an aggressive update schedule, for example AC usually releases a hotfix right after a new build. What would be important is to minimize the downtime as much as possible, as for example in my league the server has been almost empty the last few days because people can't join or given up trying. Maybe it would be a better idea to release new builds at the beginning of week, presuming that nothing gets released in the weekend.
I have read to much in here but cant find the solution of the servers issues. Cant understand how you, ISI, dont solve them inmediatly. I know you can be wrong in some things but this is the worst one. If we cant enjoy the servers, what can we do, go back to 382? Please, work harder in this.
Got to ask the bleeding obvious, how could testing have found this issue? If no-one else is running the same build online you can't join or get mod anyway.
The internal testing team has servers up for each build so we do run tests. But I presume this is a load-related matchmaker issue, which his why it didn't appear in testing with only a couple of clients.
You are not going to find it, they are working at it, the issue is known to them, they are searching for a fix.
Exactly. And unless it takes a very small number of clients to trigger the problem the testers wouldn't have been able to put enough servers up to find the issue existed. Had to go public for that..or is my old brain missing something?
Did YOU read it? Some ppl use it for more than lazy entertainment and run groups of ppl. Of course you dont care. Stop telling people they shouldn't be putting their feelings about the new build up here, fanboys!! Edit: Can ISI please implement an event donation/entry-fee refund function (with messaging) into the next release, so that any organisers can simply and easily abandon events based on your software? Cheers.
I don't think it has nothing to do with MM load. MM is shared between all builds so the load has been always similar. I have not perceived any type of difference in behaviour since yesterday morning. When you use B382, MM is working fine. You can even get mod from servers running B494, which you need to "refresh" when doing the same having started online withn B494. How is it possible that for some, the only possible explanations to problems can only be related to star alignment or other kind of unpredictable phenomena? What are you basing one to affirm what you say? Until some years ago internet updates did not exist and even gaming software was distributed in cassete, diskette or CD. At those times things had to work perfectly after testing phases when the product was released or the prestige of a company would considerably suffer. Isn't it just possible that something is being done wrong, considering that there were critics to the stability testing method even before the unstable was released? It is not acceptable to use general public for stability testing in a non beta software as this without alerting them. I guess many will own a nvidia card and they are probably aware about the possibility to get alerts when updates are released. You can include or exclude beta versions from those alerts. You can also see the amount of time that usually goes from the beta version to the stable version. View attachment 11827
Any chance I can get a HINT of when this JOIN fix MIGHT be implemented? Only because of the Cup this weekend ...it is 5 days away..??
The difference with the old and new build is that in the old, everything is queried at once from MM when you open multiplayer, or press refresh. In the new build, it starts a separate query each time you click on a server, as to my understanding. That's why the join button remain grey for a while. Also there are some new features, for example spamfilter and new information on the multiplayer main screen (forced cockpit view etc). I suspect all these things increase the load on MM. But who knows, if build 382 can still query fine maybe it's also a build issue. But if it was purely a build issue it should have been detected in the internal testing, which we didn't, and we do test joining servers.