Jeremy Miller tweets (Senior Programmer at ISI)

You edit your post , So when you plan to introduce a moddification that can impact a product in use, you inform everybody in advance before launching patching ...
This to at least avoid complains.

 

Yes with this I could agree

 

Were they released separetedly? As far as i remember this was a full package. So MM update was part of build update. Or im i loosing something? Is MM something from outer space that no one tested it?

Couldnt be fully tested? Whys that? Why one of (not to say the most) important thing was not tested? Time? We waited 2 months, probably no one would kill themselves for waiting a couple more hours/days.

MadKing is fully right on what he said. Even ISI would give a much better impression if all this "stupid" problem didnt reach public.

We all here who read ISI fórum understand the problem, imagine the dozens who simply dont even know theres a fórum....today they are screaming at the wind that RF2 is the worst of all.

Time to sit back, look at this project (if i can call it that), plan some objectives and here we go.

Even for a RF2 enhtusiast it keeps gettting harder and harder to understand its development, its interaction with RF2 users/buyers.

Was changing car in practice a so demanded thing among rf2 users? I simply dont understand. Im ok with most of the build content even if it lacks much things i was expecting, but some others i simpy dont understand.
Sometimes posts look a little harsh but believe me...i strongly try to be polite and objective.

 

The new build crashes due to a specific combination of server names. The build was tested a week internally and then the unstable for a day and then even the stable for a half a day before someone had a server with silly characters in its name that caused it to crash. Up until then the build never crashed on the MM list. The randomness is a little odd though. When I spoofed my own MM I could pass it a static list of unfiltered servers and it would crash 99.9% of the time but every now and then it would work. When I filtered all the crap server names from the list it would work 100% of the time.

Anyways, this build could have been tested for months and the MM may have never crashed. Sh$t happens.

 

In the end thats what 90% of rf2 users really care about. They dont care if its a letter or if its a car or a track.
The other 10% are here trying to figure out this issues, like me and you.

Anyway, this is like banging the head on the wall....no point in keeping discussing how, when and what. Lets hope MM stays stable in this stable.....

 

Right the fault is in the build. But all the arguing about releasing a build that very clearly has a problem is nonsense. It didn't very clearly have a problem until the stars aligned.

 

The stability testing should be done massively in the same way as when a stable build is released. For that it needs to announced publicly. There is nothing wrong to release something alerting that it needs to be tested. If the build is considered stable there would not need to change anything. It would be enough to say so after the week or the time considered sufficient to consider a build stable. Not even necessary (I don't mean it would not be useful) to change the number. Just a confirmation from ISI about its stability confirming the build for the next months before a new "ready for stability testing" build is released.

With respect to leagues providing feed centralized on one person I think is a good idea and already proposed it to tim via email and he liked the idea. We are happy to collaborate with ISI in many fronts.
I know I appear sometimes to be against everything but it is mainly because the way I write. I should use more emoticons.
enviado mediante tapatalk

 

Can you give it a rest? You're turning into a right know-it-all bore.

 

I know. If a problem server or two were up, you'd not have had a time where it worked well, if at all. Earlier in the day they weren't up, and obviously this led to it being hard to detect.

 

Good post as always.

I wish people would understand this first, before posting things like "ISI should do this, ISI should do that". It's awkward to propose solutions without understanding the cause in first place.

 

Sure with this kind of information from Noel, it's better to understand, So ISI should change its way of communication and give more information on the plans on going.
Tim said the game is considered as out of beta since June 2013, and 6 mounth later, it's impossible to use the game in league context, so now we need clear plan to calm down the guys wanting to use the game every day in online !!! Patience has limits, and now the limit is close to be over passed, so close collaboration and clear communication is needed.

 

Sorry but that's a 100% false statement.

 

Tire model not finished
Real road behaviour not finished (too accelerated) and not realistic under drying phase
Damage model not finished (no deformation)
Limited to less than 30 drivers due to 32 bits limitation

I stop the list here, I talk about serious championships, not just fun open.

 

I can't argue with your expectations as they are yours, but your statement is still 100% false.



When ISI says something is finished and walks away, then I'll get worried.

 

So what you're saying is rF1 wasn't good for league racing either? Strange, I raced in rF1 for 9 years.

 

rF1 had none of those things. GPL was limited to 19 (20 if the host raced). Were they suitable for championships? Because I raced in a GPL league, and I am racing right now in an rF2 league.

I understand you have an expectation of your own requirements to use rF2 in a league, but the word impossible doesn't fit that.

 

I will give it a rest when I feel satisfied. You probably didnt read carefully what I wrote which took me some time in the effort to explain why I think that what happened was predictable. I think you just wrote an easy post criticizing me but with no added value in your message. I am claiming that the unstable releasing is useless and proposing something I thonk it woild work better. It is not about knowing all. It is trying to improve something that simply is not operative.
It is the second time a major bug occurs and quite similar. It is the second time it is said that it could not have been detected but was actually detected within hours by the community. I prefer appearing to be a smart ass pretending to know everything but trying to help than appearing to not understand anything and without any will to improve anything.

enviado mediante tapatalk

 

As satisfaction is a deeply personal thing that it's impossible to guarantee, this sentence makes me concerned about you being allowed to continue posting here. I accept your points, but as you are repetitively stating them, becoming an annoyance to others, and seemingly reveling in "appearing to be a smart ass", it becomes non-constructive when you post, which is the opposite of what you originally intended, and stifles other opinion or debate.

 

And did you take time and effort to understand the cause of matchmaker problem, before you suggested changes in testing procedure?

Allow me to explain, what I'm referring to here.

This particular issue would not have been detected if release procedure was different. It only manifested itself, when certain racing leagues (with "something weird" in their mods) updated their servers to new build. Since leagues only update to stable builds, not unstable ones, this went undetected until stable build release.

To sum it up:
Is there an error in ISI's software that went undetected? Yes.
Could it have been detected if testing / release schedule was different? No.
Did ISI react immediately to this issue? Yes.

Please don't get me wrong here. I'm not saying that trying to help by offering advice is a bad thing. But in this particular case I believe you were not qualified to offer useful advice. You posted a proposal, without understanding what actually went wrong with the software.

This is why it's often best to leave such decisions to development team, who have better insight as to what's going on. The upside is that ISI reacted immediately and Jeremy spend a lot of time trying to get it fixed for us as soon as possible.