Quote from Scawen :

I'm pleased that you're pleased, but the best thing you can do is test before you make any comment!

Lol, sorry. I was so surprised by new patch name

But now I have check every updates.

Good job with objects, especially with armco5, it can replace barrier long (armco5=2barrier long) so it can reduce number of objects on a layout and put more objetcs but armco's edge are too dangerous when you hit it, there will be a kubica bis on lfs

Good work on tyres temp. for hotlapping. I have tried it and it run perfectly. I think it would be awesome to had this on all game mode (solo, multiplayer... etc) like if we use warmers blankets like in Formula 1 actually.

Good to see qualify now on open tracks, I hope we can use it at TEE Z30 festival (or should Yann rename it 0.6A1 festival for last round? )

Quote from Scawen :

Is there any reason why Aririo closes if it gets an unknown packet? I think all InSim programs should simply ignore unknown packets, if their size is valid, just skip it and carry on.

Well, yeah, a bit short-sighted decision. But in the past 2.5 years all unknown InSim packet types, as well as incorrect packet sizes meant just one thing: Broken (usually remote) communication. When one bad packet comes, many others always follow, and even those that look OK are in fact wrong, so disconnecting on each bad packet and auto-reconnecting the InSim app a minute later was the best and safest I could do. I think even for the future, just skipping bad packets will not work, because they mean some serious connection troubles and unreliable data. Well, that's just my experience...

PS: Also, bad packet received means some data, possibly important ones, were sent but not received and processed which can lead to all kinds of troubles. Again, the safest route is to close it all and restart the connection, getting all the data again. It works pretty well, when the InSim connection is local (server and Airio on one PC), this almost never happens, we've been running several full demo servers for days and weeks without a glitch. But it is kind of a safety measure to stop faulty communication.

Quote from EQ Worry :

...

It's not like I wanna tell you how to do your job, but shouldn't the TCP/IP (and lower network layers for that matter) handle hte error checking for you? In theory you shouldn't get a single bad packet because of the cheksumming done by OS or hardware. I know this might not work in all cases, but perhaps a bit less strict error handling policy will work fine here. I guess that AIRIO checks for the InSim version it's connected to, so when it detects an incompatible version, it could simply warn the admin and silently skip all unknown packets...

[QUOTE=Scawen;1603133]I see, you mean the collisions are not good when you hit the end of an object. No, that will not be changed in this version.

But in next version's it will be changed?

Super!

Quote from EQ Worry :

I think even for the future, just skipping bad packets will not work, because they mean some serious connection troubles and unreliable data. Well, that's just my experience...

I think you should just skip packets with an unknown type, as long as their size is a non-zero multiple of 4.

If, however, you receive a packet with an invalid size (not a multiple of 4) or a known packet type but it's the wrong size or otherwise invalid, that would be the time to close the connection.

I hope Airio and other InSim programs do handle split packets. I bet there are some InSim programs that don't do this... sometimes TCP can split packets so you don't get a full packet in one receive call, and the remaining data comes in the next call. TCP is not guaranteed to keep the packets in the same chunks as they are sent - it only operates as a stream and guarantees to keep all the bytes of that stream in the correct order. But, because nearly the whole time packets are not split, some TCP receiving programs assume it will never happen... but it does!

In case any programmers are reading, who have not dealt with this, what you must do is not process a packet if the received buffer size is less than the packet size. That means you have a split packet. In that case, wait until some more data arrives, add it to the end of the buffer and check again if the whole packet has arrived. There's a good reason why "Size" is the first byte of the packet.

Quote from insaneltu :

But in next version's it will be changed?

I don't know, I don't have the future planned out in that kind of detail.

There are different ways of detecting collisions, avoiding intersections before the collision is detected. I'd like to look into that some time, but I don't know when, because it is a big project in itself. Not something to do in the last week before an official patch.

I'm quite happy with the recent collision detection improvements for now, so further development of that is way down the list, compared with finishing the new tyre physics.

The plan is to fix the remaining bugs and release the official patch within a week. Whether that will happen or not I obviously can't say 100% but that is definitely the plan.

Yesterday i made a demo server and everything was ok, but there were 2 guys with no nicknames. I made a screenshot (only 1 of the guys in the server)

Quote from Scawen :

I hope Airio and other InSim programs do handle split packets.

Uhmmmm... I'll take a look at the way this is done and will consult it with a professional programmer, who's also working on the library, thanks for info.

Thinking about it, I'm not sure making that admin command packet type optional is a solution either. After all there are the new car hit packets and maybe also other ones that are simply sent. It seems everyone wanting to use Airio with new patch will have to update... Maybe a switch to activate version 5 packets? Uhm, I'd want too much I guess...

Quote from wolfshark :

Yesterday i made a demo server and everything was ok, but there were 2 guys with no nicknames. I made a screenshot (only 1 of the guys in the server)

http://www.lfsforum.net/showthread.php?p=1603106#post1603106

Quote from wolfshark :

Yesterday i made a demo server and everything was ok, but there were 2 guys with no nicknames. I made a screenshot (only 1 of the guys in the server)

Probably empty character, as long as he have username everything is OK

Well done !

I've removed some posts asking about this. I would like to make it quite clear.

This test patch is a test for the official version.
That means the official version will be just like this one!
The only changes will be small, like bug fixes and slight improvements.

There will be no more big changes.
There will be no new tracks.
There will be no new cars.

I hope that we can release the new official version next week.
That really should be the case, unless some unforeseen big problems come up.

Will this make the new version compatible with lower version hosts?

Quote from Jonathon.provost :

Will this make the new version compatible with lower version hosts?

0.6 version will be only compatible with 0.6 servers.
I am not sure but next version should be compatible with A1 too.
Thats why clients changed from 0.5 to 0.6

Are you kidnapping me, one or more official patches before the old ass scirocco and rockingham? ****. Glad I sold my wheel. This will take years.

I'm curious how does tyre warmers cooperate with pressure settings.
If I set 1.5bar in Road Super and 100*C, the result is tyre pumped with 30* air (optimum_temp-30*) then heated up to 100* so real pressure is something >1.5bar or,
I've 1.5bar of 100*C air in tyre?

FUUUUUUUUUUUUUUUUU You released it on Le Mans day

Has anyone checked the qualifications of the open configurations, are no errors?

Quote from Tomba(FIN) :

Are you kidnapping me, one or more official patches before the old ass scirocco and rockingham? ****. Glad I sold my wheel. This will take years.

I promise not to kidnap you.

I wrote quite a lot about this, do follow the link below if you do actually want to try and understand.

But I'm not kidnapping msyelf either, I don't think you really do want to understand.

http://www.lfsforum.net/showthread.php?t=73874

Quote from DarkKostas :

0.6 version will be only compatible with 0.6 servers.
I am not sure but next version should be compatible with A1 too.
Thats why clients changed from 0.5 to 0.6

OK,
thanks allot

Crash of 0.6A1 dedi

Quote :

Problem Event Name: APPCRASH
Application Name: LFS.exe
Application Version: 0.0.0.0
Application Timestamp: 4df23fd3
Fault Module Name: LFS.exe
Fault Module Version: 0.0.0.0
Fault Module Timestamp: 4df23fd3
Exception Code: c0000005
Exception Offset: 00030000
OS Version: 6.1.7600.2.0.0.256.48
Locale ID: 1050
Additional Information 1: 4709
Additional Information 2: 4709938665fa84e3ff99c8cbe11223bb
Additional Information 3: 0486
Additional Information 4: 04866a78672ca5c2618c4ccff33ff08b

also after installing fresh Z28 and update to new patch, when you start LFS for first time, crashes after click button AGREE. You need to press this before installing new patch

Quote from reineh :

Did we just get a new clutch/engine model?
Suddenly I have much better/realistic throttle respons on idle and low revs. And the clutch doesn't overheat so easely.

Really great feel!!!

... or aren't we allowed to talk about that here

I dont feel anything of this you just mentioned :S

Great patch tho..

Quote from DANIEL-CRO :

Crash of 0.6A1 dedi


also after installing fresh Z28 and update to new patch, when you start LFS for first time, crashes after click button AGREE. You need to press this before installing new patch

It was reported some pages ago, put "Welcome Z 0" in the cfg.txt file to bypass the agreement screen.

Quote from Scawen :

I've removed some posts asking about this. I would like to make it quite clear.

This test patch is a test for the official version.
That means the official version will be just like this one!
The only changes will be small, like bug fixes and slight improvements.

There will be no more big changes.
There will be no new tracks.
There will be no new cars.

I hope that we can release the new official version next week.
That really should be the case, unless some unforeseen big problems come up.

I'm asking this just out of cuiosity, why are you axactly moving to a new letter? Why can't it be another 0.5Z patch?