Complaints about 3.1 (Was "maps by Ray")

[zydonk]

3- I haven't seen anything like that happen for a very long time. (Haven't played old unsupported versions such as 2.3.9 for a very long time, either.) If really wanting to work around bugs in very old versions (which don't even support true multiplayer), you can curve the cliff endings, and that might help.

"really old versions" - what? - are you counting in all the dud 3.1 betas. The "really old version" is the only one that works properly. And if you're only playing 3.1 betas then you can't be getting in much play...

[Cyp]

3- I haven't seen anything like that happen for a very long time. (Haven't played old unsupported versions such as 2.3.9 for a very long time, either.) If really wanting to work around bugs in very old versions (which don't even support true multiplayer), you can curve the cliff endings, and that might help.

"really old versions" - what? - are you counting in all the dud 3.1 betas. The "really old version" is the only one that works properly. And if you're only playing 3.1 betas then you can't be getting in much play...

If by "works properly" you mean "crashes frequently" and doesn't support multiplayer, and by "dud" you mean completely stable, then I agree.

[Originway]

"really old versions" - what? - are you counting in all the dud 3.1 betas. The "really old version" is the only one that works properly. And if you're only playing 3.1 betas then you can't be getting in much play...

said by one that hasn't played RC2? If you found bugs then did you report them on http://developer.wz2100.net/newticket as they said?
I don't know what the deal is with all the hate you guys do.
they already said the 2.3.9 version is discontinued and unsupported and all you are doing nothing but whining about it.

[raycast]

Please stop taking my maps thread off-topic into a 3.x flamewar. Thank you.

[Iluvalar]

If you found bugs then did you report them on http://developer.wz2100.net/newticket as they said?

Well, we do. here is mine : http://developer.wz2100.net/ticket/3287
A 7 month old bug... but they keep closing it without any fix.

[Cyp]

If you found bugs then did you report them on http://developer.wz2100.net/newticket as they said?

Well, we do. here is mine : http://developer.wz2100.net/ticket/3287
A 7 month old bug... but they keep closing it without any fix.

Please debug it and fix it, if you can reproduce it. It's hard for anyone to fix, when there's no way to reproduce it (and not sure if that particular issue happens for anyone else, even).

[Iluvalar]

That's where you are in a vicious circle. Because with all those changes I dont like : New synch using that annoying bullet time. Asymmetrical ray-casting, lack of unit previews, units clogging together, noob friendly power management,new depression inducing terrain blur and accuracy bug (despite I gave you a patch for that). I prefer playing 2.3.9 anyway. I'd rather help you launching 2.4.

1- I gave you a solution for a symetrical ray-casting.
2- I gave you a fix for accuracy.
3- I gave you a starting point for naval unit support.
4- I asked for easy improvement to add new research you refused to push in.

Each time, it end up into unimplemented patches. I really don't see why I should try another time to help you.

[Cyp]

...

Each time, it end up into unimplemented patches. I really don't see why I should try another time to help you.

Because if the crash is because of an actual bug in the game, a crash fix is much less likely to be controversial than an alleged fix which changes the game mechanics from weird to weirder.

[zydonk]

I prefer playing 2.3.9 anyway. I'd rather help you launching 2.4.

Hey, yay to this! Dump 3.1 and start again from 2.3.9.

On your own say-so, Cyp, the only thing 2.3.9 needs is to have the cliffs buffed up a bit. And the best thing anyone could say in praise of 3.1 RC2 was that the fans in the factory were spinning round...

In the early days of the 3.1 betas I raised bug reports, yet the basic problems persist. I'm not a techie, but it seems to me that the new terrain renderer is - directly or indirectly - the source of many of these probs. Dump this renderer, keep the netcode, and revert the balance to the original - now that there is a specific mp .wz. Call this version 2.4.1, and go on from there.

[Stratadrake]

Rumor has it that the units getting stuck together in clusters was also new as of the 3.x branch, that "they worked fine" in 2.3.9 . So the obvious question on this matter is: what changed?

[Per]

So the obvious question on this matter is: what changed?

A lot. Seriously. There are literally years of accumulated changes in 3.1 compared to 2.3.

[Stratadrake]

... and there's a readme file for that. But I intended that question to be solely about a unit's obstacle negotiation logic - the claim that you never had the logjams in 2.3 (or in original retail!) that you constantly do in 3.1 . What about this portion of the code changed?

I wonder if removing the group speed limit is related to that. I want that back as an option, damn it! If I order my T2 commander assault group towards a target, my hover repair units arrive first, followed by Lancer Wheels, the main Halftrack cannons and MGs, with the Tracked Commander arriving last. And my repair units being the first on scene is a major DO NOT WANT.

[Per]

The claim that you never had the logjams in 2.3 (or in original retail) is false. I remember trying to fix logjam issues for 2.3 as well. It has been a persistent issue since the original.

There were many problems with the old group speed limit, but the biggest was that in multiplayer it was a lie. The units would only be limited on your end, and lead to massive desyncs. A group formation code with speed limiting would be really nice, but it wasn't that. It was just a huge ugly hack that was rather broken by design in every aspect.

[Stratadrake]

The claim that you never had the logjams in 2.3 (or in original retail) is false. I remember trying to fix logjam issues for 2.3 as well. It has been a persistent issue since the original.

Thank you. But there is still an impression that it is much more noticeable now than before... do we have any isolated test conditions which show different results between then and now?

There were many problems with the old group speed limit, but the biggest was that in multiplayer it was a lie. The units would only be limited on your end, and lead to massive desyncs. A group formation code with speed limiting would be really nice, but it wasn't that. It was just a huge ugly hack that was rather broken by design in every aspect.

That's . . . interesting. Desyncs are a major problem, but I have to wonder why it never worked in multiplayer?

[Cyp]

...
That's . . . interesting. Desyncs are a major problem, but I have to wonder why it never worked in multiplayer?

Basically, formation speed limiting would make your client limit the speed of all units travelling to the same square. So if you enabled formation speed limiting, your client would think the enemy was moving their tanks slower than the enemy thinks they were.

It would have been a slightly less ridiculous, if you enabling formation speed limiting only affected your own units instead of enemy units, and if other clients were told to make the tanks move slower.

Not sure I agree that formation speed limiting was causing huge desynchs in 2.3, since that assumes that there wouldn't be huge desynchs anyway (or at least, that there would occasionally be some games without huge desynchs). Even starting two 2.3.9 clients and moving a truck around, the clients can't agree on whether the truck is driving east or west, so any talk of the quality of synchronisation in 2.3.9 (other than as a placebo) seems to me to have equal semantic content as a discussion on the quantity of invisible pink unicorns capable of dancing on the head of a pin.