0.2.8.3: It's here! Yay!

Help test release candidates for the next release
User avatar
Z-Man
God & Project Admin
Posts: 11585
Joined: Sun Jan 23, 2005 6:01 pm
Location: Cologne
Contact:

0.2.8.3: It's here! Yay!

Post by Z-Man »

Right, let's get this going. There have been code changes since the last RC, so we can't just build it directly. Instead, I'll build an alpha from current SVN and put that up for really quick testing. If I'm not too tired, I'll be doing the real thing right after the TST. Speaking of tired, the alpha build won't make it today. Should be up tomorrow morning.

Code: Select all

Changes since 0.2.8.3_rc4:
- Intercepted chat commands are now written to ladderlog.txt. Format:
  COMMAND /command-intercepted player [command arguments]
- Increased default speed of server pinging
- Reordered game menus a bit
- fast finish works again
Oh yeah, and the freezers Phytotron reports are the fault of the AI's pathfinding. I think. They do cause micro-freezes here, anyway. I can't see a good way to fix that without completely rewriting it, so I just disable the pathfinding if it is found to cause freezes. It's not a too useful feature anyway.

For those too lazy to scroll down, 0.2.8.3 is available on Launchpad and SourceForge.
Last edited by Z-Man on Sat Feb 20, 2010 8:10 pm, edited 2 times in total.
User avatar
Phytotron
Formerly Oscilloscope
Posts: 5041
Joined: Thu Jun 09, 2005 10:06 pm
Location: A site or situation, especially considered in regard to its surroundings.
Contact:

Re: 0.2.8.3: Release Process

Post by Phytotron »

Any idea on the timer hiccups, timeouts, and so forth?
User avatar
Z-Man
God & Project Admin
Posts: 11585
Joined: Sun Jan 23, 2005 6:01 pm
Location: Cologne
Contact:

Re: 0.2.8.3: Release Process

Post by Z-Man »

The timer hickups, IIRC, were problems with one of the masters not working properly. The timeouts you reported are that one server switching IPs randomly. You can't keep contact with a server as it switches IPs. Well, you can, in theory, but for that the server would need to be able to tell the clients it has switched IPs and it can't do that directly because the messages would be intercepted by the clients' NATs. It's way too complicated to work around that, and there still would be severe hickups. So just stop that server from being silly and fix its IP.

And I blissfully ignore the etc :) Oh wait, one I remember, the authentication thing. Yeah, you can't say "/login" (or press the "authenticate" menu button), then enter your global ID. That way, you only can login with local accounts. When the login prompt on your client comes up, the authority is already fixed. So, if you haven't entered your global ID, the only way to log in is to say "/login <authority>" or "/login <global ID>".
User avatar
Z-Man
God & Project Admin
Posts: 11585
Joined: Sun Jan 23, 2005 6:01 pm
Location: Cologne
Contact:

Re: 0.2.8.3: Release Process

Post by Z-Man »

User avatar
Phytotron
Formerly Oscilloscope
Posts: 5041
Joined: Thu Jun 09, 2005 10:06 pm
Location: A site or situation, especially considered in regard to its surroundings.
Contact:

Re: 0.2.8.3: Release Process

Post by Phytotron »

Good news: I don't get the freezes in the local game anymore.
Bad news: It still gets very jittery in a way that an online SP game with multiple AIs will not.

Would it be useful if a I made a recording (or two separate ones) of both my local game and an online server with multiple AIs? Now that Shrunkland and BiH are up, I can dump a bunch of AIs into them.


I still got bumped from a server(s). The differing IP's thing wasn't the only issue—really just an isolated one, and was apparently fixed by whatever Gene did. Most of the time it's just a plain time out or whatever, although with no error message. Like I mentioned, I see a lot of other players suffering from this, and it's something that doesn't happen when using the 0.2.8.2.1 client—that goes for the timer hiccups, as well.

And then one of the "so forth"s was the deal where there's a little lag hit many times someone crashes. In fact, I usually see someone's ping jump just after, including my own. Thing is, it also can affect other players. This is something I've noticed since the first 0.2.8.x version, and reported way back then as part of the "stick and slip" phenomena.


I understand wanting to blissfully ignore these sorts of things, but in my view these are issues of basic functionality, certainly more critical than special features and stuff.
User avatar
Z-Man
God & Project Admin
Posts: 11585
Joined: Sun Jan 23, 2005 6:01 pm
Location: Cologne
Contact:

Re: 0.2.8.3: Release Process

Post by Z-Man »

Well, the AIs still do some work with long range raycasts, it's natural that that causes jittering so some extent. I really don't think anything can be done about it until we can make the grid structures easier to raycast in (read: fragment less).

About the disconnections/timeouts: sorry if something important here flew under our radar. I've never seen it myself, which usually is the best thing to get me to work :) Send a recording of you timing out, best on my server (which, of course, sucks if no tournament is going on, I'll see if I can make it non-sucky one of these days, the provider upgraded the CPU cycles allotted).

Aaanyway, apparently, nothing got destroyed since rc4, and since the disconnection issues have been around for a bit, they can certainly wait a month or two. We have a release! 0.2.8.3 is live on Launchpad and SourceForge. I'm a tad too tired to write an announcement on SourceForge right now, if someone else wants to do it, go ahead.
User avatar
DDMJ
Reverse Outside Corner Grinder
Posts: 1882
Joined: Thu Jun 08, 2006 12:15 am
Location: LA, CA, USA, NA
Contact:

Re: 0.2.8.3: It's here! Yay!

Post by DDMJ »

Yay! 1310 days after 0.2.8.2 :P.

<3
User avatar
Phytotron
Formerly Oscilloscope
Posts: 5041
Joined: Thu Jun 09, 2005 10:06 pm
Location: A site or situation, especially considered in regard to its surroundings.
Contact:

Re: 0.2.8.3: Release Process

Post by Phytotron »

Z-Man wrote:Well, the AIs still do some work with long range raycasts, it's natural that that causes jittering so some extent. I really don't think anything can be done about it until we can make the grid structures easier to raycast in (read: fragment less).
OK, but then why should there be such a profound difference between local and online games? I can go into Shrunkland or BiH, add in a lot of AIs, and have no jittering or freezing issues (other than some framerate and lag). Like I said, I could do a recording or two comparing local to online, if you like.
About the disconnections/timeouts: sorry if something important here flew under our radar. I've never seen it myself, which usually is the best thing to get me to work :) Send a recording of you timing out, best on my server (which, of course, sucks if no tournament is going on, I'll see if I can make it non-sucky one of these days, the provider upgraded the CPU cycles allotted).
Er, I haven't noticed a way to predict or deliberately cause a timeout (or whatever it is). Do you not mind a long recording?

And which is your server?

DDMJ wrote:Yay! 1310 days after 0.2.8.2 :P.
Maybe he should have waited 27 more days.
User avatar
Z-Man
God & Project Admin
Posts: 11585
Joined: Sun Jan 23, 2005 6:01 pm
Location: Cologne
Contact:

Re: 0.2.8.3: Release Process

Post by Z-Man »

Phytotron wrote:
Z-Man wrote:Well, the AIs still do some work with long range raycasts, it's natural that that causes jittering so some extent. I really don't think anything can be done about it until we can make the grid structures easier to raycast in (read: fragment less).
OK, but then why should there be such a profound difference between local and online games? I can go into Shrunkland or BiH, add in a lot of AIs, and have no jittering or freezing issues (other than some framerate and lag). Like I said, I could do a recording or two comparing local to online, if you like.
Easy, the AI is solely handled on the server, to the client, it looks like just a regular (a bit retarded) other player. On the server, small hickups don't cause jittering on the clients, only usually unnoticeable micro-lags.
Phytotron wrote:
About the disconnections/timeouts: sorry if something important here flew under our radar. I've never seen it myself, which usually is the best thing to get me to work :) Send a recording of you timing out, best on my server (which, of course, sucks if no tournament is going on, I'll see if I can make it non-sucky one of these days, the provider upgraded the CPU cycles allotted).
Er, I haven't noticed a way to predict or deliberately cause a timeout (or whatever it is). Do you not mind a long recording?
No, not at all. Of course, there may be playback difficulties, but if need be, I'll debug the thing on the Mac. And if you don't think it will make the bug not show, quitting and restarting the client every hour or so would be sweet.
Phytotron wrote:And which is your server?
Those with "Z-Man" in the name :) Currently, one of the TST servers. It's still in team sumo mode. And, I think, an AFL one. If you prefer it to run a semi-shrunkland clone, just send me a config, I'll run anything.
User avatar
Phytotron
Formerly Oscilloscope
Posts: 5041
Joined: Thu Jun 09, 2005 10:06 pm
Location: A site or situation, especially considered in regard to its surroundings.
Contact:

Re: 0.2.8.3: It's here! Yay!

Post by Phytotron »

AI deal: Ah, that makes sense. I guess I'll just put up with the jitters—better than the alternative.

I'll do some intermittent recordings, see if I get (un)lucky and catch a timeout (or whatever it is).

I'll get back to you on the Mac keyboard thing here as well.
User avatar
compguygene
Adjust Outside Corner Grinder
Posts: 2342
Joined: Thu Aug 21, 2008 12:09 pm
Location: Cleveland, Ohio
Contact:

Re: 0.2.8.3: It's here! Yay!

Post by compguygene »

Not a bug report here. Just a big thank-you to Z-man, and all the devs. .2.8.3 seems pretty darn nice and stable. I like the menu changes! After reviewing the logs from the dev meeting, it was nice to see stuff start rolling along.
Armagetron: It's a video game that people should just play and enjoy :)
https://bit.ly/2KBGYjvCheck out the simple site about TheServerPharm
FroZen
Average Program
Posts: 62
Joined: Sun Jan 28, 2007 1:14 am
Location: United States
Contact:

Re: 0.2.8.3: It's here! Yay!

Post by FroZen »

Well, I have a bug to report, if it hasn't been reported already.
In 0.2.8.3 on a Mac if you toggle out of fullscreen, change windows, and come back to Armagetron there is no sound, and it stays muted till the start of the next round (in both fullscreen and non-fullscreen).
0.2.8.2.1 doesn't do this; here's the recording if it helps.
You do not have the required permissions to view the files attached to this post.
User avatar
Z-Man
God & Project Admin
Posts: 11585
Joined: Sun Jan 23, 2005 6:01 pm
Location: Cologne
Contact:

Re: 0.2.8.3: It's here! Yay!

Post by Z-Man »

Zounds! That happens on Linux, too. Serves me right for never playing with sound enabled. I hope I'll be able to blame SDL.
User avatar
Lucifer
Project Developer
Posts: 8640
Joined: Sun Aug 15, 2004 3:32 pm
Location: Republic of Texas
Contact:

Re: 0.2.8.3: It's here! Yay!

Post by Lucifer »

While I'm thinking about it, just so someone knows, my exploration last night getting the cycle motor and sparks/grinding sounds back into trunk showed that SDL_sound is indeed incompatible with SDL_mixer. No googling helped.

Just throwing it out there before I forget, in a random thread where it doesn't belong.
Image

Be the devil's own, Lucifer's my name.
- Iron Maiden
User avatar
Z-Man
God & Project Admin
Posts: 11585
Joined: Sun Jan 23, 2005 6:01 pm
Location: Cologne
Contact:

Re: 0.2.8.3: It's here! Yay!

Post by Z-Man »

Hah! While I can't blame the sound disappearance bug on anyone but myself (forgot a 'static' keyword in a place), I can totally blame the fact that it made it into 0.2.8.3 on YOU. Yes, YOU, the reader. The bug was at least in rc3 and rc4 (possibly even rc2), and wasn't noticed or reported. That, and the fact that the sound reappears the next round, tells me we needn't be in a terrible hurry to release a fix :)

Lucifer: Hmm, yeah, the two are incompatible, but have a good look at the 0.2.8 code: it has the compile time option to activate SDL_MIXER and use that for everything.
Post Reply