► V12 Bugs

[Ladios]

It seems that whenever people connect to a map I made, the water texture will not download to them.

This is because torque hates water with a passion. Any textures on water will not download even with texture downloads on. The cheap way to go arround this is make a decal datablock with the water texture so people will download that and can see the texture.

[Ad Bot]

[M]

Where getSubStr() outputs a console error if a negative length was input, getWords() simply crashes. How odd

[-Jetz-]

This is because torque hates water with a passion. Any textures on water will not download even with texture downloads on. The cheap way to go arround this is make a decal datablock with the water texture so people will download that and can see the texture.

Clever. Workaroundy, but very clever. No real downsides. I'll roll with this.

[Wizzard1]

Bug: When opening the server list, clicking the column headers sorts the list in reverse order by default... Why?? :D

IE

Click the top of the server name column sorts in reverse alphabetical... When it's previously unsorted.

[Space Guy]

You can click it again for the other order.

[Wizzard1]

Yah, I know. But it's a shame to have to click it two times. It's aminor inconvenience, but it's illogical (unless there is some other undocumented feature preventing alphabetical sorting).

[BobSevenSevens]

Even from a clean install of Blockland, when you go to join a game, after you query internet servers and the server list has loaded, I still get this in console:

Code: [Select]

Warning- HandleSimplePingResponse()- address does not match
Also, when I type trace(1); into my console, I get console spammed with Joining()something- ### followed by Leaving()something - same # as above, and the numbers look just like the time my comp crashed after I epically glitched in BL.

Note: It actually doesn't say (something), but I can't get what it says.

[Space Guy]

Even from a clean install of Blockland, when you go to join a game, after you query internet servers and the server list has loaded, I still get this in console:

Code: [Select]

Warning- HandleSimplePingResponse()- address does not match
Also, when I type trace(1); into my console, I get console spammed with Joining()something- ### followed by Leaving()something - same # as above, and the numbers look just like the time my comp crashed after I epically glitched in BL.

Note: It actually doesn't say (something), but I can't get what it says.

That's probably for things like dead or ??? ping servers.

[BobSevenSevens]

That's probably for things like dead or ??? ping servers.

Which part, the trace(1); console spam or the HandleSimplePingResponse part?

[Space Guy]

HandleSimplePingRepsonse.

For the trace, this may be your problem:

If you trace in a server with some of my mods e.g. Portal Gun or Target Events enabled, you will get console spam due to the functions I used. Type cancel($SpaceMods::SpaceTick); in the console to stop this and spaceTick(); to start them working again. (targets won't move and portals will not function in the mean time)

[Radioactive Cat]

Im getting that when I load over 10,000 bricks.

[DracoSoul]

Im getting that when I load over 10,000 bricks.

If you die while bricks are loading, that happens.

[BobSevenSevens]

Im getting that when I load over 10,000 bricks.

Someone needs to flushtexturecache(); .

[Aloz]

Updated.  I am aware of the graphical corruption errors.  It is very difficult to debug because I cannot reproduce the problem.  I've only seen it a handful of times and it usually goes away quickly. 

Difficult to reproduce, huh?  Badspot, you've got an Nvidia card, right?  I ask because I've been dealing with this issue since v12 came out (before, I would see similarly weird stuff during loading, but it wouldn't stick around) and search about, I had come to think it might be somehow related to a previous issue with Torque/ATI/Vista (though I am using XP--I remember running across people, in my searches, who experienced problems with Torque/ATI/XP as well).  If it's graphics vendor invariant, though...

I've used two different ATI cards with it, and it appears to be (ATI) graphics card invariant.  I've gone through 5 graphics driver revisions, and it seems to be (ATI) driver invariant.  Now that I'm thinking about it, I think I'll try it on my laptop.  It's got Intel integrated graphics.  Runs Blockland fine, but so slowly that I hadn't even considered it until now.  If I'm getting it there as well, that'll be very interesting.

I don't know that it starts going weird at 10,000 bricks, but it doesn't start doing its little glitchy thing until I have a lot.  I'm not sure how many 'a lot' is, but I think it had to be more than 10,000.  If it's not 'a lot', no problems.

Now, some pictures (of a 60,000+ brick build in progress).

Loading:


After-loading:

Note how the glitch is twitching/moving even though the view point is still.  It didn't do that much until my build really started getting large.

Also note:  naturally, this all had me worried for the state of my video card.  I ran it through two video card stress testers (one D3D, one OGL).  Nothing wrong with it.  I considered that bad RAM causing nonsense to make it into the video card can lead to graphical nonsense (I've seen this before, thought it was, if anything, even crazier and caused other random things), but 8 hours of memtest86+ turned up nothing--did that a couple times, actually.  Other games run fine.  Other things work fine.  To my delight, nothing's wrong with my machine!

Anyway, it seems to definitely be tied to the number of bricks being loaded and planted in some way.  If you like, I could upload/attach the save featured in my pictures.  You've probably already got saves with a asston of bricks in them but, what-the-hey, couldn't hurt to offer.

Also also note:  flushtexturecache(); does not seem to do anything to help the matter.  But it did seem to, a little bit, before I reached quite that many bricks.  I think.  Don't know if that helps.

EDIT:  Size certainly enters into it.  Loaded a 11,737 block file.  No problems.

UPDATE:  The weirdness appears on my Intel graphics during load, but goes away afterwards.  No problems after load, though (aside from being ass-slow on min settings).  Interesting, I think.  Oh, and it's 80,000+ bricks, apparently.  I misremembered earlier.
I should note that the different ATI cards I tried it with are similar in architecture.  One's an HD 4830 (RV770LE), the other an HD4890 (RV790XT)--so they're both in the R700 family, anyway.  I'm currently using the HD4890--I've got the other card, still, but don't have anywhere to put it, so I can't test with it again.

[Doomonkey]

Flush texture cache ();