Who wants to try a new map?

[hee]

I am glad that some of you are finding the hidden stuff. There is a new version at ensle.org where I have removed all of the lights in the void and fixed the dark spot near the left red lift.

The bump open and navigation points: I am not sure that is a problem. I tested the paths extensively by leading or following the bots through the various paths.

Thanks for the updated version. A couple of points

1 Revised versions of maps should have a different name to the original version - you can add a suffix like 1 or ][ or rev, etc - so that there are no mismatches;

2 Of the two pathnodes with no incoming connections, PathNode417 looks to be quite important as it will force bots to use one particular route. PathNode17 is not very important in terms of routes but since it has no incoming paths it is 'useless'. If you use the Editor in 469b, right click on Pathnode417 and you will see that the white arrows indicate that it has several paths leading from it but no paths leading to it.

I checked 417 and it was sticking in the ground so obviously a problem. (I wondered why bots were taking the long way around to get out of the bay.) I checked 17 and it was obviously too high, though the bots were just bypassing it so it didn't make any difference.

So anyway a new version (1.2) has been uploaded that finally gets rid of all those annoying black spots. Thanks for the input.

A reminder to all. The minimum number of bots is 32 [addbots 16]. Even with that I sometimes have a hard time finding them. It is a big area. It is also nice to have at least one other human player so you can have dogfights with the jets otherwise the human player has a definite advantage. Actually, if you have the bots turned up high, the jet is not as much as an advantage as you might think.

[sektor2111]

@sektor I just realized, maybe PathNodes with no incoming connections can still be useful, so if bots get in them they know how to get out? Like, if they get propelled into a high ledge, othewise they'd stay stuck up there, wandering, instead of doing the obvious thing and jumping down. (Maybe with LiftExits as a little courageous stimulus, am I rite!)

Not entirely true and not entirely false - Logic is different. If you have a small house/big-box and a Bot is thrown over it but not dead, it won't resume navigation very soon if on the house/big-box it's not a NODE heading down, only down (perhaps if it's wandering closer to a ledge and can find lower routes). Actually in this case such a Node doing a one-way path is NEEDED, but it's not needed in the middle of a road between flags. Such situation makes all that route unavailable (only an accident makes pawn to get over it coming to a nearby item and being boosted or charging at enemy) and Pawn usually will find another way - if that another way is available.
Here an advanced mapper capable to write a few lines can do a small paths controller actor removing/and restoring a reachSpec causing a route deviation from time to time concerning a direction, Reds are coming to Blues but Blues are moving on a different way, and switching these... Situation is suitable when map has a lot of complex small routes - similar to CTF-Coret and even a bit more complex.
A LiftExit without a full combo is not helping with anything, it's just a SpecialHandling process for nothing. In old patch plugins which I wrote one of patching tasks was exactly this destruction of such useless points based on guessing and "let me try this". It's all about reachSpecs and nothing more.

Self Schooling:
I recommend you to do an U like map (three tunnels) Place flags at extremes and in the middle go play with a node rebuilding paths - see what it does this node when has no incoming connections from any kind - MOVING IT UPPER. This is the learning way instead of reading some "tutorials" explaining fake ramp stories and blabbering. An advice as a general hint: Forget messing up nodes this way.
If it's hard for you to clone a bad tactical situation I can do two samples of maps: good and bad (like I did before for other stages) showing the "How To" and which will have around 2-3 downloads - demonstrating level of interest in learning but asking for "good tutorials" - perhaps direct samples are not good enough for clever artists... neither stock samples already done.

Mini Techs: When PathNode looks strange any you have doubts about it, use advanced actor editing and look at its upStreamPaths[x] array. There should be some data not only full filled with "-1". If this node has upStreamPaths 1 to 15 full filled with "-1" and this node is placed in a tunnel being an important part of a route, you can forget that route because node has no reachspecs as entry/entrance/coming to this node, all it does is breaking a route.
If this situation is found in a FlagBase, only God can say if your Bot will find this flag - future plans for my custom CTF would be dealing with Flag here without hacking paths... this work of me looks needed because of some "smarts" designing flags without anything at upStreamPaths because they don't have minimum amount of "inspiration" required for this task. Flag can have all required incoming upStreamPaths even if it's placed in a higher spot if mapper is capable to fool that dumb Editor... and no, PathNode nearby FlagBase doesn't help with anything because of internal logic routines from these devs at C++ never explained as it should.

[OjitroC]

So anyway a new version (1.2) has been uploaded that finally gets rid of all those annoying black spots. Thanks for the input.

A reminder to all. The minimum number of bots is 32 [addbots 16]. Even with that I sometimes have a hard time finding them. It is a big area. It is also nice to have at least one other human player so you can have dogfights with the jets otherwise the human player has a definite advantage. Actually, if you have the bots turned up high, the jet is not as much as an advantage as you might think.

A couple of points on the new version :
1 Map names should not contain a . (as in 1.2) - UT reads this as an URL and so trying to run the map results in a 'connection failed' message;
2 It would be a good idea to include a readme with the map - particularly if you have recommendations on the ideal number (or range) of 'players' - and to explain anything else that should be noted about the map so that people who obtain it from sources other than that linked in this thread are aware of that useful information.

[hee]

So anyway a new version (1.2) has been uploaded that finally gets rid of all those annoying black spots. Thanks for the input.

A reminder to all. The minimum number of bots is 32 [addbots 16]. Even with that I sometimes have a hard time finding them. It is a big area. It is also nice to have at least one other human player so you can have dogfights with the jets otherwise the human player has a definite advantage. Actually, if you have the bots turned up high, the jet is not as much as an advantage as you might think.

A couple of points on the new version :
1 Map names should not contain a . (as in 1.2) - UT reads this as an URL and so trying to run the map results in a 'connection failed' message;
2 It would be a good idea to include a readme with the map - particularly if you have recommendations on the ideal number (or range) of 'players' - and to explain anything else that should be noted about the map so that people who obtain it from sources other than that linked in this thread are aware of that useful information.

Thanks, will do.

[OjitroC]

Mini Techs: When PathNode looks strange any you have doubts about it, use advanced actor editing and look at its upStreamPaths[x] array.

What does one use for advanced actor editing?

[Gustavo6046]

Mini Techs: When PathNode looks strange any you have doubts about it, use advanced actor editing and look at its upStreamPaths[x] array.

What does one use for advanced actor editing?

Presumably the clipboard?

And, thanks for your insights Sektor! I'll make sure play around with some nodes

[OjitroC]

Presumably the clipboard?

Yes, indeed - thanks for that info.

[sektor2111]

Cough ?
Advanced actor editing it's not really a Clipboard task but in this case a Copy Paste of actor in notepad shows what is about.

However... did you even see in forum this command ?

Code: Select all

EditActor name="PathNode0"

How about MapGarbage ?

Also Buggie's builder tool for Advanced Actor editing does the same thing.
viewtopic.php?f=5&t=14061&p=127451&hili ... me#p127450
And then ? And then any map can be checked visually BEFORE to be tested by others in outside and facing identical problems.

[Gustavo6046]

Cough ?
Advanced actor editing it's not really a Clipboard task but in this case a Copy Paste of actor in notepad shows what is about.

That IS what the Clipboard is, silly. Copying and pasting are clipboard operations. Duhhhh.

[sektor2111]

No, I wasn't copy-pasting, figuring these nodes in 469b 469a 436+XCGEv224 is easy as long as reachspecs are marked as arrows, it can't be more simple than looking at Node.

[Gustavo6046]

Yeah. I mean I felt like you were talking about copy-pasting (which is what "clipboard" means, an operating system or windowing system's clipboard is what makes copying and pasting text possible, see Xorg or Windows).

You yourself seemed to agree, even--

Advanced actor editing it's not really a Clipboard task but in this case a Copy Paste of actor in notepad shows what is about.

So it is a clipboard task? I'm confused.

[EvilGrins]

So it is a clipboard task? I'm confused.

It's an old Windows reference, when you copy & paste prior to the pasting part the content is in Windows clipboard.

Sometimes indicated by a clipboard icon.

[sektor2111]

Bah, boys, don't you get that these two things are doing the same result ? You can see upStreamPaths data by firing Advanced Actor Editing OR Copy-Pasting this in a Text Editing app - yeah even NotePad++ not only NotePad... The goal is figuring data doesn't matter HOW. What the heck is all the rest of discussion about ? Clipboard discussions are OFF-TOPIC. Go post them into other OS forums. Discussion it's about testing and reporting bugs - I described what do you need to look at - upStreamPaths, for a ClipBoard definition - get a PC guide for dummies if you are having issues with/in/about using a computer.

Chapter Testing map - try new map
As long as there are nodes which are not having ALL navigation requirements Pawn will discard that route following another way. Why ? BECAUSE I told this X times: PathNode comes from UNREAL and it aims SKAARJ not your UT bots - that's why.

Code: Select all

//- add new pathnode to level at position spot
void FPathBuilder::newPath(FVector spot)
{
	guard(FPathBuilder::newPath);
	
	if (Scout->CollisionHeight < 48) // fixme - base on Skaarj final height
		spot.Z = spot.Z + 48 - Scout->CollisionHeight; // sektor Note: Scout is bigger than Bot and not a DeathMatch player - expect funky reactions
	UClass *pathClass = FindObjectChecked<UClass>( ANY_PACKAGE, TEXT("PathNode") );
	APathNode *addedPath = (APathNode *)Level->SpawnActor( pathClass, NAME_None, NULL, NULL, spot );
	//clear pathnode reachspec lists - // sektor2111 - And LEAVING SPECS IN LEVEL, GENIUSES ? GIVE ME A BREAK...
	for (INT i=0; i<16; i++)
	{
		addedPath->Paths[i] = -1;
		addedPath->upstreamPaths[i] = -1;
	}

	unguard;
	};
// ReachSpecs manipulation should be mandatory implemented if you were using //FixMe because these will need corrections impossible to do WITHOUT Native/Uscript support.

If this node it's on a higher spot (or moved without knowledge) even a three years old kid can figure that he cannot walk in AIR - yeah, people are having PHYSICS issues which even kids are aware of these impossible physical actions, exactly like UnrealEngine1, and UE1 won't create paths heading to an air node unless it will be a R_SPECIAL one causing an impossible move and Pawn will run in place trying unsuccessfully to get to the node - forget such practices if you have mental issues in understanding movement physics and physics generally...
Either way I don't understand why did they do this:

Code: Select all

		distance = (int)path.Size(); //fixme - reachable code should calculate // sektor2111 - Fixing WHO ? You forgot to distribute source-code...
		if ( reachFlags & R_SWIM )
			distance *= 2;

I won't ask if these fixme were touched in 469 because nobody will answer anyway. ut99.org is probably not an UT99 place .

[Gustavo6046]

So it is a clipboard task? I'm confused.

It's an old Windows reference, when you copy & paste prior to the pasting part the content is in Windows clipboard.

Sometimes indicated by a clipboard icon.

Umm... except, not really. It has always been called a clipboard.

When you copy something, the text is copied to the clipboard buffer; when you paste, whatever is in the clipboard is inserted into the text cursor's location. Cutting moves the text, by erasing it from the original source location after sending it to the clipoboard.

It's just that in Windows there is no icon for it anymore.

In Linux, the clipboard is managed by the X.Org windowing system. That is why you can't use it in a text terminal, like tty2 for instance.

Bah, boys, don't you get that these two things are doing the same result ? You can see upStreamPaths data by firing Advanced Actor Editing OR Copy-Pasting this in a Text Editing app - yeah even NotePad++ not only NotePad...

Yeah, yeah, I see what you mean. Not everything about an actor's state can be accessed from the Editor's Actor Properties window. Many of the variables are 'hidden', and have to be accessed by copying the actor from the Editor interface, which copies a textual serialization of the actor's state into the clipboard, whereupon you may paste it in your text editor of choice. Thank Epic for choosing to use a textual serialization! It's funny how human-readable this format is. They could've used anything else, and yet they chose to use T3D for clipboard operations. Whew.

The goal is figuring data doesn't matter HOW. What the heck is all the rest of discussion about ? Clipboard discussions are OFF-TOPIC. Go post them into other OS forums. Discussion it's about testing and reporting bugs - I described what do you need to look at - upStreamPaths, for a ClipBoard definition - get a PC guide for dummies if you are having issues with/in/about using a computer.

It's not a discussion about computer maintenance. You complained that it had nothing to do with the clipboard, without even knowing what I was actually talking to. I did mean copying and pasting, because that's exactly what a clipboard is!

Your mind probably went to this instead:


That is not the kind of clipboard I meant.

And in fact, I was only defending myself, and upholding what I stated about clipboard operations being able to peer into hidden state of actors in the Editor. I did that because you said I was incorrect, without even knowing what I was talking about precisely.

I believe this is more of a language barrier issue than anything else, so I advise you start to learn English by turning off whichever Google Translate browser extension you're using right now. That'll be a very good start.

Code: Select all

		distance = (int)path.Size(); //fixme - reachable code should calculate // sektor2111 - Fixing WHO ? You forgot to distribute source-code...
		if ( reachFlags & R_SWIM )
			distance *= 2;

I won't ask if these fixme were touched in 469 because nobody will answer anyway. ut99.org is probably not an UT99 place .

The OldUnreal team is bound by a non-disclosure agreement with Epic, preventing them from distributing the source-code proper. Perhaps this NDA also prohibits them from talking (aka disclosing details) about unreleased parts of the source code?

I still suspect Epic will lift the NDA once they see v469 in a nice, working state. I can't wait to see UT99 open-source at last! Imagine all the source ports! Better late than never, ugh, it's been so long!

[sektor2111]

Then check those upStreamPaths and let me know if anything else is "Barrier".

Like I said, those Flag errors to me are pointless, I have a tool heading me at enemy base doesn't matter hidden passages, and following normal shortest routes.