SRB2 Doom Builder

Status
Not open for further replies.
I keep getting this same error when I try to start up BuilderSRB2.exe:

Error 457 in LoadConfiguration: This key is associated with an element of this collection.

I installed the latest version of Doom Builder and I extracted the zip file into that directory, and I get this error when I start it up.

EDIT: Nevermind I just saw the fix Oogaland posted for that problem. ^.^
 
JJames19119 said:
What I'm reporting is likely a suggestion more than a bug but SRB2DB doesn't notify you that a linedef is missing textures when a ceiling with the flat F_SKY1 is covering it. While this is normal behaviour in DB, in SRB2, F_SKY1 is more or less an "invisible" flat, so any linedefs the sky ceiling is covering is still visible, making the Hall of Mirrors effect pop up when you don't set proper textures. What would be nice is if DB told you that you needed a texture on a linedef even if it's being covered by a ceiling with the flat F_SKY1. This makes it to where you can avoid the hall of mirrors effect more efficiently.

I just realized it does the same to floors with the flat F_SKY1 though I can't imagine anyone who would run into that situation...

Anyways, a handy suggestion would be putting that sector height information that you get making an FOF through the dialog and putting in the sector properties dialog. There's well enough space for it under the sector height info and it would be really useful for making stairs, same height sectors or anything that concerns heights.

EDIT: Found another bug.

http://img.photobucket.com/albums/v340/JJames/a9123889.png

Now, whe I select the red linedefs by normal clicking, they're selected normally. When I lasso select the blue linedefs however, the green linedef is selected with them, despite not even being near them! Do you know what might be causing this odd bug or do you need the wad I demonstrated this in?
 
JJames19119 said:
What I'm reporting is likely a suggestion more than a bug but SRB2DB doesn't notify you that a linedef is missing textures when a ceiling with the flat F_SKY1 is covering it. While this is normal behaviour in DB, in SRB2, F_SKY1 is more or less an "invisible" flat, so any linedefs the sky ceiling is covering is still visible, making the Hall of Mirrors effect pop up when you don't set proper textures. What would be nice is if DB told you that you needed a texture on a linedef even if it's being covered by a ceiling with the flat F_SKY1. This makes it to where you can avoid the hall of mirrors effect more efficiently.
I'm not sure if this is quite what you mean: http://homepages.inf.ed.ac.uk/s0569864/sky.wad. The linedefs of the innermost sector should have upper textures on their reverse sides, but they're not flagged as such. However, it doesn't quite seem to match with your described problem, as the sky does obscure the linedefs of the middle sector (so it's not just functioning as a null flat in this case).

As for the line selection problem: I have a nasty suspicion that this is the sort of thing that'll disappear if you close DB and open it again. If the problem persists, though, it would be useful for me to have the wad if you're willing to send it.

Danke schön.

Oh, and I finally twigged what the right-click-to-end-linedef-drawing is for: it's designed for occasions when you use it to split sectors by bisection, rather than to create a new sector within an existing one. And it does seem to adjust existing sidedef sector references, after all.
 
Again, sorry for the double post but this is important.

The drawing sectors counter-clockwise bug strikes again but this time, on a much worse scale.

http://img.photobucket.com/albums/v340/JJames/makesnosense1.png
http://img.photobucket.com/albums/v340/JJames/makesnosense1.png
http://img.photobucket.com/albums/v340/JJames/makesnosense1.png

See the right most sector there? I drew that one counter-clockwise...when the edit sector properties came up, it gave me the properties to the thok barrier, which I found very strange. Especially since the sector wasn't even connected to it at all. I hit cancel and I got the selections you see in those pictures. The red dot displays the mouse position in each one. When I went to the Display Map Errors dialog, it told me that six, yes six sectors were unclosed. I attempted to fix all the linedef references only to find out that one of the sectors was unclosable...it had linedefs that I could not select, furthermore it made saving the map completely impossible because the nodesbuilder rejected the map each time due to the corrupted sector. The sector was supposed to be around the top right vertex on the sector I drew but no matter how hard I searched, I could not find any linedefs that belonged to this sector. It forced me to use a backup. I really suggest this be fixed ASAP, it's made me scared to drew sectors now.

EDIT: Okay, after restoring my backup I tried drawing the sector clockwise...DB messed up and reproduced the error. Now I can't draw the sector at all...

EDITv2: Aha! I got it, it's not how I draw the sector...it's where the starting vertex is placed. If I press Insert in linedefs mode and start drawing the sector from either the top right or top left vertex in those pictures, I get the bug I mentioned but I start anywhere else, I get a clean sector. Now I can draw sectors at peace...
 
KingofFlames said:
I'm getting sick of this sector-drawing bug, I literally can't map until that damned thing is fixed.

If it helps...:

JJames19119 said:
EDIT: Okay, after restoring my backup I tried drawing the sector clockwise...DB messed up and reproduced the error. Now I can't draw the sector at all...

EDITv2: Aha! I got it, it's not how I draw the sector...it's where the starting vertex is placed. If I press Insert in linedefs mode and start drawing the sector from either the top right or top left vertex in those pictures, I get the bug I mentioned but I start anywhere else, I get a clean sector. Now I can draw sectors at peace...
 
KingofFlames said:
I'm getting sick of this sector-drawing bug, I literally can't map until that damned thing is fixed.
Sorry about all the trouble this is causing. I'll do my best to fix this and get a new build out tomorrow.
 
A small Suggestion for SRB2builder

Hi, Oogaland

Some suggestions for SRB2builder

1) In the "Find and Replace" dialog, there is no way of changing light levels in a sector.

2) Have a way to select all sectors or something to that effect.

3) You should start making documentation for all of the features that are in SRB2builder so a newbie can figure out how to use each feature.

4) Get SRB2builder to save and run a map properly when the blockmap is very big. (Examples, MR - AGZ2, Ree-c's Sunken Palace, etc)

Regards,

Sonict

EDIT: You are late with the build that you were suppose to release. :P
 
r17

New build: http://homepages.inf.ed.ac.uk/s0569864/builder_srb2-r17.zip. [Insert usual alpha-caveats here.] New this time:

Code:
! Texture previews weren't updated correctly when the info bar was vertical.
! CS space-finding algorithm was STILL buggy.
! CS hint thing being outside a sector was flagged as an error.
! Silly bug that would prevent DB from loading if show3dsechighlight wasn't set.
+ Sectors containing FOFs are drawn in a different colour.
! When lassoing linedefs, non-lassoed linedefs whose vertices were previously
  selected would always become selected themselves.
+ Adjacent sector info shown in sector dialogue.
! Additive selection couldn't be forced when it was off by default.
! Drawing linedefs on top of existing ones was broken.
! Texture requirement identification wasn't working correctly with zero-height
  sky-containing sectors.
+ Revamped the testing options.
+ Axis-drawing and FOF-colouring can now be disabled.
+ Light levels can now be searched for.
! Find-dialogue height behaved a little strangely.
+ Added Select All, Select None and Invert Selection options to the Edit menu.
+ Added an 'allowed to be missing after nodebuild' flag for map lumps. Reflagged
  BLOCKMAP accordingly.

Drawing linedefs on top of existing ones was broken.
This is the one that's been messing up sector references hither, thither and yon. The bug only used to occur when drawing one-fracunit-wide sectors, which I fixed several builds ago, but apparently caused this mess in the process. In the interest of completeness, I'll point out that the bug still exists if you contrive to draw a sector with a width of less than half a fracunit...

Sectors containing FOFs are drawn in a different colour.
It's disabled by default, though, since it can make things look overly busy with even a few FOFs.

When lassoing linedefs, non-lassoed linedefs whose vertices were previously selected would always become selected themselves.
This one was old -- and for once, not of my making. Odd that it hadn't been noticed 'til now.

Many thanks for all the bug reports and suggestions, particularly to JJames19119 for his several gargantuan posts of late.
 
Glee! I'm downloading this now and I'm very excited to try this out! It has features and bug fixes which I was really looking forward to! :D

And GCFreak: http://www.sepwich.com/ssntails/srb2/specials.html#s998

EDIT: Guh, beaten to the punch by Shuffle. >_>

EDIT2: Okay, I just tried testing out and no matter what parameters I feed it, it makes SRB2 come up with a weird "P_InitPicAnims: bad cycle from AFALL1 to AFALL2" error, making testing impossible...
 
JJames19119 said:
EDIT2: Okay, I just tried testing out and no matter what parameters I feed it, it makes SRB2 come up with a weird "P_InitPicAnims: bad cycle from AFALL1 to AFALL2" error, making testing impossible...
Could you set the executable in the Testing options to this EXE, please, and post what it says when you try testing? It'll echo the arguments passed to it on the command line. There're one or two other things wrong with testing, so I'll release a quick patch later today, assuming I can get to the bottom of this one.
 
KingofFlames said:
Is the evil sector-drawing bug fixed?
Yes:
I said:
Drawing linedefs on top of existing ones was broken.
This is the one that's been messing up sector references hither, thither and yon. The bug only used to occur when drawing one-fracunit-wide sectors, which I fixed several builds ago, but apparently caused this mess in the process. In the interest of completeness, I'll point out that the bug still exists if you contrive to draw a sector with a width of less than half a fracunit...
Sorry for all the trouble that one caused.
 
Status
Not open for further replies.

Who is viewing this thread (Total: 1, Members: 0, Guests: 1)

Back
Top