EDGE-Classic v1.37 (Updated December 2023)

[Lobo]

Macro, any objections to us adding in your Heretic changes to the next EC?

[macro]

6 hours ago, dasho said:

 

Just so you know, we tracked down the culling issue. Looks like non-MSVC builds in general had the -Ofast optimization applied to it, which from reading up in the GCC manual on what that actually entails sounded pretty sketchy. That flag is now disabled in our CMakeLists and future MSYS/MinGW/GCC builds should no longer have that problem.

 

We do have options in the Gameplay Options menu that cover some of the usual major sticking points for people from a compatibility standpoint, like "True 3D Gameplay" (which infinitely tall actors is regulated by among other things), "Shoot-Through Scenery", Jumping/Crouching, and Autoaim.

 

As far as Strict Vanilla or Strict Boom or complevels in general, these are not goals for EDGE-Classic, and arguably weren't even goals for the original EDGE. I would not have even added the ability to toggle MBF21 lines and sectors On/Off except that some historical mapsets (including some IWAD content) erroneously have those bits set which accidentally cause behavior like certain lines becoming player-blocking if it is left On. The idea is that EDGE-Classic should be about EDGE content first and foremost, with "enough" compatibility that people can enjoy a large subset of the vanilla/Boom/MBF oriented PWADs that exist.

 

This is not to say that we don't work to improve our Boom/MBF/MBF21 features. For instance, last September saw the addition of Boom's Accelerative and Displacement scroller types, which was a huge leap in Boom compatibility as these are often used to trigger/"script" events.

 

I am happy you guys found the root of the problem. I am well aware that solving these small things is very important for EDGE-Classic.

 

As for complevel stuff, it was just an idea. I don't care too much either, just a thought or something. Because there is a thread here in Doomworld were someone built a huge map that needed to be strict Boom compatible. That map has dozens of doors with SR Door Stay Open but without sector tags. They work in gzdoom (also in EDGE, I already tested), but not in PrBoom, so this complevel idea just came to my mind.

 

What is MBF21, btw? Has it something to be with MBF source port? I don't even know exactly what UMAPINFO really is, besides some sort of LUMP with map information or something. But then, why that UMAPINFO thing is such a great deal?

 

 

1 hour ago, Lobo said:

Macro, any objections to us adding in your Heretic changes to the next EC?

 

Feel free to do it. But just in case, would you do some fast testing by yourself in case I missed something?

Also I am already working in the next patch that:

 

Spoiler

[FIXED/CHANGES - ]

- Further Adjusted Gargoyle dash attack

- Empowered Phoenix Rod added proper weapon attack
- Tomed Phoenix Rod added PRESERVE_MOMENTUM attack_special
- Necro Gauntlets added missing frames
- Adjusted Damage of all the weapons and attacks

- Fixed wrong scrolling direction in floor in MAP05, location Wings of Wrath secret area

 

[BUGS/TODO - ]

- Phoenix Rod Tomed Attack Cooldown State still need implementing.
- Phoenix Rod missing pushback effect with normal attack
- Fine Tune Phoenix Rod
- Review Weapon Mace
- Necro Gauntlets fix attack sound
- Fix Staff puff
- Check power-ups duration

- If all ammo is used up while using tome of power, if you get ammo again, Crystal Wand shoots normal projectiles, not powered ones

- Change Inventory Style to Heretic Style when using original Heretic Status Bar

 

This list of changes (that I am working on now and not published yet), includes the PRESERVE_MOMENTUM Attack_Special for the Phoenix Staff. Not sure if that EC that you mention will have such Attack_Special implemented.

As for what you see in the github above, everything should work with 1.34.1.

[Lobo]

UMAPINFO is basically just a way for prboom-type ports to have access to a subset of similar functionality to ports like EDGE or Zdoom (stuff that can be done via levels.ddf/games.ddf or ZMAPINFO/Decorate). It's a big deal for those types of ports because, up to now, they have never had the possibility of defining certain things(i.e. sky, music etc) on a map-by-map basis.

 

For your Heretic stuff, I'll wait till you feel it's ready before merging it into EC.
Idea: for the "Phoenix Rod missing pushback effect with normal attack": you can add something like 
PHNX:D:4:NORMAL:MOVE_FWD(-3) to the weapon ATTACK states to push the player back each time it's fired.

[macro]

3 minutes ago, Lobo said:

UMAPINFO is basically just a way for prboom-type ports to have access to a subset of similar functionality to ports like EDGE or Zdoom (stuff that can be done via levels.ddf/games.ddf or ZMAPINFO/Decorate). It's a big deal for those types of ports because, up to now, they have never had the possibility of defining certain things(i.e. sky, music etc) on a map-by-map basis.

 

For your Heretic stuff, I'll wait till you feel it's ready before merging it into EC.
Idea: for the "Phoenix Rod missing pushback effect with normal attack": you can add something like 
PHNX:D:4:NORMAL:MOVE_FWD(-3) to the weapon ATTACK states to push the player back each time it's fired.

 

About UMAPINFO, I see. Thanks for the info.

 

As for the MOVE_FWD(), I had no idea it was available from DDFWEAP! I will add that right away. I was trying to figure out some overcomplicated way.

 

Also, I just finished the Gauntlets, both normal as well as powered ones. Basically before, when you shoot them only the "hold fire" sound was played. I managed to make it an initial attack sound, and then a hold sound.

The way I did was simply to make the [GAUNTLETS] to play the initial attack sound when you fire (the explosion-like sound), then they BECOME a [GAUNTLETS_HOLD], and the attack sound is the normal holding sound (than kind of electric sound).

After the player releases the attack button, the [GAUNTLETS_HOLD] simply BECOME [GAUNTLETS].

 

Similar thing for tomed gauntlets.

 

Only left now is to make the tomed gauntlets to drain health from enemy and give to the player. 

After that, I will check the last weapon, I think it needs some fine tuning, then you can merge it and have all the weapons done as well as a few other fixes.

 

 

[Lobo]

The tomed gauntlets already do drain health from enemy and transfer to the player. It has the attack special VAMPIRE set.

Or I am misunderstanding you?

 

[macro]

8 hours ago, Lobo said:

The tomed gauntlets already do drain health from enemy and transfer to the player. It has the attack special VAMPIRE set.

Or I am misunderstanding you?

 

 

Yes you are correct. I didn't even notice about that.

[macro]

I completed all the weapons, except for the last weapon, the mace.

How urgent you need these patches? If the release of the next EC is immediate, I can upload everything that is completed so far. 

[dasho]

6 hours ago, macro said:

I completed all the weapons, except for the last weapon, the mace.

How urgent you need these patches? If the release of the next EC is immediate, I can upload everything that is completed so far. 

 

I wouldn't say it's incredibly urgent; been doing some bug-squashing lately and testing to make sure there isn't anything else that needs attention at the moment, other than what is already posted to our Issue tracker on the Github repo.

[macro]

I will try my best to work of the last weapon over this coming weekend.

That planned release, does it have the Preserve_Momentum already implemented? Sorry for the question, but I am not familiar with github at all.

[Lobo]

Yes, the next release will have  Preserve_momentum implemented.

 

Take your time: we've no rush and we are very grateful for your contributions.

[Redneckerz]

On 9/10/2023 at 12:02 PM, macro said:

What is MBF21, btw? Has it something to be with MBF source port? I don't even know exactly what UMAPINFO really is, besides some sort of LUMP with map information or something. But then, why that UMAPINFO thing is such a great deal?

 

  Reveal hidden contents

[FIXED/CHANGES - ]

- Further Adjusted Gargoyle dash attack

- Empowered Phoenix Rod added proper weapon attack
- Tomed Phoenix Rod added PRESERVE_MOMENTUM attack_special
- Necro Gauntlets added missing frames
- Adjusted Damage of all the weapons and attacks

- Fixed wrong scrolling direction in floor in MAP05, location Wings of Wrath secret area

 

[BUGS/TODO - ]

- Phoenix Rod Tomed Attack Cooldown State still need implementing.
- Phoenix Rod missing pushback effect with normal attack
- Fine Tune Phoenix Rod
- Review Weapon Mace
- Necro Gauntlets fix attack sound
- Fix Staff puff
- Check power-ups duration

- If all ammo is used up while using tome of power, if you get ammo again, Crystal Wand shoots normal projectiles, not powered ones

- Change Inventory Style to Heretic Style when using original Heretic Status Bar

 

Ole

[macro]

Regarding this topic:

 

 

Have you ever think about a version of EDGE-Classic that is designed with performance in mind?

 

I was testing a few things, and I noticed that Sunder 2407 MAP11 gets almost unplayable even in my PC that is much more powerful that the OP in that aforementioned thread. I have to set drawing distance to a maximum of 2900 or 3000 to make them playable.

 

I also noticed in MAP15 and 20, even with draw distance to a minimum, the game runs slow, but not like a constant low framerate, more like jumps between high framerate and low framerate, but not as if running out of RAM or VRAM where games get stuck for a fraction of a second. It is more like just jumping between framerates every second or so. I have 12Gb of RAM, so memory shouldn't be an issue.

 

But interestingly enough, these same maps in PrBoom+ or DSDA-doom runs really well, with some light fps drops here and there, but 90% of the time at 30+ fps. Also zDoom runs quite reasonable, although about 20 fps when many actors and missiles are around.

 

But if you play Sunder 2407 with -nomonsters flag, any map will run fine.

 

I have no idea what could be the reason behind this, so just sharing what I found.

[dasho]

31 minutes ago, macro said:

Have you ever think about a version of EDGE-Classic that is designed with performance in mind?

 

I was testing a few things, and I noticed that Sunder 2407 MAP11 gets almost unplayable even in my PC that is much more powerful that the OP in that aforementioned thread. I have to set drawing distance to a maximum of 2900 or 3000 to make them playable.

 

I also noticed in MAP15 and 20, even with draw distance to a minimum, the game runs slow, but not like a constant low framerate, more like jumps between high framerate and low framerate, but not as if running out of RAM or VRAM where games get stuck for a fraction of a second. It is more like just jumping between framerates every second or so. I have 12Gb of RAM, so memory shouldn't be an issue.

 

But interestingly enough, these same maps in PrBoom+ or DSDA-doom runs really well, with some light fps drops here and there, but 90% of the time at 30+ fps. Also zDoom runs quite reasonable, although about 20 fps when many actors and missiles are around.

 

But if you play Sunder 2407 with -nomonsters flag, any map will run fine.

 

I have no idea what could be the reason behind this, so just sharing what I found.

 

As I (partially) alluded to in the thread you mention, BSP traversal and using OpenGL 1.x/sending all of the geometry every frame is the cause of this. Draw distance culling won't help if there are enough lines in the non-culled view that traversal chokes up; this is probably accompanied by a lot of small fixed-function GL calls happening in rapid succession which will compound the problem. The likely cause of your framerate peaks and valleys that you describe is that you are looking at geometry with varying degrees of complexity. 

 

EDGE-Classic will always be designed with performance in mind, but the engines you mention have also had over 10 years of active development since the last release of EDGE which became our codebase. The changes needed to improve on maps that you describe will take massive changes to the renderer, and will not be quick by any stretch of the imagination.

 

EDIT: I will also add that I personally don't care much about performance on maps like Sunder, except as a stress-test or benchmarking tool.

Edited September 14, 2023 by dasho

[Redneckerz]

If your gpu is OGL 3 capable, the best answer would be Helion - Sunder MAP15 is a cakewalk when it comes to perf.

[Scyon]

Hi.

 

I am not sure if this question will go here, but if not please move or delete it.

 

I am creating a project for EDGE-Classic but I have a question. One of the maps has two switches that must be activated to lower the floor. The way I wanted to do this is by activating a certain RTS script that increases COUNTER01 by +1, then check if COUNTER01 is 2 or more and if yes, then lower the floor.

 

The script is something like this:

 

RADIUS_TRIGGER 0 0 -1
        TAG 1
        TAGGED_DISABLED
        
        Give_Benefit COUNTER01(1)

        On_Condition COUNTER01(2)
            Lose_Benefit COUNTER01(2)
            Enable_Tagged 2
END_RADIUS_TRIGGER

Script number 2 (enable_tagged 2) is the script that lowers the floor, but there is no particular issue with that.

 

Apparently this simple script should work but the issue that I have are several. Both switches linedef actions are set as action 418 S1 Activate EDGE RTS. The first issue is that when I press either of the switches, the script doesn't seem to run. I added a simple TIP "HELLO" right before Give_Benefit, but it doesn't seem to work.

If I remove the ON_CONDITION stuff, the script runs, but there is another issue that is when I press switch number 1, then switch number 2 automatically changes its texture to activated. If I press switch 2 that changed to activated texture, it changes texture to back deactivated but the script doesn't run.

 

So far the only solution that I could find is to have one script for each button to increase COUNTER01 and then another script with just a ON_CONDITION stuff.

But ideally I would like to have one single script handling both things.

 

At this point I am not sure if this is a bug or it is how it should be, but in either case I would like to know if I am doing something wrong.

[mmx]

Are there any plans of adding multiplayer support to the WebGL build? Would be great for those non-tech savvy friends.

[Lobo]

3 hours ago, mmx said:

Are there any plans of adding multiplayer support to the WebGL build? Would be great for those non-tech savvy friends.

I'm afraid multiplayer support is a long way off at the moment.

It's always at the back of our heads, but there is so much other stuff to do first :(

[Lobo]

@Scyon RTS scripts are processed all at once by EDGE. So, as you have seen, if you have an ON_CONDITION in your trigger then the whole trigger depends on that condition. If you have multiple ON_CONDITION in the trigger then they ALL must be satisfied even if you have other code placed between them.

 

RADIUS_TRIGGER 0 0 -1
        TAG 1
        TAGGED_DISABLED
        
        Give_Benefit COUNTER01(1)

        On_Condition COUNTER01(2)
            Lose_Benefit COUNTER01(2)
            Enable_Tagged 2
END_RADIUS_TRIGGER

This trigger will not run until activated via the TAG(i.e. a switch linedef).
Even then, once it has been called, it will ONLY run if COUNTER01 is 1 or greater. It will NOT do the GIVE_BENEFIT COUNTER01(1) and then continue to the ON_CONDITION.

 

From what you describe, the correct approach is this:
"... is to have one script for each button to increase COUNTER01 and then another script with just an ON_CONDITION stuff."

Edited September 22, 2023 by Lobo

[Lobo]

Just dropped a new release; it started off as a minor/bugfix type of release but gradually snowballed into a new version as more features were added. The full changelog can be found on the latest release page on Github, but here's a TL;DR:

- "None" option added for Blood Level; will disable blood splats, crush/death states, and will teleport enemies away on death a la Chex Quest
- Complete overhaul of faux skybox drawing: New options are Mirror, Repeat, Stretch, and Vanilla
- Sector-specific and level-wide indoor/outdoor GL fog options have been added
- New line specials to automatically stretch the height or width of their texture to match their length
- Wall glows are now implemented (were previously referenced in the engine but were no-ops)
- UDMF spec updated with the following supported fields:
  - linedefs: blockplayers, blocksight
  - sidedefs: scalex_top/mid/bottom, scaley_top/mid/bottom, offsetx_top/mid/bottom,
offsety_top/mid/bottom
  - sectors: xpanningfloor/ceiling, ypanningfloor/ceiling, xscalefloor/ceiling,
yscalefloor/ceiling, gravity, fadecolor, fogdensity
  - things: health, alpha, scalex, scaley, scale
- Game Music Emu replaced with libVGM: All but one chipset (Wonderswan) supported by the VGM format will now work
- Better support for Unicode characters from international locales in the program's path
- Many bugfixes, including some still remaining from legacy EDGE

[Lobo]

@macroNo pressure on the Heretic stuff: we can add it for the next release so take your time. We just really had to get this version out as it has some important bug fixes and performance improvements.

[Scyon]

17 hours ago, Lobo said:

@Scyon RTS scripts are processed all at once by EDGE. So, as you have seen, if you have an ON_CONDITION in your trigger then the whole trigger depends on that condition. If you have multiple ON_CONDITION in the trigger then they ALL must be satisfied even if you have other code placed between them.

 

RADIUS_TRIGGER 0 0 -1
        TAG 1
        TAGGED_DISABLED
        
        Give_Benefit COUNTER01(1)

        On_Condition COUNTER01(2)
            Lose_Benefit COUNTER01(2)
            Enable_Tagged 2
END_RADIUS_TRIGGER

This trigger will not run until activated via the TAG(i.e. a switch linedef).
Even then, once it has been called, it will ONLY run if COUNTER01 is 1 or greater. It will NOT do the GIVE_BENEFIT COUNTER01(1) and then continue to the ON_CONDITION.

 

From what you describe, the correct approach is this:
"... is to have one script for each button to increase COUNTER01 and then another script with just an ON_CONDITION stuff."

 

I see. So basically it doesn't matter the order of statements inside each script. 

 

But that doesn't explain why when I activate one of the buttons, the other button changes its texture to activated as well, and after that if I press that second button, it reverts the texture back to deactivated but the script doesn't run anymore in, so only runs once. 

 

This happens if I make both buttons to execute the same script. 

[Lobo]

Can you post the wad here so we can take a look?

If you don't set the trigger as 

"tagged_repeatable"

then it will only run once.

[Lobo]

Here is an example wad of what I think you want to do.

scyon.zip

[Scyon]

On 9/23/2023 at 2:48 PM, Lobo said:

Can you post the wad here so we can take a look?

If you don't set the trigger as 

"tagged_repeatable"

then it will only run once.

 

I just noticed your reply now. I am sorry, but I cannot share the wad.

 

I checked your example and that is what I currently have.

What I don't understand is why if both switches activate the same script, when the player uses one of the switches the other switch is also "used", so the texture changes although the script runs only once as expected. The usual behavior should be that only the switch that is used by the player changes texture.

Then, if the player uses any of the switches again, both of them change their texture to deactivated.

 

I will attach a modified example because it might be difficult to understand just with words.

scyon.zip

[Lobo]

The decision was made a very long time ago in the EDGE codebase to, by default, switch all switches that have the same tag and special.

It's not vanilla behavior but it's been that way since 1998 so we will not change it so as not break any EDGE maps that rely on that behaviour.

 

However, there is also a SWITCH_SEPARATE special flag for DDFLINE for the exact purpose of not having all lines with the same tag switch when one is pressed.

[dasho]

I'm happy to announce that the latest release of EDGE-Classic, 1.36, is now available on our Releases page.

 

I'll post the full changes below, but it is likely worth clearing out your /cache folder and letting nodes be rebuilt since we made a small fix to AJBSP's 'fast mode' causing some potential visual oddities (it had a chance of selecting self-referencing lines for partitioning). 

 

In addition, it might also be worth starting with a clean config since there are some quality-of-life improvements regarding gamepads. We have migrated from legacy SDL_Joystick usage to the modern SDL_GameController functions. This means user-friendly button names, automatic trigger detection, better hotplugging detection/handling, and default gamepad bindings.

 

New Features:

Spoiler

- ATTACKS.DDF: 
  - DAMAGE_UNLESS_BENEFIT and DAMAGE_IF_BENEFIT subcommands added
    - Can be used anywhere that DAMAGE is used (sectors, attacks, etc)
    - DAMAGE.DAMAGE_UNLESS_BENEFIT: Player will take damage unless they have at least one of the benefits listed
    - DAMAGE.DAMAGE_IF_BENEFIT: Player will take damage if they have at least one of the benefits listed
  - Attack special "PLAYER_ATTACK" makes attack behave as if originating from the player, even if 
    it's an indirect secondary attack i.e. the player gets his VAMPIRE health and KillBenefits.
  - DAMAGE.FLASH_COLOUR field added (Courtesy of akaAgar - https://github.com/edge-classic/EDGE-classic/pull/576)
    - Allows a custom color to be used as the pain effect when a player is hit by this damage type
    - Can be used to emulate effects like Hexen's cold/poison damage

- WEAPONS.DDF: 
  - Can now define a 3RD and 4TH attack. Equivalent DDF syntax to SEC_xxx used for second attack
    - DDF Fields: 
      - 3RD_ATTACK, 3RD_AMMOTYPE, 3RD_AMMOPERSHOT, 3RD_CLIPSIZE, 3RD_AUTOMATIC, 3RD_SPECIAL
      - 4TH_ATTACK, 4TH_AMMOTYPE, 4TH_AMMOPERSHOT, 4TH_CLIPSIZE, 4TH_AUTOMATIC, 4TH_SPECIAL
    - DDF States:
      - 3RD_ATTACK, 3RD_RELOAD, 3RD_DISCARD, 3RD_WARMUP, 3RD_FLASH
      - 4TH_ATTACK, 4TH_RELOAD, 4TH_DISCARD, 4TH_WARMUP, 4TH_FLASH
    - DDF Actions:
      - 3RD_SHOOT, 3RD_REFIRE, 3RD_NOFIRE, 3RD_NOFIRE_RETURN, 3RD_CHECKRELOAD, 3RD_FLASH
      - 4TH_SHOOT, 4TH_REFIRE, 4TH_NOFIRE, 4TH_NOFIRE_RETURN, 4TH_CHECKRELOAD, 4TH_FLASH
    - For consistency, all the old SEC_xxx ddf commands can also be used with 2ND_xxx instead 
      - i.e. 2ND_AMMOTYPE can replace SEC_AMMOTYPE
  - REFIRE_TO frame command added
    - Uses same syntax as JUMP, but uses refiring instead of % chance as the test condition
    - SEC/2ND_REFIRE_TO, 3RD_REFIRE_TO, and 4TH_REFIRE_TO also supported

- GAMES.DDF: "NO_SKILL_MENU" boolean command added (Courtesy of akaAgar - https://github.com/edge-classic/EDGE-classic/pull/577)
  - If true, will skip difficulty menu and start selected episode at skill 2 ("Hurt Me Plenty" for Doom)
  - Intent is for mapsets that have a Quake-style initial skill-select level or where the author may not implement/want
    to present skill levels
  - Default value is false

- RSCRIPT: "EXACT_" prefix added for ON_CONDITION checks (Courtesy of akaAgar - https://github.com/edge-classic/EDGE-classic/pull/578)
  - Allows check for the exact value of a benefit/etc, versus the default of greater-than-or-equal

- THINGS.DDF: 
  - New Action command: SET_PAINCHANCE(percentage) which will dynamically change a monsters painchance.
    - Possible usage: if we do SET_PAINCHANCE(0%) at the start of a lengthy PAIN state, this will make sure the 
        whole animation is played out and won't be interrupted by another shot which would normally make us 
        re-start the pain animation again. Remember to set it back to normal at the end of the lengthy animation.
  - New Special flag: IMMOVABLE which will make this thing be unaffected by thrust from attack impacts 
    irrespective of whether it has a small MASS or a very large one.

- UDMF:
  - Sectors: Support added for the `lightcolor`, `rotationfloor`, and `rotationceiling` keys

 

General Improvements/Changes:

Spoiler

- Detection/functionality for mikoportals implemented
  - Voodoo dolls that fall onto a floor at exactly -32768 will preserve their X/Y/Z momentum and be moved to their sector's ceiling
    - Friction and drag ("air friction") will not be applied to voodoo dolls in this situation
  - Regular players and things will adhere to normal physics
- Windowed Mode resolution list will now dynamically populate based off of available display information instead of using presets
- Video Options crosshair selection will not be shown when switching to/from weapons that have DDF custom crosshairs
- Slightly re-organized the Key binding menu to accomodate the new 3rd and 4th attacks.
- Implemented 'episode' UMAPINFO key; EDGE-Classic now fully supports the UMAPINFO spec, Revision 2.2
- ZMAPINFO/MAPINFO parsing updates:
  - Episode definitions and the 'clearepisodes' directive will now be respected
  - ZMAPINFO 'ResetInventory' level definition key will result in the RESET_PLAYER DDFLEVL command being applied
- Heretic: numerous improvements and tweaks courtesy of DW user Macro (https://github.com/Macro914/EDGE-Classic-Scripting-Patches)
- Heretic: improved Mace attack courtesy of CeeJay
- YMFM:
  - Added "Apogee" instrument bank, based on Apogee Sound System and AdPlug instruments
  - Removed "AdPlug", "WadSmoosh", and "DMXOPL3" instrument banks
  - Replaced "DMXOPL" instrument bank with one tailored to vanilla/single OPL3 chip limits
  - Removed ability to load WOPL instrument banks as it was only partially implemented
- COLMAP.DDF: Colormap entries with no gl_colour, lump, or pack file defined will no longer throw an error
  - New behavior is to remove all existing colormap entries with the same name
    - Intent is to be able to define a colormap as "nothing", even if prior entries had a definition
  - Colormap entries later in the load/parsing order with the same name will still be added
- Improved level warp cheat behavior with custom mapnames
  - Previous behavior was to only match a numerical entry with maps named either MAPXX or EXMX
- Improved handling of gamepads
  - Automatic detection of analog triggers
  - User-friendly button names in menus ("A Button" versus "Joy 1", etc)
  - Menus can now be navigated via gamepad without needing to set explicit bindings
    - "Menu Navigation" section removed from Key Bindings menu
  - Added default bindings for gamepads
- Added configuration files for use with Doom Builder.X, Ultimate Doom Builder, and Eureka
  - Can be located in the /docs/mapping directory

 

Bugs Fixed:

Spoiler

- Legacy bugfix: Boom generalized sector types were not preserved when saving/loading levels with said sectors
- Fixed assertion failure when referencing DSDehacked frames past the last frame explicitly defined in the [CODEPTR] section
- Fixed viewpoint modified by the r_fov CVAR not changing the vertical view slope appropriately
- Fixed region properties not working properly with a vertex sloped floor that rises above its sector's original floor height
- Fixed occasional sector glow thing link/unlink errors
- Fixed AJBSP using self-referencing linedefs that were perfectly vertical or horizontal for partitioning
- Fixed floor/ceiling rotation not working when using images that have a different width and height
- Fixed MODEL_ROTATE not being applied to attacks
- Fixed savegames not preserving the 'is_voodoo' convenience boolean for mobjs, causing issues when loading a game that had them present
- Fixed berserk and other powerup effects not modifying model colors accordingly (bug introduced by transition to VBOs for model rendering)
- Fixed crash if player dies while TARGET_JUMP or FRIEND_JUMP weapon specials are active

 

[Lobo]

Updated FP.

[dasho]

Our example web player (https://pbdot.github.io/edge-classic-site-webplayer/) has also been updated to be in sync with the latest release. It is probably a good idea to clear cache/force reload of the site, although this will result in fetching the Freedoom IWAD again on the first run.

[RataUnderground]

There is one EDGE issue that I find rather uncomfortable, so I thought it was a good idea to expose it here.
(Note in advance that my operating system is Windows).
I play on a setup with multiple screens and often want to switch applications between screens. This on most sourceports is simple, I just press the windows key and then drag the window to the screen I want. However, no key combination seems to work in the case of Edge, neither the windows key nor alt-enter to toggle between window and fullscreen. I have to press ctrl alt del and then I can finally move the cursor out of EDGE.
I think it would be a nice quality of life improvement if this was taken into account.

[dasho]

1 hour ago, RataUnderground said:

There is one EDGE issue that I find rather uncomfortable, so I thought it was a good idea to expose it here.
(Note in advance that my operating system is Windows).
I play on a setup with multiple screens and often want to switch applications between screens. This on most sourceports is simple, I just press the windows key and then drag the window to the screen I want. However, no key combination seems to work in the case of Edge, neither the windows key nor alt-enter to toggle between window and fullscreen. I have to press ctrl alt del and then I can finally move the cursor out of EDGE.
I think it would be a nice quality of life improvement if this was taken into account.

 

When I press the Windows key, the Start Menu appears and the mouse cursor is disengaged from the program. Alt-Tabbing also works as expected, after which I also can use the mouse freely. Are you saying that you cannot regain mouse control without using Ctrl-Alt-Del?

 

As far as Alt-Enter, we could work on adding some handling for that.