Any modern RMB (reject map builder) derivatives?

[Grassy chunks]

Scouring the doomwiki, i came accross the RMB page where I saw that the program was not updated in over 20+ years. And also coming accross the map 02 of requiem, and seeing it's blind demon done through a custom reject table; i really wanted to build a custom reject table myself.

The problem is that the program is 16bit and i also tried running it in dosbox but with no avail..

 

So, anybody that can recommend me it's modern derivative for use with newer ports???

Or there are none out here.

https://doomwiki.org/wiki/RMB

[Doomy__Doom]

Reject table building is typically a nodebuilder capability nowadays. For manual editing of the already existing lump there's always slade.

[Grassy chunks]

1 minute ago, Doomy__Doom said:

Reject table building is typically a nodebuilder capability nowadays. For manual editing of the already existing lump there's always slade.

But the folks that made requiem probably generated the reject table for map 02 manually, using RMB. And if you could tell me how could I recreate the blind pinky demon in that map using Slade then I would be very much benefited.

 

Speaking of that, opening any map in Slade and viewing it's reject lump there in text mode shows an empty textfield.

 

[Doomy__Doom]

Remind me to never go for quick replies on work breaks, never works out :) You can export the lump from slade, hex-edit it and re-import. Honestly idk if there's a convinient calculator of which bytes you'll need.

[Grassy chunks]

2 minutes ago, Doomy__Doom said:

Remind me to never go for quick replies on work breaks, never works out :) You can export the lump from slade, hex-edit it and re-import. Honestly idk if there's a convinient calculator of which bytes you'll need.

Thanks for your reply, but, I'm not that smart one who knows how to hex edit anything. I only know python and Java, no c/c++.

[Professor Hastig]

Looking at that wiki page already gives the reason why there's no modern versions: RMB has never seen a source release. And with the algorithm essentially being a trade secret it is not surprising that it saw little work in later years.

 

[Grassy chunks]

29 minutes ago, Professor Hastig said:

Looking at that wiki page already gives the reason why there's no modern versions: RMB has never seen a source release. And with the algorithm essentially being a trade secret it is not surprising that it saw little work in later years.

 

What a real bummer. The RMB 3.0 page at idgames archive dating to 1996 also has very low rating so I guess that one probably doesn't work anymore.

 

Edit:Also, someone here who is knowledgeable in hex editing/programming maybe could guide me how to do it nowadays.

Edited February 6, 2023 by Grassy chunks

[Hitherto]

3 hours ago, Grassy chunks said:

The problem is that the program is 16bit and i also tried running it in dosbox but with no avail..

It SHOULD work in DOSbox. For me it works even in winXP cmdline.

 

Here is another problem exist, all this old stuff can't work with really large modern levels. Few day ago I was trying to build reject table for https://www.doomworld.com/forum/topic/128344-seriouslywad-mbf21-re-release/

m31 but zennode can't handle it.

 

Doomy__Doom, do you know about such capable nodebuilders? I know there should be something, because @Ribbiks for example do it for cosmogenesis maps.

 

[Claymore]

REJECT usually gets generated by node builders nowadays, some do it, some ignore it, however the closest you can get to building a custom REJECT table is by feeding RMB options to ZenNode.

It doesn't support all the options, but afaik it's the best you got without manually hex editing the lump in SLADE.

[Grassy chunks]

10 hours ago, Claymore said:

REJECT usually gets generated by node builders nowadays, some do it, some ignore it, however the closest you can get to building a custom REJECT table is by feeding RMB options to ZenNode.

It doesn't support all the options, but afaik it's the best you got without manually hex editing the lump in SLADE.

Sounds great, could you maybe elaborate more on how could I feed RMB option in zennode, maybe even provide an example wad. Cause, I'm having a hard time understanding how to go about this....

[Hitherto]

RTFM for RMB. Also there some example wads.

[Redneckerz]

12 hours ago, Hitherto said:

I know there should be something, because @Ribbiks for example do it for cosmogenesis maps.

 

While the concept is understood, its spec is less defined.

 

Ribbiks afaik uses a custom made reject generator script. I remember a post where it was said there was a download, but i couldnt find it. (The DL, not the post)

 

Rejects are still useful these days for extreme maps.

[Grassy chunks]

15 minutes ago, Hitherto said:

RTFM for RMB. Also there some example wads.

What's rtfm?

[Hitherto]

Nobody can Read That Fucking Manual for you. Except yourself.

Edited February 7, 2023 by Hitherto

[Redneckerz]

@Hitherto @Grassy chunks here is that post i was referring to. Likely @Ribbiks uses LUA for the reject script.

 

Sadly @Insane_Gazebo's SU15_MakeMeATable.Zip (Which is the SU15 REJECT table without a wall) link is dead, but possibly that link had a template to make a REJECT table yourself. I am tagging the aforementioned users so they may see this thread.

 

FWIW, more options for Reject table generation are always welcome.

[RjY]

On 2/6/2023 at 2:55 PM, Claymore said:

REJECT usually gets generated by node builders nowadays, some do it, some ignore it, however the closest you can get to building a custom REJECT table is by feeding RMB options to ZenNode.

 

On 2/7/2023 at 1:23 AM, Grassy chunks said:

Sounds great, could you maybe elaborate more on how could I feed RMB option in zennode, maybe even provide an example wad. Cause, I'm having a hard time understanding how to go about this....

 

It's not your fault. You most likely did everything correctly. The problem is, there is a vital piece missing from the code to support RMB options in ZenNode. It parses the instructions, but does not pass the results to the reject builder. So it looks like it is going to work, but then silently fails to do anything useful.
 

The link has been restored in ZokumBSP, but you will need a build of the latest development version to use it.
 

On 2/6/2023 at 1:16 PM, Hitherto said:

Here is another problem exist, all this old stuff can't work with really large modern levels. Few day ago I was trying to build reject table for Seriously.wad m31 but zennode can't handle it.

 

FWIW the aforementioned latest ZokumBSP has some improvements to allow building reject tables on large modern maps. (At least, it can build Sunder 15 now, which was the point of the exercise.)

[zokum]

On 2/9/2023 at 11:19 PM, RjY said:

 

 

It's not your fault. You most likely did everything correctly. The problem is, there is a vital piece missing from the code to support RMB options in ZenNode. It parses the instructions, but does not pass the results to the reject builder. So it looks like it is going to work, but then silently fails to do anything useful.
 

The link has been restored in ZokumBSP, but you will need a build of the latest development version to use it.
 

FWIW the aforementioned latest ZokumBSP has some improvements to allow building reject tables on large modern maps. (At least, it can build Sunder 15 now, which was the point of the exercise.)

The code in Zennode has a variable that will overflow and cause problems. I raised it from 32bit to 64bit to fix this. This is most likely why it can build some bigger maps that earleir versions or Zennode failed on.

The other problem is that it uses a doom blockmap internally to generate the reject map, and this limits the size of the map. This causes problems on maps that are near the limit on the blockmap since the reject code needs a more complete blockmap for its generation, which will go over the limit. This is fixable though, and some work has been done a few years ago to facilitate this. Code was added to increase it from a 16bit signed blockmap to a 32bit blockamp, but it isn't finished. That would increase the practical internal blockmap limit from about 65k to over 4gig. With this limit removed, it might run into other limits when building very large maps.

And yes, Zennode/ZokumBSP generates the blockmap twice. Once for the blockmap entry in the map, then once more if you plan to build a reject entry...

Edited February 16, 2023 by zokum
less screaming, more clarification