[Ninjakitty]

Wait... it hasn't even been a month yet? Wow... It sure felt like a lot longer, but perhaps that's just because I've actually been doing stuff lately.
LOL

[Master O]

Maybe we should get that Giant Sumo from Shadow Warrior to sit on him.

[MusicallyInspired]

Ah give him a break. It's not like there's any real standard of consistency in any of us that he has to measure up to. Realistically, he's one of the only ones who's done the most work here in a while (not counting the VoidPoint team).

[Kyanos]

Drek, on 06 August 2017 - 09:21 AM, said:

That's a chicken or egg type question. I read it as first a new port, then a new episode, with neither actually happening

I was just kidding, thought a bit of goading would encourage him to not bail this time...

[Player Lin]

Drek, on 21 September 2017 - 11:21 AM, said:

I was just kidding, thought a bit of goading would encourage him to not bail this time...

I would say the Real-Life™ just get ICD for sure, but it's still too early to give up.

[Romulus]

[Tea Monster]

When it's been more than a year since you've heard from him, then you read on a forum post that he's working on another project all together, THEN you can give up ;-)

[Kyanos]

What state is this level editor in? Anyone use it much?

[Mark]

Back when ICD was still active I placed a few SE lights and props for testing in a map. That was about it. I think it was far enough along to actually be useful.

[RPD Guy]

Very nice!!!

[Hendricks266]

Ice's editor can never be "done" because it's a hacked up Duke editor, not an SW editor. It's an entirely wrong path to go down, and as usual he rejected the sense I tried to talk into him.

[Kyanos]

Hendricks266, on 06 October 2017 - 10:30 AM, said:

Ice's editor can never be "done" because it's a hacked up Duke editor, not an SW editor. It's an entirely wrong path to go down, and as usual he rejected the sense I tried to talk into him.

Can you tell me some example cases that hold back usability?

Also this.

Helixhorned, on 14 April 2011 - 12:20 PM, said:

All you need to do is to copy/link the SW.GRP to DUKE3D.GRP, since that is what Mapster32 is looking for. All necessary files like palettes should be then loaded from the group file. The only issues I see with using Mapster32 to edit SW maps are minor graphical ones (e.g. text colored differently, parallaxed skies not displaying properly).

[Hendricks266]

Map scale

Forcing of duke3d-specific stuff like sprite animations, and the absence of any SW ones

Lack of support for the SW tag editing keys (?)

Total absence of SW alt-pals since they are defined in the source

No ability to preview ROR

Have you looked through SW's bstub.c to see what else you would be missing?

[Kyanos]

The two big ones then are ROR preview and map scale. I made a palette, the animations are simple/tedious changes and the tagging system did get some work done to it, not sure how far ice got with it. As for bstub.c, yes I've read it over and the largest difference is the tagging system, I didn't notice anything that isn't on your list.

[Kyanos]

I just finished putting the ST1 effector names in, this old version of Mapster32 I'm tweaking is from before TROR where the codebase is closer to SWPbuild so I've just sort of ported stuff from there and got a good working version of what I call here SWMapster32. Voxels are in and pal swaps work, parallax skies look a bit off but the last big things on my todo list are to get ambient sounds playing and floor over floor preview hooked up, it's there already but set to duke3d's se40 sprite.

[Hendricks266]

Drek, on 10 October 2017 - 01:45 PM, said:

this old version of Mapster32 I'm tweaking is from before TROR

Using such an old version of the codebase defeats the purpose of your work.

[Forge]

Hendricks266 said:

.......

[Kyanos]

Hendricks266, on 10 October 2017 - 08:37 PM, said:

Using such an old version of the codebase defeats the purpose of your work.

Yes and no. It's like 6 in one a third of a dozen in the other.

I stopped working on it when icd put out his editor, but since there is no other alternative at the moment I decided it was still worth the effort. The true purpose of my work was to learn, so if a semi useful editor comes out of it then that's a bonus.

[Hendricks266]

Drek, on 11 October 2017 - 12:42 AM, said:

The true purpose of my work was to learn, so if a semi useful editor comes out of it then that's a bonus.

Cool, I can get behind that.

[Kyanos]

Put together a help file, SW is much more complex than Duke3d in the tagging of actors, effects, tracks... lots of reading.

[Kyanos]

Another update, then the next time I post about this I'll start a thread and include a download.

I chopped out duke3ds names.h and any hardcoded duke specific sprite behavior is now gone. Ambient sounds are playing in 3d mode and showing their radius in 2d mode. It's odd in SW how the actual sound numbers and ambient sound tags are different and overlap, I'm currently writing something to handle that so that F2 still can load all the SFX for sampling (as of now I just used the ambient sounds only and loaded them by those numbers.)

[Hank]

Drek, on 17 October 2017 - 02:23 PM, said:

Another update, then the next time I post about this I'll start a thread and include a download.

I'm like this again.


Looking forward to what ever you are willing to do for Shadow Warrior

[enderandrew]

Hank, on 06 August 2017 - 09:26 AM, said:

@ Nancsi - It started as an intended Total Conversion using Shadow Warrior and Eduke32 codes without worrying about the licenses nightmares involved, and now seems to become a port first and foremost, with a lot of hard coding work ahead.

I'm 'watching' it on Github, which sends me updates when ever they are available. FMOD is gone in favour of OpenAl and am waiting what 3d model format will be implemented.

Bottom line, only icecoldduke knows where his project will go to.

SW and Eduke32 both have pretty permissive licenses. There shouldn't be license nightmares combining those. In fact, it seem the initial decision to do that and the heavy lifting there was @Hendricks266.

I read through every page of this thread even though I'm a copy/paste coder who rarely does any development. It seems to me in all the threads I read from ICD that he wants to jump in and make something. So he isn't afraid to experiment or rewrite things, which usually leads to consternation from Evan, who doesn't want his code changed or tinkered with. FWIW, reading over all these threads it seems like we should appreciate new work being done and it helps to be open to new ideas, but sometimes ICD runs into issues because he doesn't know the existing code base as well as others.

I see Evan constantly claiming that ICD refuses to listen to any feedback, and then ICD begging for specific feedback. FWIW, I've also see ICD compromise or make changes given feedback. But literally never have I seen Hendricks266 be open to accepting any code from ICD. This impass saddens me. I get that ICD hasn't finished the big rocks he started, but his repositories are open and there has been clear progress.

I wish there was more specifics given for the technical differences between the sides so we can end up with a usable SW port.

As for why ICD is bouncing forth between new features in this port (when some people just want a basic SW port that plays the base game and 2 expansions to start) is precisely because ultimately ICD wants to make a mod for a new episode. I don't see this as a contradiction or a shift in direction. The base game can't run the mod he wants. A straight source port can't either. Playing with new features informs what he can and cannot do with a mod. He also split his repository so that one branch is more of a straight port, and other stuff is separated as new features for the mod.

FWIW, I do often fall on ICD's side of thinking in that bringing in a separately well maintained library for a function just makes sense. As for the lengthy argument over which model format to support, I keep wondering:



Let modders use whatever works for them.

I hope some of the negativity hasn't chased off ICD and that his work will continue in time. He has disappeared before, and come back again when time allowed.

[Hendricks266]

I don't have the time to thoroughly debunk all the things you piled on in your post. It's not about me, it's about the quality of his code.

enderandrew, on 31 October 2017 - 07:52 PM, said:

I read through every page of this thread even though I'm a copy/paste coder who rarely does any development.

This is exactly the problem. Ice announces things that sound cool on the surface, and of course the common user is going to react positively to basically anything someone could promise in this position, no matter its merit. Apparently I am the only one willing to speak up about the significant costs and consequences of basically every decision that he makes. So even though Ice promises something shiny, what you're really seeing is a liquid mercury coating on top of shit.

[Phredreeke]

One thing that puzzled me was his work on getting Polymer to work with Shadow Warrior while at the same time not planning to add voxel support to it. Maybe it's just me that considers voxels a defining feature of the game.

(also what are the problems with combining SW and eduke32? I thought 3D Realms already dealt with getting GPL exceptions for it)

[Jblade]

Quote

This is exactly the problem. Ice announces things that sound cool on the surface, and of course the common user is going to react positively to basically anything someone could promise in this position, no matter its merit. Apparently I am the only one willing to speak up about the significant costs and consequences of basically every decision that he makes. So even though Ice promises something shiny, what you're really seeing is a liquid mercury coating on top of shit.

This and he seemed to start/drop projects and change what he's working on all the time.

[enderandrew]

Hendricks266, on 31 October 2017 - 09:06 PM, said:

This is exactly the problem. Ice announces things that sound cool on the surface, and of course the common user is going to react positively to basically anything someone could promise in this position, no matter its merit. Apparently I am the only one willing to speak up about the significant costs and consequences of basically every decision that he makes. So even though Ice promises something shiny, what you're really seeing is a liquid mercury coating on top of shit.

I don't have to know the technical details to understand ICD asking for help and feedback and seeing you respond that you're unwilling to provide specific criticism, only to shit on him in a generic sense. How is that productive?

Occasionally you do point out specifics and then he is able to respond in kind and adjust his code.

Presumably you'll want a better SW port for your Deadly Kiss project as well. ICD has developed clear bug fixes for the base port and separated out his new features into a new branch. Doesn't it make sense to try and find a way to work in those fixes first?

When ICD first started to tackle getting SW running in Eduke32, you reacted with clear anger and derision that someone even wanted to tackle that. You made an initial effort yourself. You could have explained the lessons you learned and the technical hurdles that prevented you from finishing the project. Others jumped in with criticism of how ICD did his Lua scripting implementation. ICD listened to the criticism and changed his code accordingly, because the criticism was specific and actionable.

Offering specific and actionable criticism is productive.

Telling someone everything they do is terrible and will never be accepted drives away community contributions and is not productive. Even if you are right from a technical perspective, the behavior doesn't make sense.

Jblade, on 01 November 2017 - 01:55 AM, said:

This and he seemed to start/drop projects and change what he's working on all the time.

Several people in the community seem angry at him for not finishing projects, but these are big rocks and will take time. If he disappears for months in between because of his day job, I think that is fairly understandable. I'm not sure his projects are fully abandoned yet, though the way the community treats him certainly isn't encouraging if people want to see them completed.

He started on a new renderer which showed great progress (though I personally would have targeted Vulkan). While this project isn't finished, it does compile and run. You can test the current renderer. He also merged that code into the SW port presumably because he plans to finish it.

[Hendricks266]

Phredreeke, on 31 October 2017 - 10:01 PM, said:

(also what are the problems with combining SW and eduke32? I thought 3D Realms already dealt with getting GPL exceptions for it)

There are no licensing problems.

enderandrew, on 01 November 2017 - 06:41 AM, said:

I don't have to know the technical details to understand ICD asking for help and feedback and seeing you respond that you're unwilling to provide specific criticism, only to shit on him in a generic sense. How is that productive?

You may have read this thread, but have you also read the PolymerNG thread? I stopped trying to give specific criticism, because every conversation went like this:

me: <long, in-depth post explaining the situation>
ice: no

enderandrew, on 01 November 2017 - 06:41 AM, said:

Presumably you'll want a better SW port for your Deadly Kiss project as well. ICD has developed clear bug fixes for the base port and separated out his new features into a new branch. Doesn't it make sense to try and find a way to work in those fixes first?

I've combed through his commit history, and his "fixes" are hacks that will break other things.

enderandrew, on 01 November 2017 - 06:41 AM, said:

Others jumped in with criticism of how ICD did his Lua scripting implementation. ICD listened to the criticism and changed his code accordingly, because the criticism was specific and actionable.

Ice did not listen to my criticism about his Lua support. The "change" you refer to was an outright patch from Zombie.

[Striker]

Hendricks266, on 31 October 2017 - 09:06 PM, said:

I don't have the time to thoroughly debunk all the things you piled on in your post. It's not about me, it's about the quality of his code.


This is exactly the problem. Ice announces things that sound cool on the surface, and of course the common user is going to react positively to basically anything someone could promise in this position, no matter its merit. Apparently I am the only one willing to speak up about the significant costs and consequences of basically every decision that he makes. So even though Ice promises something shiny, what you're really seeing is a liquid mercury coating on top of shit.

Aye. The code, especially with it's far-reaching changes, heavy-handed hacks, and experimentation in the mix, is not in a state that would be considered useful for EDuke32 or clean enough to be suitable to merge. It's strict what gets used, and what doesn't, and for good reason.

[Mark]

I think a big part of the friction between devs is that H266 is going to be the one left with all the bug fixes and other maintenance in the upcoming years. Hence he would rather deal with his own blocks of code instead of ICD's. Whether it works or not, it will be tougher to figure out someone else's coding.

And of course there is more than one way to reach a goal. They happen to have different opinions on which way to go. Its a tough nut to crack.