![](https://forums.duke4.net/public/style_images/cgs/_custom/switch.png)
Graf Zahl Razes EDuke32 game code from his fork
#271 Posted 25 August 2020 - 08:41 AM
#273 Posted 25 August 2020 - 11:18 PM
#274 Posted 28 August 2020 - 03:52 AM
#275 Posted 28 August 2020 - 11:10 AM
#276 Posted 28 August 2020 - 11:29 AM
#277 Posted 28 August 2020 - 02:27 PM
Master O, on 28 August 2020 - 11:10 AM, said:
Raze, verb. To completely destroy what came before.
Graf's conduct shows Raze's posture.
The development we've seen so far has consisted of replacing our functioning systems with code mostly written by Graf, providing less value in the process by breaking feature after feature as they strive to satisfy their Not Invented Here syndrome.
Ken's software renderer was ripped out. I presume they want to write their own. I wonder what will happen when they discover Build doesn't hand them a BSP tree on a silver platter like Carmack's map format does. Even now their hardware acceleration still uses Ken's Polymost algorithm. Just the GPU dispatch and texture management have been replaced.
There is a branch on their repository that removes libdivide, a small library that optimizes integer division by divisors known at runtime but not at compile time. I am sure that the sole reason behind this is that it doesn't fit with Graf's rigid coding style. In practice, this removal is purely a performance downgrade, while the benefit of the library is critical on some platforms such as ARM.
Accurate palette emulation, accurate sound mixing, Fury's new game menu, SSI file loading for the Sunstorm expansions, and Adlib music from the best-in-class Nuked OPL are all casualties of Raze's development principles.
You, Master O, are not one to talk. You have a bad habit on this forum of finding a post from an important person in a thread, and immediately asking them an annoying question unrelated to the topic. Moreover, you originally named this thread "New Build Engine Being Developed", which demonstrates your ignorance and willingness to parrot exaggerated claims.
Phredreeke, on 28 August 2020 - 11:29 AM, said:
It's not just jabs, Graf makes outright false statements to prop himself up. For example, his "findings" that he quotes in the thread above are 1) not the source of the issue he originally went looking to diagnose, which actually stems from details of how our timers, game loop, and interpolation work, but they also 2) are a problem he entirely brought on himself by ripping out the texture atlas we implemented to minimize state changes. It only would have come up otherwise when using hightile, which none of the games do, Ion Fury included. Furthermore, our graphics engineer examined his supposed fix and found that 3) while it might increase performance on Nvidia cards, they would likely make it worse on all other vendors. Nvidia cards already get the best performance, so this demonstrates the project's backwards priorities.
Graf made a big deal about linking all the games together into a single binary. However, when he tried to merge the Rednukem and EDuke32 cores together into a single set of code, he abandoned the idea because it was too hard. Likewise, when Rednukem got Duke 64 support which included a port of its renderer to OpenGL, he complained because this code got in the way of his ability to siphon off changes we make as part of our own development that we are capable of doing but he is not. Current plans for unifying the games' modding support consist of bolting on ZScript, a bespoke language based on DECORATE scripting for Doom, "just for the HUD and menus".
Separately, I have been informed that if EDuke32 adopted Raze's code as it stands now and continued development from there, Graf would consider his job done. This is detached from the reality of the damage he has done to it and shows he isn't in it for the long haul. He puffed out his chest and claimed he was saving EDuke32 from itself but has caused more harm than good and his lofty claims have yet to come to fruition. There are formidable challenges ahead of Raze leaving "alpha". They have torn down many pieces that contribute value to our projects, rearranged some other parts, and stabbed in systems from GZDoom, but I have yet to see any significant Construction.
This post has been edited by Hendricks266: 29 August 2020 - 06:29 PM
#278 Posted 28 August 2020 - 02:50 PM
#279 Posted 28 August 2020 - 02:56 PM
Hendricks266, on 28 August 2020 - 02:27 PM, said:
My understanding is that they're just gonna do Polymost with a software rasterizer.
#280 Posted 28 August 2020 - 02:57 PM
Phredreeke, on 28 August 2020 - 02:56 PM, said:
They may want to look at Ken's old code for rendermodes 1 and 2. (0 being classic software, 3 being GL Polymost, 4 being Polymer.)
#282 Posted 28 August 2020 - 06:23 PM
Or it's a pure ego trip for him. Or both. Knowing him any are equally likely.
#283 Posted 28 August 2020 - 08:11 PM
Honestly, if Graf wasn't the lead on the biggest sourceport he'd be persona non grata across the whole community.
#284 Posted 28 August 2020 - 09:21 PM
Morpheus Kitami, on 28 August 2020 - 08:11 PM, said:
Honestly, if Graf wasn't the lead on the biggest sourceport he'd be persona non grata across the whole community.
There's probably something to be said for the fact that Raze seems more interested in playing oneupsmanship with eduke and less interested in making a good source port.
I know it's in alpha but the foundation appears to be built on pure spite rather than any general sense of wanting to provide the best experience possible.
Maybe I'm wrong and I sort of hope I am. But given Graf's history as well as how mishandled GZDoom has been in the past in terms of user-friendliness, I kinda have my doubts.
Multiple source ports can be a good thing but I don't have a lot of faith in Raze, personally. Too much stained history to see anything other than spite for eduke.
#285 Posted 28 August 2020 - 09:22 PM
I have seen this same sort of drama in the Wolfenstein 3D community. If you think this community is small, then you really don't know what small is until you've been there. Everyone knows everybody there. Wolfenstein 3D modding has always had its quirks that only Wolf 3D modders truly understand. ECWolf comes along, which supports all the games (at least Wolf 3D, SOD, and the "Lost Episodes") under one EXE, doesn't require you to do nearly any of the things you'd have to do with the DOS code or even SDL code to make a mod, and it caught on very quickly with the Doom community. ECWolf having similar systems to what you'd see in a port like ZDoom made an impact in giving fans of Doom the chance to make some Wolfenstein 3D mods in a way familiar to them, but it didn't make the same impact with the Wolf 3D community. Instead, a few took strong interests into ECWolf, while many others chose to ignore it and stick with the original DOS or SDL way of doing things and it has been that way ever since. I think a very similar situation will happen here, but the thing ECWolf has going for it is you can mod it more like ZDoom, but with Raze it doesn't give you any advantage to developing a Duke mod compared to EDuke32, so what is the advantage then here?
#286 Posted 28 August 2020 - 09:34 PM
#287 Posted 28 August 2020 - 11:37 PM
markanini, on 28 August 2020 - 09:34 PM, said:
I know you're just saying that since you've lackeyed up to Graf after getting banned from the Blood and Duke4 Discords.
markanini, on 28 August 2020 - 09:34 PM, said:
There is always more to build. Could you tell me where in GZDoom I can enable three-point filtering?
#288 Posted 29 August 2020 - 12:22 AM
Hendricks266, on 28 August 2020 - 11:37 PM, said:
There is always more to build. Could you tell me where in GZDoom I can enable three-point filtering?
So what? Raze fixed issues for me since the first release, many of the same issue I brought up with the eDuke32 team over the years that are still unresolved. I'd still give credit to your team for the ground work. Exactly what good does do if I offer you sycophantic loyalty you seem to demand?
3-point filtering is moot for Build and Doom games, none shipped with it. DN64 used it's own renderer.
This post has been edited by markanini: 29 August 2020 - 12:22 AM
#289 Posted 29 August 2020 - 12:37 AM
#290 Posted 29 August 2020 - 03:16 AM
Raze would first have to get there, and since Graf seems to be busy taking it apart to turn it into what he would like it to be, I just don't see the point in waiting for something we already have. Playing Duke3D with a GZDoom-like system or menu is probably an amusing gimmick, but if you look beyond that, there isn't really much. That's why I just tried it once, went back to EDuke32 again and never looked back.
This post has been edited by NightFright: 29 August 2020 - 05:15 AM
#291 Posted 29 August 2020 - 05:09 AM
#292 Posted 29 August 2020 - 05:27 AM
markanini, on 29 August 2020 - 12:22 AM, said:
It's funny you mention this given Graf's complaints about Nuke putting the DN64 renderer in the RedNukem source code.
#293 Posted 29 August 2020 - 06:55 AM
lol
#294 Posted 29 August 2020 - 07:07 AM
I suppose I don't share an adoration for Graf that a lot of people seem to have because Doom isn't really that important to me as a game. It's not even my favorite Doom Engine game. That distinction goes to Strife. I started using source ports really late, mostly because I hadn't concerned myself with anything other than Dosbox until around January of 2019. GZDoom was the first port I tried and used it for about two or three weeks until I switched to Chocolate Doom for the main games and Crispy for mods. As a result, I don't have any emotional attachments to it and if you couple that with the fact that I'm simply not that interested in Doom mods, you can understand why I am not in constant awe of his work. It's not a disregard of it, but it simply plays no role in my gaming experience for the most part. I usually stick to playing the official releases of Doom Engine games myself. I have tried several mods, but not enough to consider GZDoom a necessity. There's a reason I have a duke4 account and not a zdoom or even doomworld one.
Now, regardless of my feelings towards Graf, I was looking forward to trying out his port. I can separate the creator from the creation just fine, and any new attempt to create a source port for games I like is something that I welcome and wish to see succeed. For this matter I went in with an open mind and put aside my bias as best I could to appreciate Raze in separation of any petty squabbles. I'm not a programmer nor am I familiar with coding or anything related to it, so I can only give you the perspective of a regular user. In that sense, there simply wasn't anything I needed from Raze that I wasn't getting already from the ports I had. If anything, it currently offers me less because it doesn't have the classic renderer that I use for all Build games with the exception of Ion Fury, for which I prefer Polymost. I can't vouch for some sort of increased performance or stability that some users have experienced with it: personally, I felt no particular improvement on these regards as opposed to what I experienced with EDuke32 and variants. I'm just speculating here, and keep in mind that I don't mean this in any sort of negative light but more of a reflection of events, but the strange identity crisis that Raze suffers from has me wondering if it is due to how it came about: Graf wanted his own way of playing Build games and therefore made his fork to tweak the experience to his liking. That doesn't sound like a passion project to me, and the end result speaks of that. Perhaps after he was done playing the games in the manner he wanted, it has now became obvious to him that Raze served it's purpose and this is why development has stagnated somewhat. Maintaining GZDoom alone must be hard enough work, and to add the majority of Build games to the workload, with many people also expecting to see extended modding capabilities added as well has to really weigh in on anybody, specially if the project isn't coming from a place of passion and fandom. Now like I said, this is just speculation I've engaged in and for all I know Graf has a lot of passion for Raze and is currently working on some super duper update that will turn the thing around into an essential addition to all Build ports. Granted I don't believe that is the case, but my point is that I'm happy to wait and see what happens.
Since I know my post could be misinterpreted I should end things by clarifying my stance regarding some things: I don't hate Raze and I wish to see it succeed. I also don't hate Graf himself, just not a fan of his attitude and the way he can conduct himself at times. I like and enjoy Doom thoroughly, and the fact that I have little interest in its modding scene is in no way a criticism or dismissal of it. It's simply a reality of the type of player that I am, being more into Quake mods and Build mods due to personal preference. The fact that I like them better doesn't mean I think other modding scenes suck.
This post has been edited by Putrid Pete: 29 August 2020 - 07:23 AM
#296 Posted 29 August 2020 - 07:42 AM
Salty Chicken said:
Source: https://www.doomworl...um/post/2178930
#297 Posted 29 August 2020 - 09:23 AM
Phredreeke, on 29 August 2020 - 05:27 AM, said:
That's because I'm not Graf. It tells of how your're unaware that people outside your circle can operate as individuals.
#298 Posted 29 August 2020 - 10:08 AM
![:)](https://forums.duke4.net/public/style_emoticons/default/laugh.gif)
#299 Posted 30 August 2020 - 03:19 AM
markanini, on 29 August 2020 - 09:23 AM, said:
So what? TBH, the same thing can be apply to them too, if you really following both side.
![:)](https://forums.duke4.net/public/style_emoticons/default/tongue.gif)
It(saying such things) won't get anywhere and just increasing this so called "hostile" situation...
This post has been edited by Player Lin: 30 August 2020 - 03:21 AM
#300 Posted 30 August 2020 - 03:23 AM
Player Lin, on 30 August 2020 - 03:19 AM, said:
![:)](https://forums.duke4.net/public/style_emoticons/default/tongue.gif)
It(saying such things) won't get anywhere and just increasing this so called "hostile" situation...
Have you ever owned a business?