Talk:Comparison of source ports

From DoomWiki.org

I'll be simple about it, this is of cause a work in progress. I've layed out most of the menu's, now we only need to fill them. Please, feel free to contribute but try to keep it factual. Eonfge 10:15, 4 June 2009 (UTC)

64-bit?[edit]

Since 64-bit is now main stream, we should list the ports that support 64-bit. These would be Chocolate Doom, Odamex, PrBoom, and ReMooD. Status of ZDoom and GZDoom would not be known. GhostlyDeath 20:31, 10 June 2009 (UTC)

is*, last time I heard they don't support 64-bit GhostlyDeath 20:32, 10 June 2009 (UTC)

Useless page[edit]

I don't see the point. I've fixed some of the more blatant misinformation, but I still don't see the point. These "comparison" tell nothing except their last release date, which is bound to become obsoleted as new releases are, well, released, and this page isn't updated. Especially when the editors doesn't bother checking for facts first -- the latest release of Doomsday is still on the Doomworld frontpage! Then there's that second table with only Chocolate Doom in it, I see the point even less. Is SvStrife Doom-true? Oh, and I see you've re-edited the abandoned status for Doomsday. So here's a link for you: http://www.dengine.net/blog/ That's the Doomsday development blog. EDGE isn't abandoned either by the way. Another link for you: http://edge.sourceforge.net/ -- lurk around the forums a bit, you'll see they're quite active for something "without a community". EDGE merely progresses slowly because its primary developer also works on other projects such as OBLIGE. But I've just taken a look at their SVN and the last log entry at the time I'm typing this is "29 minutes ago" -- how's that for "abandoned"? Last Doomsday SVN entry 13 hours. I'll spare you the Eternity and PrBoom links because you haven't reverted them. --Gez 10:36, 4 June 2009 (UTC)

I think it's a potentially useful page to have around. It's nice to be able to compare the features of different source ports. Clearly Eonfge has made a few mistakes in the information he's added, but that's nothing that can't be cleared up. I also think it's unfair for you to accuse him of trolling just because he's added some inaccurate information - you should assume good faith here.
I've fixed up the information about Chocolate Doom, which does run on MacOS of course (along with many other systems) and also has a client-server multiplayer system. Fraggle 11:15, 4 June 2009 (UTC)
Nice from you to ad some information. Just had my lunch break and I know that some info was still incorrect. As you might have seen GEZ, it's a work in progress. Yes, it's odd that only chocolade doom is in the second list, but that's why we should expend it. I wanted to give it a [1] look, but that of cause takes time. I do believe it's good to make a comparison page, because this is a 'wiki' and with so many source ports, one should be able to make a desent comparison, especialy if you see the Source portpage. That one ain't very pretty or useful either! so please, to all, help us making a nice comparison list.Eonfge 11:32, 4 June 2009 (UTC)
It'll only be potentially useful if it's kept up to date. And I wouldn't bet on it. A system of templates such as this one, but for each port, could make it more useful as it would update both a central activity comparison page and the ports' own articles, but that too rely on it being maintained (nobody bothered to update the EDGE page until I just did it right now). --Gez 11:36, 4 June 2009 (UTC)
That's why I advice to keep it 'general'. So no deep backgrounds on all kernal-supports and render-modes. As long as it provides a accepteble comparison between the different ports. Eonfge 12:01, 4 June 2009 (UTC)

Comparison by content[edit]

I can honestly say I don't find the selected criteria to be very relevant. Where is, for example, "boom-compatible" or "limit-removing"? Those are two of the most commonly found requirements for mods on /idgames. "Doom-true" implies that adding new features makes it "disallow the original Doom experience" even if said new features aren't used. --Gez 11:46, 4 June 2009 (UTC)

Instead of complaining, you better come up with a catchy term. I Couldn't think of anything to make a quick comparison with Chocolate doom and Zdaemon possible, in terms of doom-ish. Both are Doom, I totally agree. But it's no denying that Zdaemon is a lot more different then Chocolate Doom. Limit removing is a good one, will ad that one too.Eonfge 12:01, 4 June 2009 (UTC)
I was thinking about something like this:

<snip>

It allows to showcase each port's strengths and weaknesses. --Gez 12:23, 4 June 2009 (UTC)
Ergonomically it might be best to split each group of columns into its own table, like this.  The current version is so wide that the ads might cover up parts of it, depending on the user's setup.  Also, the port names will eventually scroll off the left edge of the screen, making it hard to interpret (not everybody has a 1600x1200 display or good eyesight).    Ryan W 17:15, 4 June 2009 (UTC)
I think you're right, we better make some priority's an categories on how we should index it. I am honestly suprised because when I left home 5 hours ago, this page was only in 'v0.1'.Eonfge 17:28, 4 June 2009 (UTC)

I think it would be better to add the license and OS support too the 'general box'. What are the criteria for 'custom content'? I think it's save to say that all source ports support PWADs. Also, at 'demo' there is v1.9 and Complet-N. Is that not the same? if so, I would say we go for 1.9 or 'vanilla'Eonfge 17:28, 4 June 2009 (UTC)

People have had trouble making PWADs work on cell phones and handheld game platforms (there was a Doomworld thread about this recently).  It could also be a problem for homebrew console ports, with their limited RAM and read-only media, though maybe those are all dead now.
"Vanilla" can mean one of at least six demo formats (see the PrBoom documentation).  COMPET-N used one of them, the "doom2.exe" version of v1.9, but so did many other projects.    Ryan W 18:34, 4 June 2009 (UTC)
"Custom content of a nature impossible to create with the vanilla engine, even with deutex or nwt" would be a little too long for a header. So I don't mean "new graphics/sounds/etc." but new obstacles/monsters/weapons (without cannibalizing others with dehacked). --Gez 19:33, 4 June 2009 (UTC)

I would dispute the need for the "vanilla accuracy" column. The other columns are all non-debateable yes or no answers. But the afore mentioned column is something open to considerable debate. Not to mention that it's not even clear as to what Doom engine game it's talking about .--Verm 4 June 2009

Not that much debate. ChocoDoom is accurate down to the original bugs and limits, if you find a Doom quirk that's not replicated by ChocoDoom it'll be treated as a bug and fixed asap. PrBoom is practically as accurate if you use the appropriate settings. ZDoom, on the other hand, fixes many vanilla bugs even if in doing so it alters the behavior slightly (e.g., the blockmap glitch which makes hitscans sometimes pass through monsters and players without affecting them, especially noticeable with a BFG9000 and a Spider Mastermind; where people were arguing that because of this fix it made the E3M8 fight too easy). Legacy altered Doom behavior even more, fixing some harmless vanilla bugs (such as voodoo dolls...) and introducing many more. Odamex and ZDaemon are based on older versions of ZDoom which had less deviated from the original Doom codebase, so they're more vanilla-accurate than current versions of ZDoom. And so on. It's pretty easy to quantify, really. --Gez 23:24, 4 June 2009 (UTC)

Hacx 1.2[edit]

Should Hacx be added to the "supported games fields, given that the "official" v1.2 release is a standalone Iwad? —The preceding unsigned comment was added by Verm (talkcontribs) .

I consider it a borderline case. On one hand it has had a commercial release which gives it a certain weight amongst WADs. But on the other hand, the commercial release was small and relatively insignificant. I'm not sure whether the 1.2 release has gained much mainstream attention, but to me it looks like it hasn't. Overall my personal estimation is that Hacx isn't comparable to Chex, Strife or the Raven IWADs in notability terms and should be left out of the supported games listing because of that. -- Janizdreg 16:23, January 29, 2011 (UTC)
So far, source ports with support for the Hacx IWAD include ZDoom, GZDoom, EDGE, Chocolate Doom, and Eternity, AFAIK. I know that DaniJ was also interested in adding it to Doomsday as well. Given that it's now a feature for many of the most popular ports, it would be justified. --Gez 17:41, January 29, 2011 (UTC)
Well, the features compared on this wiki article are biased toward the editing features and supported games of each port (over things like game modes and visuals, which was why I added a section for the later). I agree with Gez that adding Hacx1.2 is justified, with indication that it specifically refers to the 1.2 version. (I.e. the field is called "Hacx1.2" or something).Verm 20:10, January 29, 2011 (UTC)

Rename Page?[edit]

I'd like to suggest that this page be renamed "Comparison of source ports", instead of "Comparison of Doom source ports" (with a redirect for this title of course). —The preceding unsigned comment was added by Vermil (talkcontribs) .

Good idea. The list already covers several things related to other Doom engine games, after all. -- Janizdreg 23:45, 30 September 2011 (UTC)

This page is very bad.[edit]

This page very bad. It should instead use sortable columns with automagically generated data based on every source port in the article so that the page wouldn't have to be maintained ever again. Also fixing column colors after fixing something is a bad thing to do also, it should do the colors for us not the other way around. GhostlyDeath 01:54, 1 October 2011 (UTC)

EDGE > 3DGE[edit]

Should EDGE be replaced by 3DGE on the list, since it's the direct continuation of EDGE?

Compatibility WIP??[edit]

Use of "WIP" as a compatibility status is meaningless. Whether or not Chocolate Strife has been officially released yet, which is under no one but fraggle's control, it's still 99.999% compatible with the vanilla game engine to the point it has demo compatibility. Please don't belittle our work by labeling it as a mere "WIP" in this manner. --Quasar 23:55, 15 February 2013 (UTC)

The name isn't ideal but I think Eris Falling has a point [2].  Our typical reader doesn't follow the dev discussions (or even look at release notes) and will assume that if a feature is listed, that means it was included in a package at some point, or at least a public beta.  Are the SVN builds seriously intended that way?  I thought SVN builds were by design experimental and sometimes unstable, even Budko's.
I suppose we could add a footnote linking to the main article and saying that the work is 99.999% done — although strictly speaking even that is non-NPOV, since Fraggle's and Quasar's reputations contribute heavily to the belief that a release will eventually occur.  :>     Ryan W 00:08, 16 February 2013 (UTC)
Chocolate Heretic and Hexen are, for all intents and purposes, complete, and have been so for a long time. If built from the v2.0 branch, they even have netplay, which was arguably the only thing missing previously. Chocolate Strife still has a couple of things left to be done on it but other than that, it's complete and has had two "beta" releases already, with a third to possibly follow soon. I think the delay in getting them released is, at this point, due only to feature creep and a general lack of time from fraggle. I just don't think that "WIP" describes a state of compatibility adequately, given that Choco Heretic and Hexen were already 99% vanilla compatible when fraggle first got them compiling - the work since then has consisted of merging into Choco's shared framework, bug fixes for things that didn't work like they did under Watcom/DOS, and changes for portability. What I'm saying, in short, is that the state of development isn't necessarily related at all to compatibility. --Quasar 10:07, 16 February 2013 (UTC)
How about "Branch" instead of WIP or whatever, to say it's considered complete but is not in the official trunk yet for unrelated reasons? People can always refer to the main article (in this case, Chocolate Doom) to get in-depth explanations. This page ought to be mostly tables. --Gez 11:26, 16 February 2013 (UTC)
How about "Hurry up and release v2-branch, fraggle!" ;) No I'm kidding. I wouldn't object if you think it's a good idea. It'd be nice if the text did link to something that explains what that means though. --Quasar 14:43, 16 February 2013 (UTC)
Sorry for adding that to your comment Ryan W, but I had no recollection of that edit until you showed it to me. Now as that link shows, I changed the fields to WIP because it was suggested that it is what they were. Like Ryan W's typical reader, I do not follow dev discussions. OK, in retrospect I probably shouldn't have made an edit with no knowledge of what I was actually doing, but the fact that it simply "Exists" and wasn't marked as "Yes" says to me - a non-developer - that it's not finished. When it comes down to it, only those who follow the dev discussions or specific users of the port will know that the work is 99% finished. I think if these things are 99% complete, then why not simply mark them as that? --Eris Falling 09:54, 17 February 2013 (UTC)

Is it necessary to have Skulltag on the list anymore?[edit]

Zandronum is the direct continuation of Skulltag, so I don't see why it's necessary to have Skulltag on the list anymore. Most everyone who used Skulltag outside of a few people have switched over to Zandronum by now. (Though I could be wrong) Death Egg 22:40, 26 May 2013 (UTC)

The question mark after Vavoom's license.[edit]

Why is there a question mark after Vavoom's GPL license? Is it GPL or not? Also, why has Doom Legacy not dropped the DSL license when Heretic became GPL'ed?

Odamex and ACS[edit]

Re: this edit: What to believe? The license the ACS interpreter was released under is not GPL complaint [sic], so there is no ACS support in Odamex. or The ZDoom 1.22 core engine, as well as the ZDoom 1.23 beta33 ACS interpretor[sic].? --Gez 20:23, 11 December 2013 (UTC)

I talked to both RjY and dr_sean on #doom-tech and they say Odamex has ACS and uncapped framerate. So I approved the revision. --Quasar 22:17, 11 December 2013 (UTC)

Vavoom[edit]

Now that the homepage is dead and only preserved by the Internet Archive, maybe it'd be appropriate to remove it from the listing? The page should focus on current ports, which is why EDGE was replaced by 3DGE and Skulltag by Zandronum. --Gez 13:01, 9 January 2014 (UTC)

Doom Retro compat rating[edit]

I take a bit of an issue with Doom Retro being rated Very High in compatibility when it apparently doesn't play back vanilla demos according to the same table. That would rank it higher than Eternity which does play back demos, as well as support multiplayer, a very fundamental part of the Doom experience. Suggest changing this ranking to "High" if others agree that's appropriate. --Quasar (talk) 20:58, 22 October 2014 (UTC)

I took "compatibility" to refer to map compatibility, and nothing else, so that's why I put "Very High". But demos are indicated in the table, so sure, I'm okay with the change.
It's "accuracy", not compatibility, per se. Every port ought to be very compatible! But some do introduce gameplay changes, fix glitches, etc. and that gives a slightly different experience from the original. --Gez (talk) 22:07, 22 October 2014 (UTC)
I have so far in my actions with regard to this article reserved the value of "very high" for ports that go so far as to do partial emulation of the x86 DOS machine's opcodes, memory layout, and actions under instances of undefined behavior, so as to act exactly like the vanilla executable, guaranteeing more or less 100% demo comp, mod comp, and gameplay experience comp. The only two ports that do this on an extensive basis in all three areas are Chocolate Doom and PrBoom-Plus. Eternity has a bit of this, but not nearly as much. EE also takes much more significant liberties with play experience, such as using a psycho-acoustic model for sound priority rather than vanilla's rude FIFO sound channel assignment behavior. --Quasar (talk) 16:26, 23 October 2014 (UTC)
Out of curiosity then, what defines average, low, and very low accuracy? I could probably argue that breaking demo compatibility deserves at least "average". I can understand that ZDoom is considered low since it intentionally makes some game play related changes by default (using Raven invisibility behavior for example), but besides that, isn't having identical core logic (demo compatibility) the differentiation between high and average? Not related to Doom Retro, but isn't "very low" basically the same as "exists." Blzut3 (talk) 01:48, 26 October 2014 (UTC)
There's no set rule, so it's very arbitrary at the moment. I'd suggest this:
  • Average: released source code without gameplay changes
  • High: tweaked code to restore vanilla demo compatibility (undoing Bernd Kreimeier's changes)
  • Very high: same as high, but went even further to handle freak demos (overflow emulation, etc.)
  • Low: tweaked code to fix bugs and add new features
  • Very low: same as low but even further? Not sure there's much point to keep this category.
Also there's the issue of assessing the default configuration vs. assessing the most accurate configuration. I'd favor the later. So PrBoom+ in Doom complevel, ZDoom in "Doom (strict)" profile, etc. --Gez (talk) 07:04, 26 October 2014 (UTC)
Well, in that case I guess Doom Retro should be "low" then. Not wanting to misrepresent my port... --bradharding (talk) 07:07, 26 October 2014 (UTC)
As I mentioned in my last posting, very low would be a good way to designate incomplete support. Some more non-tech friendly descriptions could be:
  • Very high: Suitable for demo playback and recording, including where normally undefined behavior exists.
  • High: Suitable for most demo playback and recording.
  • Average: Demo compatibility not restored (similar state to source release). Suitable for playing vanilla levels.
  • Low: Contains non-breaking game play tweaks. Suitable for playing vanilla levels.
  • Very low: Incomplete or otherwise breaking changes. Not necessarily suitable for playing vanilla levels.
These descriptions do leave average and low kind of arbitrary, but I think we would be able to fairly judge between them. As for judging the default vs best case. I think I agree with you there. Blzut3 (talk) 14:17, 26 October 2014 (UTC)

ReMood[edit]

Given that this port hasn't been updated in about 6 years, though it's website is kept up and features news on the authors personal life, should this port be removed from the list?

I mean, Vavoom's website is now back online since the earlier discussion above, but it hasn't been re-added to this list due to the port not receiving an update for some time?

I will be adding ReMooD back into the list now, since I released in August 2016 and plan on a rapid release cycling now. GhostlyDeath 18:26, 23 August 2016 (CDT)

Does Mocha Doom qualify for this list?[edit]

I asked this on the forums recently, but would Mocha Doom be notable enough or have a large enough player pool to be added to this? I'm pretty sure it still is in development, has a few interesting features the rest might not have, and is probably about as notable as DelphiDoom or ReMooD. Death Egg (talk) 14:35, 22 January 2017 (CST)

No, I don't think so. It has had no development for over three years and Maes doesn't seem to be promoting it anymore. --Gez (talk) 16:34, 22 January 2017 (CST)

QZDoom[edit]

Once QZDoom is officially merged with GZDoom (And already has in the daily builds, just not in an official release) should it remain on this list? It's already been confirmed it will become just a test vessel for potential new features for GZDoom afterwards. Death Egg (talk) 11:38, 27 March 2017 (CDT)

Doom Legacy removal[edit]

Should Doom Legacy still be on this list? It's been years since it was considered a port of choice for many people, with most of its mods having fallen into obscurity, found support by other source ports like GZDoom, or been upgraded to other source ports if they're even still in development. (Only The Ultimate Chex Quest and DeimWolf come to mind) wesleyjohnson still updates the port so it could still be considered 'active', but it's mostly just been maintaining compatibility with Windows 10 rather than actually adding new modding features. It's not that I have anything against the port; 10+ years ago it was my port of choice, it's just now it's about as used as ports like Doom95.

Death Egg (talk) 16:30, 17 April 2018 (CDT)

Classic RBDoom 3 unnecessary[edit]

I don't think the recent addition of this port to the list should be kept. It's nothing against the port author, but as it stands there's nothing outwardly unique about it, and it hasn't reached any kind of serious popularity outside of a tiny niche. --Death Egg (talk) 19:34, 11 May 2018 (CDT)

You're probably right. I let it lie for the time being to see what other people thought. --Quasar (talk) 07:28, 15 May 2018 (CDT)

Perhaps rebrand this page as 'active source ports' and set a date of updated in the last two years or so to be included? —Preceding unsigned comment added by 80.0.251.15 (talk)

The page source ports should work fine for that, since it lists all ports based on when they've been updated. If it was renamed, would it be when a last official release was or when the last GitHub release happened? Odamex doesn't have any active developers but a very active community, if that rule was followed it'd be removed eventually. Death Egg (talk) 14:15, 19 May 2018 (CDT)

What players want is DMing new mods and streaming new mapsets.  If the major ports are in a quiescent period where developers mostly work on plumbing rather than new features, as happens every so often, then unmaintained ports can remain active as you note.  Eventually new features are released and modders use them, but still the community waits to see if other developers adopt an older project, before abandoning it completely.  I realize I'm not answering the question but I think it's a judgement call — nobody knows how to quantitatively measure "active community" and define the exact point where it drops too low to recommend to newbies.  A long time between packages might mean that, or it might not: Doomkid's tutorial still lists Legacy v1.42, which is pretty friggin old.
With purist ports we could be even more inclusive, given that mature apps tend to keep working until a new OS breaks compatibility.  I believe all the listed programs currently have "semi-active" maintainers at least, who reappear at such times to update libraries and whatnot.
The OP is merely a side effect of creating multiple lists.  People link their favorite port from as many articles as they can find, for additional exposure.  If they aren't wiki regulars, I for one don't expect them to understand due weight (which is subjective anyway, and our written guideline could be interpreted to mean every port is equally significant).    Ryan W (living fossil) 12:51, 20 May 2018 (CDT)

I honestly think we should find a better way to phrase it than ports that are "active and popular". I do think popularity could still have something to do with it, as all of the ports listed have some form of significance across the community. The way I've been adding to and developing this article is to make it geared toward having a detailed list of some of the best ports for people getting into the community to start out with.

If we listed every port that was active, that definition can spiral out of control very quickly. Would we add Russian Doom and CnDoom to the list? What about ports with no official releases like Doom Legacy 2, D2K or KBDoom? All of them still get updated, so they have forms of activity. (Assuming KBDoom IS real here) Despite not being updated by Maes anymore, Mocha Doom has been forked off and occasionally worked on by AXDOOMER. Would we include Delta-Touch despite being the conglomeration of four other ports? What about the slew of other Android ports like DoomGLES and Gameception? If we're counting community alone, there's STILL people who swear by Doom95.

I don't mean to come off as being sarcastic or giving too many examples, these are actual concerns I have about developing this list. It's why I removed Doom Legacy despite still being "active"; I don't think anyone would recommend a Doom newbie use a port that's fallen so far behind in modern OS compatibility. Death Egg (talk) 12:12, 4 June 2018 (CDT)

It's obviously going to be difficult to come up with rigorous and impartial criteria. Thing is that the page is also there to give an overview of the important names. To address a few points:
  • CnDoom: it's a niche port for people who want to compete in a specific subset of the Doom community. People aren't going to use it as a generic port for casual playing.
  • Russian Doom: likewise, it's a niche project for a specific subset.
  • KBDoom: there's no point in listing source ports that are not publicly available.
  • Mocha: even AXDOOMER's fork hasn't had a single commit to it this year.
  • Odamex, on the other hand, has had a commit last February. Nothing exciting, but it's not dead yet.
  • Android ports: if anything, those should be on a separate list, in a separate article about mobile ports. You can't make an apple-to-apple comparison with the computer ports.
The thing is that there's a balance to keep between inclusiveness and simplicity. Just like it's easy to imagine a bloated list, it's also easy to imagine it being reduced to two or three names. Heck IIRC originally the entire comparison article appeared as a compromise because another page recommended newbies to just use ZDoom, and then it was changed to Chocolate Doom. --Gez (talk) 12:51, 4 June 2018 (CDT)

Well perhaps then one of the first new restrictions of the article could be to focus on ports that are primarily PC ports. Starting that might seem obvious to us but for other people who might try to add Gameception or some other port it might not be. Death Egg (talk) 14:30, 4 June 2018 (CDT) Death Egg (talk) 14:30, 4 June 2018 (CDT)