On 5/30/20 11:04 AM, Tobias Frost wrote: > On Sat, May 30, 2020 at 02:38:09AM +0200, Roland Plüss wrote: >> On 5/29/20 10:07 PM, Tobias Frost wrote: >>> On Fri, May 29, 2020 at 12:32:42PM +0200, Roland Plüss wrote: >>> >>>> I'm a bit cautious to allow installing games into the user home >>>> directory. Game files can quickly grow large (up to GB of data). One >>>> reason why I opted for /opt in the first place. >>> Speaking of games, are there free (FOSS) games available? A quick Internet >>> research yielded no results, so I'd appreciate if you could provide >>> pointers. >>> >>> Thanks! >>> >> Not yet. The engine has been released public for the first time >> recently. Available right now are example projects including >> ready-to-run delgas: https://github.com/LordOfDragons/deexamples . The >> engine can be used for both FOSS and commercial alike without troubles. >> To makes games though you need the engine in the first place. It's sort >> of chicken/egg problem here. > Ok, thanks for being honest. (I guess the "GB of game data" is nothing to be > concerned right now and when it becomes a thing you can always extend your > programm to look in more than one location or the games provide some launcher > script. (To be clear, this does not change anything about /opt being > unsuitable.) > > But there is now another chicken-egg problem: In Debian we don't try to > package > every piece of software; e.g it should have have already some relevnace, > users, > games, etc… Please don't think that being in Debian will boost the > userbase significantly. > > We have already plenty other game engines in the archives, so maybe elaborate > a bit what makes yours special over the others? > > One plus I immediatly see is that there are free examples (they should be part > of the package!) and seems to work well with a FLOSS toolchain only. (at least > I have the impression it does). But on the other side, it has yet to proof in > real games that it fit enough for them. > > One the minus side I see that you are alone on the project and despite the > project being quite old (I see featurelists dated 2014) it seems not to have > gained traction. Hinted by the absence of games and contributions to your > project, so I fear that the project is missing relevance. > > I write above not to discourage you, I just want to be frank with you that I'm > not sure that the software is ready for Debian. I want to avoid that you put > significant work into the packagaging just to find it rejected later. > > On the other hand, fixing the issues mentioned can also improving the > quality of your project. > > The videos you have published have kind of impressed me, but _disclaimer_ I'm > not a game developer ;-). > The reason for this is quite simple. I did not wanted to go down the "release early, release crap, become a mess" route so I did not release the source base to the public until recently. The entire design philosophy and workflows are different than other engines so getting the workflows working solid and fast had been important. This required especially waiting with a release until the various parts reached the goal level so changes and additions do not break things left and right (looking at source engine here in particular).
Games typically compile engines into the file project or at least link hard-link against engine versions. This creates maintenance costs, tends to break with OS/Platform updates and requires rebuilding/repadistributing for different platforms. This engine fully separates the engine from the project. You do not link against the engine in any way. Delgas are thus truly cross platform and you are not forced into licenses. Even after after a game is release ant not maintained any more if new updates and features get into the engine the game still benefits from it with 0 maintenance cost. This means you can make a game on windows and people can run it on linux although you never touched this OS at all. From the developer side this is a directory based engine so no need to import/export all the time or fighting with assets. Edit an image with GIMP in the project directory and test-run. Export from Blender right into the project directory and run. You can even change the file while test-running and load the file again. The used file formats are open. There are no closed proprietary formats (nor media nor assets) used. For the engine defined file formats Blender scripts are present so you can export anything from Blender straight into your project. The distributing is also WYSIWYG: what is in your project directory becomes the distributable. No hidden magic, no cooking. Due to the directory nature this engine integrates well into existing workflows. Combine all kinds of tools you want. As long as the result ends up as a file in the project directory your golden. Another point is safety for the user. Bugs in games can trash your system (been there, done that, Valve for example trashed my machine with a broken uninstaller). Game distributions can be tempered with to snuggle malicious code into it. Delgas are no binaries. The game is run native (no VM) but confined into a VFS preventing the game to do anything you as user do not explicitly allow. In particular your home directory is off limits. No chance for a game to read your private data not even data of other games. The user can safely download and run delgas from the internet. The worst that can happen is that the delga is prevented from running. Another point is that game developers work hands-off with the engine not hands-on as it is the case with others. You don't tell the engine "how to do it" but "what you want to do" and the modules do their best to deliver the best possible result. In the end everybody wants best graphics, best network performance and best physics if possible. Using this separation the developer does not have to worry about target platforms and their abilities/performance. The installed engine is configured by the user once to give him the best personal experience. Development becomes faster, users are more happy (if your system runs the engine it runs any delga you throw at it). These are a few points. Now if you insist this will be rejected because not having yet finished games projects out for years then there is not much I can do about it. This engine had been already at a game show and used by players and didn't falter so I'm positive it is ready. That said I'm doing things like supporting new OS and packaging for OS also to improve the project. So if this should be rejected, no problem to me. I've got other path I can walk too. But please if you really think this has no chance to enter Debian no matter what I do then tell me this please now. Then I will at least improve things using the lintian and go on from there. -- Mit freundlichen Grüssen Plüss Roland Game Development and Game Engine Technologies https://dragondreams.ch
signature.asc
Description: OpenPGP digital signature