Return Styles: Pseud0ch, Terminal, Valhalla, NES, Geocities, Blue Moon. Entire thread

/prog/nix ideas

Name: Anonymous 2017-03-02 2:48

What software should be included in the new ENTERPRISE QUALITY /prog/ UNIX distribution?

Put your ideas below, please.

Name: f̲a̲g̲g̲o̲t̲ 2017-03-02 2:54

diff

Name: Anonymous 2017-03-02 2:54

Anon core utils

Name: Anonymous 2017-03-02 4:17

Cudder's browser.

Name: Anonymous 2017-03-02 7:10

AfterEgypt

Name: Anonymous 2017-03-02 7:15

>>2
these guys would need to implement it first

Name: Anonymous 2017-03-02 9:24

A window manager that.
1.Allow only maximized windows.
2.toolbar to switch windows is icons without text(to fit like ~100 icons in 1 line).
3.Said toolbar pops up for 10 seconds when you hover the mouse pointer in the top-right corner or some other area.

A terminal which natively supports drawing 32-bit pixels:
1.Terminal window can contain mix of graphics and text.
2.32-bit color pixels(8bit alpha/opacity)
3.Programs grab control of terminal graphics and use it as OpenGL framebuffers(without the verbose setup, like in QBASIC)

A task manager that is carbon-copy of Process Explorer on windows.
A network monitor that is simple like TcpView with ability to accumulate packet stats instead of removing them as connections close.(TCPview can quickly tell "which process owns the connection" as connections change in real-time)
A chess-engine GUI which would support most protocols and chess variants:
1.It needs to support multi-engine matches and tournaments(like Arena).
2.It needs support for connecting with online chess servers/sites.
3.Custom piece bitmaps and move/capture animation support.
4.Low-overhead for blitz

A minimalistic browser which:
1.Doesn't support JavaScript or CSS, but creates a compact layout from loaded elements. Image/Video/Audio support.
2.Designed to load everything as fast as possible
3.Has Built-in adblock and domain filter.
4.Is built for secure browsing: everything is verified(buffers/requests) and no persistent data about the user remains(all data is for one session only).
5.Support for multiple protocols: irc, torrents,etc which normal browsers don't have.

A text editor which has
1.Custom Syntax highlighting schemes.
2.Spellcheck with user-editable dictionary
3.Regex-search/replace
4.Notepad++-like interface
5.Code-block folding.
6.Code auto-complete based on #2
7.Multi-document view& visual compare(visual diff)

A wiki database editor:
1.With web interface
2.Support for multimedia/file embeds
3.Scaling for multi-GB wikis
4.Support for auto-linking words(i.e. words typed become links if there is an article about it)
5.Tag-cloud support for pages
6.Password protection and approved editor lists
7.Smart diff, which shows how content is being edited with highlighting: yellow- content being removed and put back again, red-removed content, green- new content.

A Second Life clone like OpenSimulator with
1.Much better in-game editor (with scripting and 3D modeling support).
2.More optimized for multiple connections and scaling up with system resources(GPU acceleration)
3.More realistic graphics renderer(client-side)
4.Native support for Virtual Reality headsets and haptics devices.

A graphics editor like MSPaint with more modern tools and brushes for quick editing, optimized for performance and low memory use.

Name: Anonymous 2017-03-02 9:42

>>7
hey FV why do you want a chess engine and SL clone to be a part of the operating system?

Name: Anonymous 2017-03-02 9:45

>>8 A part of "/prog/ UNIX distribution"

Name: Anonymous 2017-03-02 9:59

>>8 A more complex answer: it provides
Creative outlet/Entertainment(SL) and mental skill training(chess), which replaced most of "muh games" windows appeal at minimal bloat, since SL can replicate any game with sufficient scripting and Chess variants can model any board game.

Name: Anonymous 2017-03-02 10:39

>>10
first of all, make sure to check my dubs. now, to the point: the idea that scriptable chess and an SL clone can replace games people play on windows is based on a fundamental misunderstanding of why and how people play vidya. first of all, this limits the number of games that can be played to those scripted inside chess engine or SL - so you might be able to play 'a role-playing game' but it will probably not be the role-playing game you want. I mention RPGs specifically because this is important for narrative-driven games: there's a world of difference between 'an RPG in a post-apocalypitc setting' and 'Fallout 2'. this is of course also important for multiplayer-focused games: you might play 'a multiplayer first-person shooter' but this won't be the same as e.g. 'playing competitive Quake 3'.

more generally, chess engine and SL clone provide so little for most of the games that aren't chess or SL that you'll need to program most of the games from scratch or port them. and this doesn't offer anything new: 'gaming *nix which can play the same games as Windows' would be a big deal, 'a *nix which can play all the games written for that *nix' is every single *nix ever.

tl;dr saying that chess engine and SL clone will remove the need for Windows-based games because they can be theoretically supported is like saying that Brainfuck removes the need for C because it's Turing complete.

Name: Anonymous 2017-03-02 10:45

>>11
It not about giving every user a windows-game clone.
Its about the potential of using OpenSimulator and ChessVariantGUI to create your own games and distribute/play/share them.
Its basically democratizing the game industry and allowing /prog/nix users to be game creators. Its like a secret ingredient that GNU misses, the entertainment aspect of using the PC and these 2 programs are the minimalist solution.
The alternative is Linux distros with 20GB of random games of which less than 10% are enjoyed.

Name: Anonymous 2017-03-02 10:46

All touhou games

Name: Anonymous 2017-03-02 10:46

A Dubs checker program

Name: Anonymous 2017-03-02 11:05

>>14 #define isdubs(x) ((x%10)==((x/10)%10))

Name: Anonymous 2017-03-02 11:31

>>12
board game engines that allow for easy creation and distribution of board games exist, and they have the advantage of not being strictly based on chess so you don't need to bend chess rules beyond recognition to implement something not chess-like. I don't see how a chess-based one would be superior.

also, while I see the merits of a board-game engine as it can more or less translate into anything turn-based, I don't see the point of SL clone. I guess this is meant for everything 3D but the fact that it would be geared towards MMOs will probably hurt anything that's supposed to be offline. wouldn't a general 3D game engine like Unity be better?

Name: Anonymous 2017-03-02 11:35

A package manager that has no dependencies itself/is bundled with all of its dependencies. You'd think it is absolutely obvious that the program that deals with dependency problems should not ever have dependency problems, but apparently it isn't. Fucking Portage. This requires proper static linking, so you would need a libc that allows it (read: anything except glibc). While you're at it, nuke everything related to Drepper, Poettering and the other ``modernization'' retoids who constantly break shit to sell you the fix later.

Name: Anonymous 2017-03-02 12:16

A personal computer should be a single user system - boot straight to desktop without login. Instead programs should have access control lists(instead of users)

Name: Anonymous 2017-03-02 12:21

>>18
I don't disagree with the idea of a single-user system (ok, I kind of disagree because it's less universal and it wouldn't be trivial to repurpose it as a server), but a personal computer shouldn't even be able to boot straight to desktop. the disk should be encrypted.

Name: Anonymous 2017-03-02 12:25

>>16
board game engines that allow for easy creation and distribution of board games exist, and they have the advantage of not being strictly based on chess
A chess variant is easier to design, with template and tweaks to rules. To design a new board game from scratch takes weeks.
To create a new chess variant, a few tweaks(e.g. allow bishops to reflect off boards).
wouldn't a general 3D game engine like Unity be better?
You don't understand. Imagine yourself, with near-zero 3D engine skills in VR headset, loading a simple field and tutorial. You could sculpt a 3D object by hand, paint it and pop-in a virtual keyboard, add a simple script and the object will start running it instantly. Now imagine this done in Blender/Unity/whatever. Its like writing code in Common Lisp inside a live system (Read-Eval-Print).

Name: Anonymous 2017-03-02 12:43

>>20
A chess variant is easier to design

yeah, so what? it has little appeal to people who want to design video games as there's not much self-expression there (at least when compared to normal game design), it has even less appeal to gamers as it's just endless variations on the same centuries-old game and it doesn't even have much appeal for chess players as they usually try to get better at traditional chess. it's something you'd like because you like chess variants but you must admit it's a niche interest - even when compared to pre-existing Linux gaming (e.g. NetHack - this is far from mainstream but it has a more understandable appeal than 'chess but you must check dubs').

You could sculpt a 3D object by hand, paint it and pop-in a virtual keyboard, add a simple script and the object will start running it instantly. Now imagine this done in Blender/Unity/whatever. Its like writing code in Common Lisp inside a live system (Read-Eval-Print).

the key element in your description is the usage of motion control for the creation of 3D graphics combined with a VR-based live coding environment. this sounds less like SL and more like an extremely modernized Smalltalk environment.

Name: Anonymous 2017-03-02 13:01

>>21
yeah, so what? it has little appeal to people who want to design video games as there's not much self-expression there (at least when compared to normal game design)
It has appeal to people interested in Chess Variants and Chess-like games. Chess is one of the fairest and balanced board games. Its ancient legacy is proof of its appeal, while most board games rarely get played after a few decades.
extremely modernized Smalltalk environment.
Yes, however Smalltalk is incredibly slow and unoptimized. I'm talking about modern C++(SecondLife) and C# code(OpenSimulator) getting an "upgrade". A form of creative virtual reality when you design the "game rules" and "game contents" as you see fit, and then bring friends(or bots) in.

Name: Anonymous 2017-03-02 13:11

Name: Anonymous 2017-03-02 13:30

>>23
ok, now write it as a script for a chess engine

Name: Anonymous 2017-03-02 13:47

>>24
Thats where ChessVariantsGUI comes in. You plop in a template(16x16 chess), add pieces and their rules, write some scripts for actions/spells/special moves or copy them from a preselected list(dropdown), and then save it as a new game.

Name: go > chess 2017-03-02 14:01

go > chess

Name: Anonymous 2017-03-02 14:15

>>25
I'll believe it when I see it. in practice, the scripting will either be too limited to allow such a complex variant (at least without some really complex and obscure tricks) or it will be so general-purpose that your chess engine is actually a board game engine (which I still think would be a better idea)

Name: Anonymous 2017-03-02 14:24

>>27 A board game engine is too general and less user-friendly.
The idea is that ChessVariantsGUI will have many premade templates/rule templates, units, spells/abilities that creation of a new variant will be accessible to practically everyone and advanced scripts/customization for the experts.

Name: Anonymous 2017-03-02 14:25

Name: Anonymous 2017-03-02 14:34

>>28
like I said: I'll believe it when I see it. are you going to write such an engine?

Name: Anonymous 2017-03-02 14:39

>>30 Gaming projects are low priority for me. I've just answering "What programs a /prog/ UNIX distribution" needs.
If you have plenty of free time, you can write one yourself - thats the beauty of open source(smug.jpg).

Name: Anonymous 2017-03-02 14:40

>>31
I'm actually working on a game, but it's hopefully more interesting than an extensible chess engine

Name: Anonymous 2017-03-02 14:48

>>32
What is replay value of your game?
Can you modify it substantially with a slight parameter/option/rule tweak?
Can your game be a basis of online competition/esport?
Is your game more of an art piece or game mechanics showcase?

I have some game design articles on my wiki if you like suggestions:
https://www.reddit.com/r/frozenvoid/wiki/pages/

Name: Anonymous 2017-03-02 15:07

>>33
What is replay value of your game?

nothing yet because it's not even close to being complete. target replayability is similar to that of RPGs with multiple endings: the main storyline can go into one of the different branches (a bit like different paths in a VN) and each of those branches has multiple endings. plot is actually what I'm working on now, I already have most of the backend code but there's no GUI yet.

Can you modify it substantially with a slight parameter/option/rule tweak?

no, because this isn't the priority. I don't have much gamedev experience and the scope is already pretty big so it probably won't be. it will support different game modes but they do require a bit of programming to implement

Can your game be a basis of online competition/esport?

no, because this is supposed to be a single-player game. competitive or cooperative mode would be possible to implement because the rules are fairly simple and 'board-gamey' but once again, I'm not doing it now because the scope is already big.

Is your game more of an art piece or game mechanics showcase?

both, but more of an art piece. mechanics are deliberately simple but complex situations do arise.

I have some game design articles on my wiki if you like suggestions:

I read some of your game design stuff a while ago and found it interesting but I don't think it applies much to my project as it mostly seems geared towards MMOs. but feel free to link me specific articles you think might be relevant (my game is supposed to be a strategy/city builder with a text-based, VN-like narrative if that's relevant).

Name: Anonymous 2017-03-02 16:29

The best unix distro is vms.

Name: Anonymous 2017-03-03 22:26

>>35
Gentoo.

Name: Cudder !cXCudderUE 2017-03-04 4:13

>>35
The best unix distro is vmware.

Name: Anonymous 2017-03-04 17:10

>>37
le cudder sage

Name: Anonymous 2017-03-04 17:23

>>37
The best OS is x86? That doesn't even make sense.

Name: Anonymous 2017-03-24 23:01

>>36
just use openbsd, negro

>>35
how is your login.com doing?

no, seriously why just don't you use RESEARCH[/u]security[/u] operating system as your desktop.

Newer Posts
Don't change these.
Name: Email:
Entire Thread Thread List