IRC Channel - #unixhub-distro

Proposed general outline: tbd (bleeding edge, minimal, rolling, etc.)

After some semi-serious conversation in the IRC, the idea of restarting, forking, or making a new distro inspired by ArchBSD was brought up. In an attempt to shift the focus of Unixhub to a more collaborative coding community, a project such as a distro, released under the Unixhub license, could be a major asset in both spreading knowledge of distro construction, GNU/Linux, as well as the inner-workings of working on large-scale projects.

I believe that the focus should be on a strong integration of a bleeding edge and rolling release model, including default Wayland integration, as well as a serious focus on a politically neutral development ethic. As for using a BSD kernel at the core, however, I do believe that to make use of the massive community of developers and corporations (Intel, Google, etc.), a Linux kernel is simply more reasonable. Desktop performance is simply not there in the BSD kernel; this is not a BSD vs GNU/Linux thread, simply a factual statement that BSD can not compare to the Linux kernel for desktop performance, which is the target audience for a project like this (desktop, not servers, or embedded). Modern hardware can only be utilized to the fullest extent with a Linux kernel.

I have never made a distro, or at least never finished one. I have played with LFS before, but nothing with a team, and nothing to completion. That being said, there is a lot that goes into distro; the many levels of development, in both complexity, and specificity, would allow the project to take off quickly with many developers of various skill sets.

Is there anyone that would be interested in a collaborative project of this sorts, to really exercise this community's abilities?

Fixed title. -NeoTerra

I think this would be an interesting project. I'll most likely have some free time coming up, and I would gladly throw in a couple of hours of work on this. But I cannot and will not do this by myself, so I hope there should be other people jumping in on this. Honestly, the "participation" in the UH Portage overlay somewhat scared me off community projects, at least with the folks around here.

When it comes to the base of the system, I think personally I would prefer to build it on top of BSD, alone for the fact that in the Linux world, it's hard to find a niche that hasn't been filled yet by some more or less obscure distro. And if we are building something, I would prefer it to be something different, and not just another fork of something with some tweaked default settings. Also, I think most of the people here are too autist for something as mainstream as Linux. But hey, I think we can find something that everybody can agree on.

(09-02-2014, 02:34 AM)crshd Wrote: Honestly, the "participation" in the UH Portage overlay somewhat scared me off community projects, at least with the folks around here.

The thing with the portage repo is that not a lot of people on UH are using gentoo/funtoo. That explains the lack of interest.

I know there aren't a lot, but at least some. And from the responses in the thread, I assumed I'd at least get some submissions.

(09-02-2014, 02:34 AM)crshd Wrote: But I cannot and will not do this by myself...

When it comes to the base of the system, I think personally I would prefer to build it on top of BSD... And if we are building something, I would prefer it to be something different, and not just another fork of something with some tweaked default settings.

I didn't want this to fall onto a single person; the whole point of a collaborative community project is that it is a community project.

Well, kernel wise, if we go for a BSD kernel, we are missing out on a huge community of developers that, in the end, has the best desktop performance. We could always offer the choice between a BSD base and GNU/Linux base, but I'm not sure how much of the code we could recycle - perhaps that's too ambitious, though.

An entire distro is quite a large project for a small community. I recommend making a software tool to make something we all love better.

Ideas?

We could make an entertainment distro that is based off of Debian for the living room which allows users to play games, watch/stream youtube or movies.
I was thinking of calling it Linux Entertainment System.

Or GNU/Lesbian for short.

(10-02-2014, 04:24 PM)NeoTerra Wrote: The other day BigE and crshd were talking about a firewall they loved for Mac, yet there was no real equivalent Linux version. We could do something like that, instead of a huge undertaking like a distribution.

I would do my best to contribute to something on the level of a per-application firewall like LittleSnitch on OSX. It is a great piece of software, and if we could clone it, I think it would be a successful and well used project.

(10-02-2014, 04:24 PM)NeoTerra Wrote: This is a good idea however here are my $.02:

1) If we're gonna do this, we need to all be on board with what's going on. That is, 100% unanimous decisions regarding most if not all design decisions.

2) I would prefer something Linux, not BSD. The other day BigE and crshd were talking about a firewall they loved for Mac, yet there was no real equivalent Linux version. We could do something like that, instead of a huge undertaking like a distribution.

3) Resources. This kind of goes without saying, we're going to need at least one server, and someone to manage it. I would be happy to do this, if we get enough momentum going.

Let's definitely keep up this discussion though, because I think this has potential. :)

EDIT: This might be my autism coming out, but it's UnixHub (camel case), not Unixhub!

I completely agree - without total understanding as to what we are undertaking, then there is no /way it can be worked on as a community. There needs to be an absolute goal in mind.

I definitely want to use a Linux kernel, for aforementioned reasons.

I have a server rack, and plan on upgrading my servers soon. I can contribute server space if necessary, in addition to what you can/will provide.

Personally I'd lean more towards a distro. Yes, it's more work than a single piece of software, but this also means that there's something to do for everybody.

Taking the firewall as an example - I don't C, so (for the sake of the example let's assume it's written in C) I wouldn't have anything to do. I do however Perl, Python, SH and what have you. So if we're working on a distro, I could leave the C stuff to somebody who knows their stuff (looking at bottomy), and concentrate on what I do best. Unless of course we decide to write the whole damn thing in C, then I'm still sh*t out of luck.

I'm leaning towards a whole distro, as well. I feel it could really spread knowledge and experience among the forum, and attract new members looking to work on a large project of the sorts.

(11-02-2014, 01:21 AM)NeoTerra Wrote: What kind of distro are we going to do that hasn't been done yet, though?

https://en.wikipedia.org/wiki/List_of_Li...tributions

I'm sure you guys are creative enough to come up with something that hasn't been done before.

We could all write drivers for FreeBSD, so every time I install it on a laptop I don't spend the next 10 hours wishing I had chosen to cave my skull in with a framing hammer instead.

Originally the idea was to create a rolling-release BSD system, similar to ArchBSD but building off the work yrmt has been doing with pkgsrc. However I'm open to ideas. I feel like a software project might be a better place to start. A distro is a pretty large undertaking, and requires a lot of unanimous decisions, and a decent amount of marketing to make the whole thing feasible.

UnixHub is gaining some pretty solid internet presence and I think if we all get behind one thing we could make something super useful. There is great potential here :)

Why not create a *nix game. A game will needs everyone participation.

If you guys want to work on a nice minimal yet modern rolling release BSD system with pkgsrc, try to help with EdgeBSD.

(11-02-2014, 03:25 AM)venam Wrote: Why not create a *nix game. A game will needs everyone participation.

If a distro is not acceptable, I am all for a game, or game engine. A game would allow many people, who may not write C or C++, but could have helped with the distro, still work on the project. I've been playing with an engine myself, and feel confident we could create something modern, powerful, and OSS that really embodies the UnixHub spirit.

I read that as "playing with something using an (existing) engine", although it's not quite clear.

Yeah you should just play with SDL.

(12-02-2014, 01:06 AM)NeoTerra Wrote: Game engines are extremely time and work intensive, I don't think anyone (that I know of) here has the skill set for a gaming engine.

Well, then, let's make a game, or a distro, if an engine is out of our capabilities. All of my GL work has been with GLFW, GLEW, GLM, although I have *played* with SFML and SDL. Were we to write a game, I would assume that we'd use GL with a language in the C family - am I correct in this assumption? If so, then we need to decide on libs and frameworks. A lot of people like SDL, and I can understand why, but I'd prefer if we used GLFW.

Does this mean that we *aren't* making a distro, or are we simply discussing possibilities?

(10-02-2014, 06:39 AM)funriz Wrote: We could make an entertainment distro that is based off of Debian for the living room which allows users to play games, watch/stream youtube or movies.
I was thinking of calling it Linux Entertainment System.

Or GNU/Lesbian for short.

I was thinking "surely nobody will get behind this...", then I saw the joke. Well played.

(12-02-2014, 03:49 PM)ichiban Wrote: Well, then, let's make a game, or a distro, if an engine is out of our capabilities. All of my GL work has been with GLFW, GLEW, GLM, although I have *played* with SFML and SDL. Were we to write a game, I would assume that we'd use GL with a language in the C family - am I correct in this assumption? If so, then we need to decide on libs and frameworks. A lot of people like SDL, and I can understand why, but I'd prefer if we used GLFW.

Does this mean that we *aren't* making a distro, or are we simply discussing possibilities?

I'm quite fond of SFML myself. As bottomy proved when helping me with my (crappy) game, we clearly have people talented in that area. I personally think it allows a bit more room for creativity, involvement (even if you don't know the language you could contribute ideas, art and music).

A distro would be great, but I think it requires a lot more effort in the name of background reading on how all this stuff works. Maybe it's just me, but I'd have no clue where to even start on a distro, and to be honest I think at the moment I wouldn't personally have time to keep up with things enough to be an active part of development. Not to discredit a distro, go for it if we want, it'd make more sense than a game, but I think you'll see less involvement if you do. Maybe I'm wrong.

Alright, then - I am okay with *starting* with a game. A distro is a large undertaking; a game is much more doable in a shorter time frame.

Still, just making *a* game is vague. 2D or 3D, procedural, linear, etc. - all design choices that will change the code base quite a bit.

If we are going to make I game I suggest making it for mobiles. We'll need some people that take care of the engine, some of the story, some of the art, some of the audio, there's a lot of area where everyone can contribute.

If anyone wants suggestions for the type of game, I would totally buy or donate for a well made clone of Fire Emblem(for anyone who remembers playing those awesome games on the GBA), a turn based strategy with some but not excessive class leveling. Recalling back to being 10, I think the game would work well on a mobile device because pinpoint precision is not important, and there are minimal extra buttons needed besides a menu button for unit and inventory management, which would be fairly nice with touch gestures for scrolling.

Thoughts? I just totally miss those games, would love to see them recreated in a modern context for hand-held gaming on a mobile device with some high resolution textures and some more polish.

(13-02-2014, 03:13 AM)venam Wrote: If we are going to make I game I suggest making it for mobiles. We'll need some people that take care of the engine, some of the story, some of the art, some of the audio, there's a lot of area where everyone can contribute.

Why would you suggest mobile?

If we're making a game, I'd be up for doing some art. Always wanted to do some game graphics. And maybe I'll try myself on some chiptunes... I guess that would fit in nicely with some TBS.

(13-02-2014, 08:02 AM)crshd Wrote:

(13-02-2014, 03:13 AM)venam Wrote: If we are going to make I game I suggest making it for mobiles. We'll need some people that take care of the engine, some of the story, some of the art, some of the audio, there's a lot of area where everyone can contribute.

Why would you suggest mobile?

If we're making a game, I'd be up for doing some art. Always wanted to do some game graphics. And maybe I'll try myself on some chiptunes... I guess that would fit in nicely with some TBS.

I suggested mobile because that's where all the new technology is heading. It can be mobile and platform at the same time if we use webtechnology.

I have literally none, but I'd be willing to learn and if I can't contribute code I'll do story or something. Also, while I've never played Fire Emblem, I know what BigE is referring to and it sounds great, I'd get behind that.

I would prefer a Distro possibly a mix between OpenBSD and Arch Linux, it can be called OPENArch and have the default wm as 2bwm; and if that's not a good idea or to much work we could make a Custom ROM for Android, it can be called UBAOS (Unix Based Android Operating System).

Must we take a vote to determine distro vs game? Perhaps a 'pros and cons' list for each.

(15-02-2014, 07:41 PM)DotDev Wrote: I would prefer a Distro possibly a mix between OpenBSD and Arch Linux, it can be called OPENArch and have the default wm as 2bwm; and if that's not a good idea or to much work we could make a Custom ROM for Android, it can be called UBAOS (Unix Based Android Operating System).

Forgive me if I'm wrong, but don't they seem to counteract one another quite a bit? Arch is about bleeding stability, OpenBSD is about packages being super duper secure with good documentation, leading to stuff being quite outdated. What would you take from either project? OpenBSD with pacman? Because in my opinion pacman isn't that great, and yrmt has his pkgin thing coming along quite nicely. Not to shoot you down, just my immediate thoughts on the idea.

(18-02-2014, 09:26 PM)ichiban Wrote: Must we take a vote to determine distro vs game? Perhaps a 'pros and cons' list for each.

Again, off the top of my head:

Game
Pros
* More room for participation
* Probably easier
* More room for creativity
Cons
* Not very 'UNIX-y'
* Falls apart if we don't have any talented artists/musicians/idea people.

Distro
Pros
* More obvious - UnixHub makes a Unix distro
* Potential to be pretty unique, though it'd need some thought.
* If we could get it going, I personally think it'd be more fun and more of a learning experience.
Cons
* Harder, more specific knowledge/skillset is required.
* Harder to come up with something original to give the project a purpose.