End goals/objectives?

Maybe I’m not remembering something that’s written already, but I’m trying to find what I think is commonly referred to as an ‘end goal’; even as a generalized term.

First, I should say I’m in support of this project.

Based on what I remember reading (paraphrasing here) ‘elive is a work of art’.

But does elive have the ambition to become ‘a masterpiece’?

I’m of the opinion that art becomes a masterpiece when it becomes timeless.
In order for something to become timeless it must not only become well known through social media, but well known as something great, and not something that fell short of greatness.

To this, I asked the question will elive continue to play ‘catch-up’ to the latest Debian or Ubuntu stable releases, become another fork in the niche abyss, or is there something more?

There are many opportunities for diversion from the norm & my question is how will it approach them?
Will it take a stand against the new launchd/systemd newage, or choose to conform?
Will it look closely at kernel security patches such as grsecurity/pax, where others have mistakenly shunned?

We know where the future is going because we’re creating it (humanity/our generation), the question is “what part will elive play in that future?”

1 Like

First Of All:
Welcome in the Elive community.

Your contribution is on the point.
It needs a detailed and well-founded answer.
Actually we are here currently in discussion about the topics you mentioned.

I assume that Thanatermesis can not resist to answer you directly and that you will be included in the discussion.

Forgive me for my bad english.
(I guess that you are a native english speaker ?!)

Welcome @cisc0ninja

I'm giving my own opinion here (of course), @Thanatermesis will certainly correct me if I'm historically wrong. :innocent:

I suppose Linus' answer "world domination" might fit. :nod:

Which seriously is just a joke. Most programs and distros (and even Linux) start out as "scratch your own itch" projects and then get a life of their own. Morphing into almost organic like growth and along the way, becoming its own goal.
Setting a strictly defined goal creates limitations IMHO, something that can stifle creativity.
Community input and demand should be what defines next steps, once it takes off. An end goal literally means finishing, something practically impossible. There's only failure or continuation to chose from.

it is not? :happy:

well, it has the ambition to be better every day / release, and "utopia" doesn't exists, utopia is just a constant improvement (everything can be always improved)

The next alpha version 3.7.x is based in debian buster (which is the next debian stable, so not even released by debian yet)

About "more", Elive is a customized debian, but a very customized system with lots of extra features and an own way of work with its very specific desktop and behaviour, in other words Elive is not just a "debian with enlightenment" (as I have read some comments on internet saying that), and that's why the website (try's) to explain this point in the homepage (check one of the first accordions that says that sentence)

Not sure if im answering what you are asking exactly :slight_smile:

I'm trying to make the next version compatible with sysvinit, but after some tests its very difficult, so too many things today depends on systemd, for example you can't install network-manager without systemd (duh!), and some daemons / features strictly depends on that... so the new alpha versions will have by default systemd while I will still try to make it working with sysvinit (seems like systemd also consumes much more ram), but I cannot say yet if will be possible to use sysvinit or we need to stay at systemd.

If in the end is possible to use sysvinit, the ideal goal is to be able to select between systemd or sysvinit in the installation, so your resulting system will use the one you choice

All is included by default (debian security & updates repositories enabled by default), the problem with 3.0 is that is based on wheezy repo's which are not maintained anymore by default, so those security things cannot be updated in 3.0, but will be in the new versions of Elive (based on buster, which will stay for some years)

An alternative OS focused on usability, userfriendly, lightness, and with some things "in a different / better way to do it", since knoppix appeared to the world I disliked common desktops and how things are set up, that's how Elive was born in its early stages

2 Likes

agree :nod:

:boogie:

Thank you to everyone, especially Thanatermesis, for the replies!
My questions were basically answered by the replies but a few things I wanted to be more specific about:

  1. Regarding systemd vs sysv , if staying with the debian mainline/mainstream then systemd is probably inevitable. What I should have mentioned was whether or not alternatives such as Upstart would be looked at?

  2. In regards to Grsecurity/PaX, this has never been part of the mainstream Debian kernel patches & has recently gone commercial. That said, I believe the there are older versions which still may remain free for the linux-3.14.51 kernel & so I wasn't sure if this was an opportunity to branch off from the latest stable Debian into a more self-supported model?
    Are there any tools/tips/tricks which would make custom kernel compilation easier or harder on elive?
    Many users, especially security conscious people find this statement highly controversial:
    "Due to how internal security works on Linux, it is almost impossible to get a virus."
    Kernel hardening has been proven to have the capabilities at stopping entire bug classes from being exploitable. While I wouldn't consider myself advanced in this area, I can usually point in the right direction for further discussion on this.

  3. I would agree that I too dislike common desktops. But with that said, I'm still of the opinion you should both have & know your 'target audience'. From my perspective (which may be very niche & not realistic) I see the following audiences in the Linux community:
    a) Gamers - the more fps we can squeeze out the better games are running these days & steam support is a major factor (don't remember if this is supported yet or not)
    b) Developers - I need to look into how many & what type of both compilers & ide's/editors are in your list of supported software; off the top of my head Golang has been quite popular. (we all know real programmers use vi) :smile:
    c) Security enthusiast - while kali linux mostly has this covered it I believe it defaults to the bulky gnome/kde interfaces. I've always thought the slimmer you can reduce system resources the more resources you can dedicate to your toolsets like password crackers.
    d) Infrastructure guys - (more of a recent dev-ops thing, with ansible, terraform, docker (containers), etc.. for instance if I wanted to either deploy elive on a multi-user scale (everyone in my home or small business) or even if I just wanted version control around changes that I personally make within my workstation; is any of this native or can this be more automated/user friendly?
    Maybe instead of stating it's impossible to obtain a virus, you took the stance of "you can easily roll your system back to a previous version & have file integrity checking to help mitigate the chance of a compromise"?

While I haven't fully explored all of the capabilities within elive just yet, I can say there are features I've been hearing about within other OS's that may be beneficial to elive if you choose to look in that direction.
An example would be system76 Pop!_OS's cuda compatibility, for those looking to develop applications which may be reliant on this.
I was honestly thinking of buying one of their higher end laptops & putting elive on it. :laughing:

Ooof, a lot of insider stuff,

= highly appreciated.

Am affraid that Thanatermesis could discuss with you for days and lose the perspectives for the 'daily business'. :wink:
From the stuff that you mentioned above
I can not really enter this discussion cause
I am not a dev or programmer - but am one
of these guys whose using that stuff every day and have to deal with it for getting things done - somehow.
Since we {Elive} (yes, YOU too, my friend!) want to make the 64bit happen, we all have to support Thanatermesis with what each of us ist good in.
Obviously you have a deeper knowledge about the things under the hood.
And so you could help in a constructive way.
Please make your mind If you want to
and then reply to Thanatermesis, he has still a lot of things in his ToDo agenda
where you can assist for sure.

For speaking technical -
init and upstart are nearly dead on the 64bit platform, so no need to discuss it,
Debian leads anyway.
Much more interesting is the fact that the other distros all drop the support for the 32bit platform.
Here is our field of the game.
Therefore we should discuss about:
init or upstart, monolitic kernel ... and so in.

Actually Thanatermesis has to ride two horses at the same time, which is not easy at all.
Additionally with a lightwight purse in the pocket,
so you can imagine,
support is welcome.

Play Back, when you are ready

Kdrds

1 Like

I have to disagree that Kali has security covered, on the contrary. Due to its pentesting goal it actually is more vulnerable and was never really meant to be permanently install able.
Others would say Tails but personally I have a strong preference to using the route QubesOs takes but then without KDE. Their downside is that one needs hardware virtualization to function, though.

I intend to create a similar setup with Elive (once 64 bit is there) for myself, just as proof of concept but ........ highly suspect I'm going to be strapped for time.
How hard can it be, heh? :madness:

As for your other a-d points: I think having sort of modular options for specific target installs like graphics, music editor, movie-maker, programmer, gamer, servers (xbmc is already there) might be an addition later on.
Each group has it's special quirks and must-haves with some so specific and niche, it would be madness to include them in one big bundle. Don't forget that Elive's first priority is to be a live system running from a USB stick.

I have 3.0.x running on several high end Thinkpads .... it flies!.:eyepopping: and I have high expectations of the coming version allowing 4.x kernels. Specifically for my X1 carbon 3d gen which has some eccentricities, hardware wise. :slightly_smiling_face:

2 Likes

Sorry, I misspoke, I agree with you Kali is for offense not defense. In fact they shipped with a vulnerable version directly prior to one of the Defcon conference where mostly everyone running it was able to be hacked.

That's a good point you made around elive's focus being on a "live cd/live usb" distro, I had honestly only thought of it with regards to a permanent install on some of my systems. I'm unsure of if continuing down this path would be the correct way for elive to go?
The more IoT coming into the world, the more support for various types of hardware could be bulky. While my say really doesn't matter, I would love to elive be adopted the framework for all Linux OS's moving forward, with a structure around being light & stable, yet powerful. It's what Linux was always meant to be in my opinion.

Well there's plenty of light and stable distros but most will do so by sacrificing looks. Elive's real strength lies in its beauty and completeness on top of that.

If you are serious about security (and don't mind the KDE clutter) I'd advise taking a look at QubesOs. The installation takes up some time but once done it's fairly easy to maintain. As said, it does requires hardware virtualization and thus a 64 bit kernel.....strictly speaking one could run 32 bit virtual machines then and subsequently parts of Elive 3..x in a similar fashion. I know I don't have the time to get that running and have some serious doubts on the uselfulness, but still. :innocent:
QubesOs is even seriously working on the "evil maid vulnerability". :nod:
I'll just take along my USB with Elive when travelling.


Here's a pdf on how its built/
https://www.qubes-os.org/attachment/wiki/QubesArchitecture/arch-spec-0.3.pdf

Actually I see hardware (and closed binary firmware) builtin backdoors as a far greater security threat. Next to free and libre software we should be vying for it to be installed on free and libre hardware.
For now, older known hardware might turn out to be far safer than all the new stuff. :thinking:

Check: Systemd VS Sysvinit , specially the last comments with the results of a test

Also, anybody wanting to do tests on an installed system to betatest replacements are welcome :slight_smile:

No, just the same way as you can do it in debian (following debian howto's), but note also that different versions of kernel would have a different structure of things, I mean, things that changes, and so uncompatible with different versions of applications / system

Steam included in the next alphas, already tested here :slight_smile:

About the virus topic, this would need some investigation, unfortunately this sentence causes controversy, but basically a virus is a piece of software which replicates itself and contagies other computers, this is simply not possible in linux systems because of how the system works (you know, the system is well architectured and you need permissions on the system to do specific things, etc, unlike windows where is just a false idea about a system with permissions), NOW, if you are able to exploit a bug in the system and reach root access and similar things we are talking about another thing, we are talking about an issue in the system, a bug, which eventually can be exploited to reach a goal, but again, a bug is a bug, if you have a bug in your filesystem it can also destroy all your data, or make your cpu burn. Basically that sentence could mean that a linux system is inmune to viruses just because of how the system is made / structured / designed (period), while as comparing to other systems (windoze) that has not a security structure that prevents specific things to be run. (and again, a bug is a bug, and can happen anything... except space time traveling, probably)

But how to describe all this virus topic in a single, short and concise sentence? :slight_smile:

BTW, your idea about "for multimedia", "for security", etc.. specific purposes doesn't sounds too bad to be implemented for the installer, I mean, not possible for the live mode but for the installed it can be as simple as a selection of extra packages to install, whereas debian (which is based) is probably the kind :slight_smile:

I mean, since there's a few users that could have specific needs for the OS, could be nice to allow them to being able to have these features :thinking: for example @triantares with the example of QubesOs, if he's able to setup this system on Elive 3.7.x, could be replicated selecting this install profile from the installer, and even if is a slow process it can be amazing to have these things in one shot

Alright, its simply not possible to include it "as default", only as an detached "extra"

I just wrote this draft suggestion Installation profiles, full modularity, specific purposes :surprise:

@triantares @Rebel450 @jfbourdeau @cisc0ninja

IIRC early Slackware (not sure about Debian) had those options at install ....... including whether to use X11 or not. :smiley14:

Oh, that is possible in debian too, in fact the installer of debian works on this way, you select what you want to have and it installs a metapackage which installs other packages, they are called task-something, for example if you want to have a gnome desktop (which you can also do in Elive), you just need to run "api task-gnome-desktop" :slight_smile:

But I was talking more like customizations / recipes made by the users, check the details in Installation profiles, full modularity, specific purposes