Mint Debian

[popsad]

having used Mepis 7 Now since the beginning of the year and 6.5 prior to that i am very keen to try out Mint debian as i have found when running mepis that it was very stable although it used debian etch.

Even during the testing on mepis 7 because it was so good and reliable i ran it as my main distro, but now that i have been using Mint although only for a few days , i am very impressed with it, so if there is an installable version of mint debian i would certainly like to test it out, i am actually using it at this moment.Although from a live cd.

Mepis has given me problems running both ide and sata drives, mainly in the boot menu setup, but when i installed mint i had no problems even booted of the ide drive which mepis could not do for some reason.

[Husse]

The Debian edition is a alpha (and Mint normally never release alphas) made to demonstrate that it is possible to "build Mint" on something other than Ubuntu. It was never intended to be installed, but a guy found it so interesting that he offered to make a Community Edition from it.
That's about where we are now
viewforum.php?f=141

[cmost]

Actually "that guy" disappeared shortly after he announced he would forge a beta of Debian Mint CE. To my knowledge, nobody has heard from him since. Further more, nobody else has taken up the mantle of the Debian Mint CE, though there is a lot of enthusiasm about a Debian flavored Linux Mint. Unless the community can pull together the talent and skills required; organize itself into an efficient team, Debian Mint CE is dead. My suggestion was to work with the Parsix Linux team because Parsix Linux is based on cutting edge Debian Testing but with a modern Gnome 2.20.3 desktop (Gnome is notoriously unstable in Debian Testing.) Parsix Linux most resembles Ubuntu, upon which Linux Mint is based. It's the most logical place to start but apparently I'm the only one who thinks so. Oh well.

[exploder]

cmost, your idea about building off of Parsix is the best idea I have seen in this thread! Using Parsix as a base would solve a lot of potential problems.This approach would provide a stable base, be less resource hungry and fast!

I took at quick look on their forums and there are a few rough edges, it seems to need Network Manager configured. All in all it appears to be solid and might be the starting point the Mint Debian project needs to get off the ground.

It's the most logical place to start but apparently I'm the only one who thinks so

You are not the only one now! You have a very good idea!

[Husse]

hmm
Was it "Zonas" that started the Debian thing?
I think I've seen him contribute not long ago with some artwork...

[cmost]

It was Zonas. He may be contributing artwork but he unceremoniously left the Debina Mint discussion with nary a word! Rude! If he didn't intend to follow through he could have at least explained his reasons why or at the very least documented his work so that others could take over. In any case, he was a "go it alone" sort of guy. Debian Mint should not be run that way. If it's going to be a community edition that it needs to involve the community. No man is an island. As I've stated (and as others will agree) the interest in a Debian Mint CE is definitely there, however, the talent, skills and organization is not. Period. Without that, there can be no Debian Mint CE. It's really too bad.

[sundayrefugee]

Exactly.

Here's the thing.

Mint Debian, IMO, should be just that. Debian with Mint on top.

Now, ironically, making it isn't going to be the hard part, given enough willing victims to dest.... I mean, enough *beta testers willing to find bugs....*.

It's the ongoing maintainance team that makes or breaks this project. Total honesty time here, now:

I could make the initial Debian Mint myself. No crap. I really could. I'm aware of Debian Gnome's problems, and I'm aware of the myriad solutions. I know how to set up repos. I know how to set up preferences files, etc... I know how to manipulate tags, how to pin, etc... I know how to put the Minty goodness into a plain-vanilla Testing GNOME once I have it fixed to my satisfaction, and I know how to make it appropriately Minty, *with ideas from the community* from there. I know how do all of the things to *make* Debian Mint, based on Debian Testing, using a combination of Debian and Mint Repos, and not having to rely on anyone else's work, happen. The process *should* be open, and it *should* involve several people at this point, but I *could* do it myself.

This, ironically, is *not* the hard part, as pain-in-the-ass as that would be

No, the hard part, and why it HAS to be at LEAST a 5 knowledgable hacker effort, plus pre-release designated bug testers, plus at least a couple of coordinating admin, is because of the perpetual borkage possible in testing. Now, it's not like Sid, where you are getting different packages every two weeks. Testing is *mostly* stable. The problem is, immediately after the launch of Lenny, Testing will reset, basically, with a dump of whatever's in Sid. So we'd be starting off with Sid and watching it get gradually more stable again. This is going to be *ALOT* of change. And it would require a large, coordinated effort to triage and bug-fix it to deliver a Level-3 safe MintUpdate bi-weekly.

This is one reason (okay, the big reason) why I was more in favor of wanting to base on Lenny, where the rolling is much slighter, and Lenny, by that time, would be stable, so the updates would, inherently, be Level 1 or 2 safe. And just eleventy times faster than Ubuntu, without it's bugs. So to produce Mint Debian, as it fits with Clems' vision, requires, I estimate, at least a council of 5 knowledgable hackers with enough time to sift through a slowly stabilizing Testing repo to ensure safe upgrades.

As cmost knowledgably says above, until the hacker volunteers are there, (and again, to those who've already volunteered to admin and bug-test, thank you), there's simply no reason to even attempt the distro.

[cmost]

@ Sundayrefugee
Since you and I seem the most enthusiastic about this project, what say you and I do what we can to make at least a workable beta for the community to play with. If we have you, we can borrow the more difficult elements from other distros (e.g., the installer, etc.) Let's communicate. PM me and we'll exchange information. I'm not as knowledgeable a Debian hacker as you, but I've been using Linux for over five years (mostly with Debian based distros.) I'm also an extremely fast learner and an excellent team player. Unless I miss my guess, once you and I release a working prototype, others in the community (or from other communities, perhaps even Parsix, DreamLinux, Mepis, etc.) will become excited and join the effort; a snowball effect with result. Let me know. Cheers!

[carlos]

I hope you 2 hook up I'd like to see it happen.

[ranyardm]

Okay.... very hesitant here....

I've been a maintainer on cAos (http://www.caosity.org) before, and my main reason for getting tired of the distro is because when you're a maintainer you can't just "use" the distro... if there's something missing you want, you feel obliged to package it.... and once you've done a lot of those, it becomes a larger and larger load just keeping the packages up to date.

All that said... I'm willing to join a team working on a debian mint edition largely because politically ubuntu has certain issues and it would be good to have a nicely polished debian for those who are political.

Hackerwise I'm intermediate, and I am a huge LinuxMint advocate (my mum uses it too!). A friend whos skills are more webby than mine is about to be a convert I do believe, but I know he enthused about debian testing when he did his last distro eval; he may become useful

Anyway, I am not very active in the community area but would be happy to sign up to help in the development of this bit.

--
Martyn Ranyard

[cmost]

...and then there were three! Welcome aboard. I agree with you that maintaining packages can become burdensome and that's why I would simply avoid doing it altogether. I would be inclined to Base Debian Mint on stock Debian Testing with Mint tools and a polished Gnome or KDE desktop on top. This will avoid having to compile and maintain scads of packages. This would also stop overzealous developers from straying too far from Debian proper (i.e., customizing Debian to the point that it becomes a fork; like Ubuntu.) If it ain't in Debian Testing, then let users compile it themselves if they want it bad enough. We could always have a community repository for packages that users donate. Anyway, I digress. I have yet to hear from Sundayrefugee as I requested, so I can't say for certain he's willing to assist with this. Meanwhile, I'm preparing to install Parsix on my laptop to get a feel for what will need to be done to Mintify Debian.

[changturkey]

+1, I would really like this to be done.

[sundayrefugee]

I still am not inclined, personally, to make it until I know the infrastructure is there to support it. I'm sorry - I know this breaks alot of hearts. I have a vision, the ability to take on the responsibility as maintainer from Clem *and actually, really know what that means* (<- lot's of bloody trouble, mostly :p, and I have the ability to make it, so I know some people are hoping that I grab this bull by the horns

But I just can't seeing myself sinking everything into a project like this that just goes nowhere. I *have* to know the supporting infrastructure is there *first*. Admins, a suitably large and *very* dedicated testing-group, and *minumum* 5 hackers with repository access who can work slavishly night and day, along with cmost, myself, and ranyardm, if he/she decides to join, to make sure weekly MintUpdate snapshots are flawless for our users.

Anything less, and I can't see the point of starting.

Although, as I've said before, I'm not in this to be a rockstar - I just think that a Debian-based Mint would be so popular, so good, 3 times as fast at half the memory footprint, and 1/10th as buggy as a *buntu-based Mint, that most users would lose interest in a *buntu variant, and so I'd like to see one done. It's the OS *I* want to use. Debian with Minty goodness and without the Ubuntu dirty hacks, bloat, baggage, borkage, and bugs? Heck. Who *wouldn't* want to use that after 3 days with each? cmost, again, if you want to take the reigns and plow ahead into the unkown, I'd be more than happy to help whenever you needed a hand with something

[cmost]

I've installed Parsix on my laptop and I'm now tediously looking to see how many degrees of separation exist between it and pure Debian Testing. As I've said, and I think most mint users would agree if they cared to test the theory, Parsix is close to what Debian Mint users want. It's Debian Testing with Parsix own tools and packages added to the mix. I see no reason why a "mintified" Debian couldn't be constructed in a similar fashion. Debian Testing, and custom mint tools, a few packages, and a mint installer thrown on top. I'm inclined to follow the KISS principle: Keep It Simple Stupid. I'll keep you posted.

[muskratmx]

I would be more than will to be a beta tester, I have three boxes, #1 an IBM pentium M 1.6 laptop with wifi, #2 a celron 1.2 compaq, #3 a pintium 1.0 home made box. Each of these are set up with ability to multi boot distros, so I have to space to install And test with no interuptions of my stable distro, Debian.

But if we're going to do this, lets do it streight from debian and leave out Parsix, Not because of dislike for Parsix, I haven't even tried that distro. But rather for the same reasons to even build another base from Ubuntu. Ubuntu and Parsix are both based off debian, so why can't Mint?

I'm currently running, Sidux and Lenny, Have an etch installed but plan to remove it. And have LinuxMint.

[Fred]

cmost,

Have you looked at sidux? It is KDE and based on sid. Another good candidate, for a KDE version.

Fred

[cmost]

@Fred I am well aware of Sidux, in fact I've used it on my workstation in the past. Sidux is based on Debian Unstable and requires a very (VERY) dedicated team of talented developers to turn something so unstable as Debian Sid into something usable day to day. Using Sidux as a base for Debian Mint would be out of the question for Debian Mint. Only Debian Testing would be manageable.

[Fred]

cmost,

Point taken.

Fred

[muskratmx]

I was going to say the same thing, fred. I have sidux and am running it on my wifes box as well, for how long I don't know. It keeps me on edge, everytime I run a distro upgrade I'm chancing braking my system, and distro upgrade is the only way to get software updates and patches and fixes.

Actually Lenny/Testing is really only a qouple of weeks behind sid. I've installed lenny on one box, I'll see if it takes some of the stress off me. If it does, with less brakage, then my wife is downgrading to lenny next time instead of distro upgrade.

I think I'll keep sidux around just to see what's on the bleeding cutting edge. I did find a very good package there for wifi roaming of open hotspots. It's call ceni. I put it in my lenny install and it works great. They tell me it's sidux package not Sid.

But Like I said I got room and boxes to do testing, if Linux Mint goes to a debian base. Just let me know the ground rules and when/where. I'm not interested in testing anything ubuntu. I'll let others do that. But if we're really going to do this debian base, even a CE. I'd love to do my part. I like Linux Mint, just a little ashamed that it's ubuntu based.

I've also never done any official testing and submitting of bugs, but I do learn quickly.

[Fred]

muskratmx,

There is no doubt in my mind that you and cmost are right and I was being overly optimistic.

I have only been using sidux for about a month; shortly after you called my attention to it. But I have had zero problems with breakage or anything else. I was just assuming the sidux developers were doing the heavy lifting as far as the update breakage was concerned. Then again, I might have just been lucky, with the sky ready to fall any second.

Fred