Hi,

I have been an avid user of Slackware from last couple of years now. Have been a fan of Slackware 14.2 liveslak version. It was the most stable OS I have seen in probably ~28yrs. Atleast for me.

I was trying to build a prototype of a router/gateway + desktop OS because now a days router/gateway are as costly as a computer. I named the project as mhtgw (my home tech gateway).

there are some constraints though because of which I am planning to use slackware64-14.1 version because I was trying to keep the most simple version of a desktop. not trying to make it as a true desktop with all bells and whistles.

That may require lot of burning hours from me because I was planning to do it as below:

It would be a merger of AstLinux (router/gateway software) + Slackware (core build) + Linux Live-Scripts (aufs based).

Keep the slackware64-14.1 version of minimal base + bump the desktop to higher versions as needed. if it is required then i may have to bump the minimal base too.

Also I might need some guidance and coordination so I can host it on slackware and get your support as well since I have not seen any project similar to it.

https://slackware.uk/

Please let me know if there are any guidelines to follow.

This is the project site
https://github.com/mhtgw/mhtgw/

Thanks
Ramesh GK

cool~

Might I ask - who would want to run a Linux desktop based on Slackware 14.1? That release has been declared 'End Of Life' per 31 December 2023, meaning it will stop receiving security updates per that date. It will only become more vulnerable to attacks over time. If yours is a new project, why not target Slackware 15.0 with XFCE as a desktop?

My choice was initially Slackware 14.2 but I am trying to with Slackware 14.1. The lower version we go, the complexity is reduced but the maintenance may be slightly high which is where I am looking up for Slackware newer versions.

The base will not change such as kernel, toolchain, etc but the software stack will be based on the least version of Slackware that is workable. It is something like you are purchasing a dog/home and fixing it whenever there is an issue. If I have to look up to take up the next step, I still have a Slackware next version.

I agree that Slackware 14.x is going EOL. but I still do like the Slackware 14.2 XFCE version. will try to manage with it as much as possible as that has become my primary desktop. and also please note that AstLinux approach packages everything as one layer like liveslak but only allows you to write to certain folders only which are persistent. so there is less risk of malware persisting beyond the system restart. It is just my thought.

Lastly, I tried to look at the desktops of users who are posting it in Slackware and 50% (probably) are still using older version of the desktop even though the software is newer. It is more psychological that they got glued to the dog/home that they started with.

In answer to post #1:

You're misguided using 14.1. You're thinking simple, but slackware hasn't got more complicated, and you're building in huge problems. For instance the old kernel headers won't build modern packages. The old glibc won't allow anything built against a newer glibc to work. And there's more, but that's enough to kill your project right there.

Go to 15.0.

He did say prototype, so the exact version does not matter if he's just starting out. Personally I'd go with 15.0, and downgrade XFCE. I am using XFCE 4.12 myself on 15.0. But you can go with 14.1, update as far as you can the kernel and other required network applications.

OK, but why does the wheel have to be reinvented?

Stock Slackware comes primed to be set up as a router/gateway, all you need to do is make a few minor tweaks to get this working: https://docs.slackware.com/howtos:mi...uick_and_dirty

OP would most likely be best served by installing Slackware 15.0 and tweaking it as required.

Slackware-14.1 is fine by itself. But if you read post #1, he's adding in stuff

• If he adds in compiled stuff, he'll be caught out by the old glibc
• If he wants to compile, he'l run afoul of the ancient kernel headers and libraries.

So it's extra work with no gain. Certainly not simple.