basically you have a lot of installation where the bus factor* for administration is about 1?
(but for installation where people are fare away you recommend some kind of Ubuntu, as you mentioned earlier, (bus factor applied already, use other system?).)
*http://en.wikipedia.org/wiki/Bus_factor

I understand this form a business point if view, put I think it is possible not optimal for the customers/people that depend on you ;-)

I don't think this graphical package manager on top of pkgtools has been mentioned yet: http://qtgzmanager.wordpress.com/

There's a SlackBuild for it on slackbuilds.org.

Eric

2 members found this post helpful.

Yes. Which is why I prefer the woods to the highway for my daily morning run.

Hi, I've given qtgzmanager a try, it seems that the last version I found is for slackware 14.0. This patch adds sl14.1 support https://www.dropbox.com/s/tc78l1bmfy...1-did.patch.gz
I've made some quick tests, seems to be working ok. So far I've used it mainly as a browser for local packages and it looks promising.

I have a copy of a really obscure front-end for pkgtools -written by Mark Tyler, the author of mtpaint. The package is called 'mtsuit' and can built for use with either gtk1 or gtk2. One could, of course, also use gtkdialog to put together a GUI for the existing tools.

The problem of dealing with dependencies remains, though, since the package format itself is not capable of handling them at all -not even of providing information. All the external efforts such as slapt-get, sbopkg, etc, all suffer from the lack of information *contained within the packages*. sbopkg can help to extend the system, but no one provides any functionality which will let you cut the system down. The standard advice of 'only full systems are supported' is, frankly, lame and unhelpful for any special needs/desires.

1 members found this post helpful.

Here's a checklist I would follow to see if Linux is even an option for your customers.

1) Iron out the customers' hardware requirements. Is the OS required to work with poorly supported wireless cards (where you need to install ndiswrapper just to get things working) or printers/scanners/faxes? Will the proprietary drivers cover your customers' older hardware or would you have to utilize open-source drivers with reduced performance (i.e. nvidia vs. nouveau)?

2) Assuming the hardware isn't an issue, what about software requirements? Will lack of a native silverlight/activex/.net environment be an issue? While I run Wine and Pipelight for various things, will this work for your customers' needs? Would a non-networked Windows XP VM be a viable solution for older applications?

3) More about software: will customers be happy with utilizing Firefox/LibreOffice/Thunderbird? Which desktop environment would work best for them?

4) Enterprise integration. Do you have any requirements for Active Directory or MS Exchange? What about VPNs?

5) Maybe take a look at the whitepaper on the Munich Linux integration. That took them several years and they had to customize a distro to make it work, although it was a much larger scope.

6) Repository considerations: you may want to choose a distro where the repository has a larger maintainer base and more packages, like Debian stable + non-free + multimedia.

7) 3rd party support. Assume at some point you may be out of the customer loop. If they have a heavily-customized Slackware install, what 3rd party will support them if something goes wrong? You may have to go with Ubuntu or OpenSUSE if this is important to your customers.

I would encourage you to branch out here distro-wise and find the best of what Linux has to offer point-and-click users. While Slackware could be tweaked to meet your needs, it is by default tailored to users comfortable with the command line vs. a distro like Linux Mint/OpenSUSE/Ubuntu/etc. You could put in the effort to tailor it to your point-and-click users, but in the long run it would be less work to use a distro that tries to do this for you. Also, what tends to happen is that the more customizations you do the more likely updates are to undo them. Finally, Linux Mint/etc. have GUI interfaces for a lot of the things point-and-click users want, and they may have quirks if you try to utilize them in Slackware due to different filesystem layouts, tool choices, etc.

BTW, Linux Mint has a version that pulls in Debian's repos not Ubuntu's. You may want to check that out.

If it was up to me, I would demo OpenSUSE, Linux Mint, and one of the Ubuntu variants and see what your customers like best.

2 members found this post helpful.

I've followed this thread since the beginning and now, since Woodsman seems to be in complete earnest and not just some troll exploiting "shortcomings" , I've gone back to revisit the prerequisites. My conclusion is that it is basically an attempt to create an OpSys as dumbed-down as Windows, where users are as free and powerful as Slackware users (and without any reliance on the terminal) yet ignorant but without any of the rules and limitations that MS has found they need to minimize tech support...Plus all built from Slackware because that would minimize Woodsman's learning curve. This, at least if none of those terms are negotiable, I think is impossible at this time and for the foreseeable future.

This, being a job relegated to the bootloader should be easy, even if it meant creating a customized install media where some (replaceable) splash is default. Many distros already do this, just not Slack.

Of the above, #2 is a deal breaker for Slackware. I think this because Slackware, as a whole, has spent very little resources even attempting to do this, and I think rightfully so. Other distros, with large dev teams and bank accounts poured massive resources into this for over a decade and only recently have something remotely solid as well as convenient and it depends on large fundamental changes to the base system and dependency on large, well-maintained repositories (thousands of packages). and #3 implies some level of "phone home" regular polling overhead at both ends, that at the very least requires considerable man-hours and/or software development with a sharp eye on security as well as viability.

Not even Mac has eliminated the terminal, they and Windows just hide it from casual users... which begs the question "Why not just sell them Mac?" since it does much of what you seem to require.

If it is because you want more involvement for whatever reason and you're more comfortable in Linux, I urge you again to try out SolydXK. It has none of the Canonical "features" you don't want and satisfies almost all of your requirements handsomely. I frankly don't see why you care if your customers have Akonadai running or not as long as the overhead is reasonable, which it is these days. Who cares if the average user finds no use for it as long as the whole does what you want for others? If for some unfathomable reason this is still a deal-breaker for you, then use the Xfce version.

However, before you judge and if you aren't motivated to just try it, here's a video from Linux Action Show that shows both versions in action. It is easy to see how minimal and lightweight the defaults are even in the KDE version. If you want to skip to the SolydXK portion of the show fast forward to Timestamp 36:40 but don't miss the default Greeter for newbs at 47:00.

The show also has a cool portion on asciicast/asciicinema a service which allows small videos journaling terminal command progressions and posting them online for linking, great for terminal tutorials and problem solving in forums such as this. There is also a short (including a One Finger Salute to Chrome) on the Unreal 4 Engine (3D) RUNNING INSIDE FIREFOX!!!

#Note - sorry for Bold...3D in Open Source Browser and a sloppy Bbbbblllltttt!! to Chrome pushed my adrenalin up

The video is Here is the LAS featuring SolydXK

The bottom line is that this is like falling off a log for a Slack user. You won't find a learning curve here and neither will your Windows user clients beyond just a few minutes worth.

Disclaimer: I love Slackware and never want to see it be this "convenient" but I'm recommending SoldXK because some people value convenience most of all and basically want "Turn the key, push the pedal, turn the wheel" level of interaction.

Those are reasonable discussion points, but would the same list apply to any person wanting a Windows system? Unlikely. Customers only want to buy a computer.

That said, I mentioned in a previous post that hardware peripherals is a concern for me. Providing a working preinstalled Linux based system is nowhere near as much a concern for me as guessing what hoops I might need to jump through to have existing peripherals function correctly.

Enterprise integration is not a concern and is not a goal. None of our immediate future customers will be enterprise users. At most we expect to run into only small business owners. No business users are yet in our radar, but as my best friend often quipped, those are problems we would like to have. I see a lot of long-term opportunity with XP now at EOL.

By the way, we now have prospective customers. I was told two days ago that three people want to migrate. I have been asked to ramp up my learning and testing. These are people who do not know what to do about the XP notifier that keeps reminding them that support has expired. Fortunately, these early prospective customers do little more than surf the web and check email. Many prospective users use web mail, which is good because that reduces data conversions.

As I shared earlier, few of these people are going to be computer savvy, geeks, or power users. By far most are ordinary point-and-click users. The plan is that through word-of-mouth advertising the local Linux support will gain speed. After a dozen customers or so we should have a much better idea how to deal with the different levels of usage.

I am anticipating --- eventually --- users who have specific Windows needs. For those with the hardware we will create a VM using the existing Windows partition in raw disk mode. Or let them dual boot, whichever they are most comfortable. For those who do not have the hardware and do not have the budget --- and there are many on fixed incomes in this area --- there will be no quick or easy answers.

I expect whichever distro we settle with will be customized by us to one degree or another to help the customers and help our own maintenance efforts. If all goes well, will be interesting to see how many of our local tweaks get merged upstream or whether we end up with a secondary "custom package set" of our own.

Currently we are leaning heavily toward LMDE. We do not want any 'buntu base, we do not want semi-annual updates, we want flexibility if the distro folds, we want a large repository base, we want a distro that is heavily user-centric, and has most of what we already want for end-users. LMDE fits that description. PCLOS was another choice, but we decided we don't like the future if the distro folds. With LMDE we fall back to Debian.

I looked at trying to mold Slackware but I kept feeling like I was pushing a size 9 foot into a size 8 shoe. That is, "so close, but..." and "almost, but not quite...." I have been using Slackware for more than a decade, but I can't --- and won't --- force my geek comfort zone onto point-and-click users. Perhaps someday I might think about my own Slackware derivative distro after I gain experience dealing with users and their needs.

Regarding third party support, that is another reason we want a flexible user-centric distro with good existing upstream support. Should we disappear users still have some place to seek help.

Gee, after using Slackware for more than a decade, and with Slackware the only OS I currently have installed on all of my systems, I hope nobody would think about calling me a troll.

That would be nice --- for me. However, for the past week since I started this thread, I have been required to start thinking like a business person rather than as a computer user or geek. Minimizing my learning curve has to yield to happy customers. I am not feeling burdened by learning a new distro, new desktop environments, or package managers. I can handle that. In the end all that counts is those first few customers we help, which in essence will be guinea pigs, are happy enough to spread the word. My learning curve will then keep pace with customer demand.

This is our plan too: just hide the terminal. There will be no terminal icon in our default panel. That said, I anticipate having customers occasionally using the terminal when they call me for tech support.

Yes, and that is what will drive our decisions. Much easier for me to adapt to a new distro than for me to customize the hell out of Slackware or derivative distros and then customize some more. Customers want to drive --- they don't want to change the oil. I am a geek at heart but I accept that usability level.

Regarding this statement and my original post, perhaps someday I will learn how these GUI tools can be migrated or adapted to Slackware. Slackware is not going to disappear from my house because I have to support a different distro.

This is definitely foreign territory for me as well. Most of the decisions I would make for my own systems are completely different than the ones I'm suggesting here. I'm used to building Linux from the ground up, choosing the login manager (if I even have one), desktop environment, programs, kernel options, etc.

I do have one small success story: I replaced our Windows 7 media center with a mythtv box. After I set it all up, got Netflix working, tweaked the IR remote commands for VLC and mythtv, and customized the menus in mythtv to launch VLC/XBMC, it got the seal of approval. It was definitely some work though.

What you're attempting is tricky as you will definitely have to manage user expectations as well as deal with hardware/software issues. I think the first step would be to let them know some of the limitations: they need to know that they won't be using Microsoft Office and that they won't be able to buy any $20 Lexmark printer and expect it to work out-of-the-box. I would probably start by having them plugin all their peripherals, run a live CD, and see which hardware Linux will support (kind of like a Windows 7 migration assistant). I believe most distros have something like this on their Live CDs. For hardware/software issues, it will be tricky to estimate the time it would take to resolve issues: if it's an easy one like installing some firmware for a specific driver, no big deal. If it's a more involved issue, you have to estimate the time well so you don't end up charging more in labor than a cheap Windows 7 PC would cost them.

You will definitely need to customize any distro you pick, even LMDE. You probably want to put some kind of company contact info in the boot loader screen so they can contact you, create a restore disk, pick which programs you want installed by default, automate the installer process, etc. If you want to get slick you could create an initrd that runs one of these hardware compatibility checks and if they hit OK will automatically install the distro you've customized for them (after asking if they want to keep existing partitions, etc.).

I like your idea about converting their Windows install into a VM: that sounds really slick. Please keep us posted on how your experience goes.

I started a project journal/diary.

You have said this a couple of times, so may I ask a friendly question? Since Ubuntu is a Debian derivative, what is so different about it that you will not use it? Just curious. (Remember, friendly question. )

One reason for the question is I was just looking at Zorin and if there was ever a Linux distribution made for XP users, Zorin is the one.

I think Zorin is used by Ken Starks in his computers-for-kids foundation.

From my perspective, Ubuntu was a Debian derivative, despite the original commonality with Debian. I just don't like the overall Ubuntu philosophy. I also do not like semi-annual updates, which leaves LTS releases as an option. LTS editions grow stagnant with respect to the speed at which free/libre software evolves. I don't want bleeding edge but I want access to regular updates. LMDE seems to be a better fit for this business.

An exact "XP experience" is not one of our goals. Instead we are seeking a "classic desktop" migration experience. KDE is too heavy for the older hardware many of our customers own. Older hardware and fixed incomes is a primary reason many have not moved from XP to newer versions of Windows. The sole criterion of older hardware pretty much reduced my choices to Xfce or MATE. I like Cinnamon but the 3D acceleration requirement is hard on people with older hardware living on fixed incomes.

There is no right or wrong here, just which pair of jeans feels more comfortable. I don't pretend to know everything. I am just trying to make reasonable decisions as fast as possible. We already have five customers waiting for me to finish building the demo system so they can start experimenting. Time played a role in my decisions here.

I agree ever since post #5.