I don’t disagree, but we’re talking about a few packages overall on any given distro. Granted, some of these are very central to how the distro works (e.g. Systemd, and I am sure there are others). If a system is running all packages with the Unix philosophy in mind except Systemd, would you say that distro needs to go to the drawing board and rethink all of the software they are using because they aren’t following the Unix philosophy close enough? Obviously not, in this (admittedly contrived) example, there is one piece of software that is a concern. That is my point.
It is not just the accountants. I can speak from my experience with my manager - if I want to refactor something, he often says no unless it will give us additional benefits. Why? Because his manager expects us to produce certain output, and if I am refactoring something then I can’t do something else. Although, in the end, good written code should be THE priority. In my particular case, my team is fortunate that most of the code we have that needs to be refactored isn’t very important and is from the early days of this team (although when we do need to take a look it is kind of a nightmare).
I am sure at least some of them are aware of the problem already, where there is a problem. I would image most corporate-produced FOSS/Linux software is probably already following the Unix philosophy, even if not intentionally. Just the cases where they aren’t, they aren’t being given the resources.
It is good to hear from someone who is working in that
commercial world today.
You view is a little less pessimistic than mine. If Linux is mostly good clean code (I believe the kernel is, for example),
there is some hope we can isolate the poorly coded bits. There is much less control over what gets included in distros, compared to what gets included in the kernel. That is where we need to press for change.
I do believe that most innovations come from deviant individuals rather than from communal efforts. Several really good distros have been started by individuals. It is quite difficult to influence those sort of people… they tend to be rebels. One way is by doing something that conveys a warning message… like Hyperbola. It is like a canary in a coal mine.5
I don’t disagree, but we’re talking about a few packages overall on any given distro.
If you believe what Hyperbola says, it is 90% of the code… by ram usage anyway. That is because it is the big clumsy monster packages, like firefox or systemd, that are the worst offenders. Dependency problems grow exponentially with package size.
. . . but, that’s my point. Something like systemd is very fundamental to how the kernel interacts with the rest of the system, and it’s not even modular in design. It’s an all or nothing proposition. I want distribution creators/designers to be able to pick and choose the parts they combine to create their distributions. If things like systemd were designed in a modular manner so the boot loader could be combined with some other system service manager or vice-versa, I’d feel like it’s at least trying to adhere to our philosophy.
GNU/Linux’s underlying philosophy of doing one thing, and doing it very well, then combining small things to do big jobs made our OS better and more secure than the proprietary alternative(s) right from the start. I just don’t want us to lose sight of that philosophy because it has worked very well over the years, and I think that’s important. I think our philosophy makes GNU/Linux great.
Now here’s the spark of a really interesting project. Maybe assign a system of priority groupings to software?
I–runit, systemd, terminal/console
II–window managers, DEs
III–production–Libre Office, GIMP, browsers
IV–secondary production–Rhythmbox, Clementine, VLC
I’m just a basic user, but I could probably pick a set that would work for most folks. As an example that has really worked, the software packaged in most Chromebooks could be a basic (if limited) and universal Linux.
Now that you’ve stopped laughing, let’s see a few suggestions of a universal Linux distribution.
It is apparenly not possible to have commercial software development driven by a philosophy.
If commercial software were never made into Linux packages what would be left in Linux?
I think Hyperbola answers that question.
I get it… you mean the alphanumeric screen or console… what used to be a teletype.
Not sure about your categories.
Another factor is whether there is an acceptable alternative package. If there is no alternative, i think a package is a high priority for a rewrite.
That puts systemd at a lower priority, because there are several perfectly satisfactory alternative init systems. Sime would say the alternatives are actually better.
Sure, let’s call it the console–the black screen that replaced the teletype. See, I AM really old.
I’m not really clear about the differences between systemd and the other init schemes–a fast boot and glitch-free, snappy interactions are sufficient for my needs. I’m going to side-by-side Peppermint (Debian) and Peppermint (Devuan) for a week or so. I think they should be pretty similar except for the systemd/runit difference.
After just a little research: I’ll run the latest Debian and the latest Devuan, both in XFCE, to compare init performance. Never hurts to verify my own mistakes.
So I’ve read through I couldn’t tell you how many post and replies, and not just today either, and some days I feel like I am finally wrapping my head around it and others I still catch myself comparing Linux and/or Unix to windows or Mac which isn’t really a sane thing to do. That’s all I’ve ever known though was Windows until I bought a Chromebook not realizing that I had picked up a very fairly priced Linux machine, albeit somewhat left wanting at times. That is though where I think Linux is really going to make up some ground for me. My Chromebook has reached its expiration date and God I never thought I’d talk about a laptop expiring about the end of last year so it’s about a full 12 months past it’s AUE and is no longer receiving updates so it’s not not keeping up with current security protocols and it’s usability is diminishing quickly so I’m going to put a Linux distribution on it I’m just not sure which one I’ve narrowed it down to a small handful and I have information but I wouldn’t know if I would consider my opinion to be one that’s informed so I’ve come asking for help and after reading through your post this seems to be the most comprehensive gathering of like minded individuals focused on a single topic albeit several different varieties so I would really appreciate someone’s advice while I finish backing everything up and making a disk image that way I’m ready to wipe her out and ideally I could do a boot but I don’t know how that’s going to work because this is all new so thank you in advance for any help that you all can send my way
Hi @hagen and welcome,
I think this should probably have been a new topic, but it does not matter.
I dont know specific issues with putting Linux on a chromebook. Someone with Chromebook experiece will give you some tips on that.
On choosing a Linux distro, if you are new to Linux choose one of the well known distros that are easy to install… like Mint, MX, Ubuntu. Definitely not Hyperbola. You will get better help with a well known distro.