Saturday, February 12, 2011

Great article about things Linux needs to change

Here's an article I found a while back on TechRadar about things that they would change about Linux. As usual, I don't agree with everything they say, but (also as usual) I do agree with most of them.

24 things we'd change about Linux

The Linux annoyances that really needs sorting out.
By Paul Hudson

  1. Fix Sound Once and for All
    I've never personally had this problem (at least on the input side of things), but I've heard about it enough to agree. I mean, call me a "user", but sound really isn't something that someone should have to worry about when setting up their PC.
  2. No More Infighting
    GOD YES. Seriously. Of course, I personally believe in what I guess I'll call the "Linux Unification Directive", in which all of the vast, excessive distros would merge into one (or even less than 5), but that's a far off dream. Nevertheless, until that day, people should stop fighting. It's ten times more stupid than the PC vs. Mac debate. "My Linux kernel with applications is better than your Linux kernel with applications!" Pfft.
  3. Guaranteed GUI fallback
    I'm not sure what he meant by "Guaranteed GUI fallback", other than "X works", in which case, I agree.
  4. Andrew says...[one package format]
    GOD YES. Seriously. All packages are are TARs with a little text file that says "put this here, this there, etc". Sure, there are differences like that Yum checks the server for every request and Apt only checks occasionally, but still: overall, they do the same thing. And yes, some people (mostly the nerdiest of nerds) probably have reasons of why they use this over that, but the point is that the majority of Linux users (or would-be Linux users) couldn't give a rat's ass about it. Please don't infer that I always jump to the "hop on the bandwagon, nerds!" defense; I hate that. I'm simply looking at it from a logical perspective for a unified package manager:
    • PRO: Only one type of package needs to be downloaded for any Linux OS. It's simpler for developers and users. Possibly will free up the people working on the non-chosen manager to work on other projects.
    • CON: Um......a few people won't like it.
    (*Side note: for the people that are really that picky about the package manager they use....why not just compile everything from source? I mean, I think that's what some distros do anyway, like maybe Gentoo or Arch or something...)
  5. Backward Compatibility Dependencies
    He makes a very, very good point. You know how pissed everyone got that apps for Windows XP couldn't run on Windows Vista? We want to avoid that. For the most point, a regular update to an dependency means that it is still the same dependency, just an update; therefore, if it's still the same dependency as before, it should still work with the same apps as before.
    Yes, I know there are exceptions. Sometimes when apps or things undergo a large update, they make it incompatible with previous versions. But that shouldn't be the normal case.
  6. Get Mac Compatibility for Games
    I have no clue how plausible this is, but I'm in full support of it. (Of course, if Valve were to just bring Steam to Linux, that would makes things a great bit easier.)
  7. A Single Name for the Wastebasket
    This really is silly and trivial. But it's one of those things that just should be fixed. Not high on the priority list....but come on.
  8. Easier Driver Install
    I would offer an opinion on this, but I really have no knowledge. I actually was talking with my brother the other day about Linux and he said "How does Linux do drivers?" and I just responded "I have no idea." Other than being able to download proprietary drivers for things like graphics cards, I have no idea how it works.
  9. Guaranteed sleep/hibernate
    A very good point; whenever I was running Ubuntu on my netbook, it would not sleep. I can remember one time I forgot and told it to sleep, then threw it in my bag and when I got home, took it out. It had been on the whole time and was burning up due to the lack of ventilation in my backpack. That's just depressing.
  10. Remove Grub
    I'm not sure I agree with this at all. I realize that GRUB is not the easiest or prettiest thing to work with, but it is so necessary that to want to remove it is like saying "Remove X." I realize that there are a few alternatives like SYSLINUX, LILO, and uh....LOADLIN(?) but honestly, in my experience, GRUB is used far, far more. If anything, Paul should have said "Replace Grub with ____" or "Redo Grub." Removing GRUB is the easy part: finding a replacement is the challenge.
    (*Side Note: I'd also like the say that GRUB2 is so difficult to work with. I finally understand GRUB enough to use it for the most part without even having to look online, and then they create GRUB2, which is 10 times more confusing, and still as frustrating. In the past several years, Linux has made leaps and bounds when it comes to simplifying things for the user, and GRUB has not. To put it frankly, I'd describe GRUB as "Last decade's technology, today.")
  11. Make Every Task Doable from the GUI
    For the record I believe the steps to get people to move to Linux are as follows: (1) Make everything look presentable, i.e., not butt ugly (Lucid -> Meerkat), (2) Make the console a last resort.
    Good example: rotating wallpapers. (At least on GNOME, for you KDErs out there.) I searched around and everyone kept suggesting scripts and whanot. You know what Windows has to rotate wallpapers? Programs. Nice GUIs that will watch a folder with Categories and such instead of having to dive into Gedit and add it to scheduled tasks or whatever. And the Wallpaper changer is a good example because all it essentially needs to do is (1) let you pick files (or folders), (2) execute a command to change the wallpaper, and (3) add a timer. Basically a GUI for a console command. Not that hard.
    (*Side note: I did eventually find one, but [at least in the default Ubuntu repos] the small number of [functioning] ones is disturbing, considering that even a non-learned person like myself wrote a simple one for Windows.)
  12. Rolling Releases
    Honestly, this isn't something I can speak on. I don't really care too much about updates, but I do agree that the release system needs to be tweaked.
  13. Restore the Desktop in KDE
    I don't even know what a "desktop plasmoid" is; I don't use KDE enough.
  14. Improve the documentation
    While this sounds great on paper (er....LCD screen), I don't see it ever happening. Basically, the best way to do this, from my view, would be a website -almost a Wiki. In order to do this, you'd need people to contribute, and in order to have that, you'd need people to not want to do everything their own way. (Sorry, but it really is kinda true.)
    Also, "examples" sounds wonderful, but they can range from the simple to the extremely complex, and to have an example for everything in between is excessive. So you're probably going to end out with users that are either befuddled because the examples are too advanced for them, or not helped because they are too easy.
    Again: I love it, I just don't see it happening. The best alternative is forums, which work great some of the times, if people would actually reply to my GD threads for once.
  15. Replace GIMP
    Again, the same basic concepts for GRUB apply. GIMP is amazing. Period. The UI is not. We don't want to replace GIMP because, for the most part, the functionality works great, we just need a few people with a knack for UI design. Yes, it will be a huge project, mostly because whoever takes on the project will have the weight of the community on them, since everyone wants the GIMP to be great. (Maybe I'm wrong about that. But seriously, if for the next GIMP release, it performed the same and had a UI as nice as Photoshop, would not not be ecstatic?) But to just say "Replace GIMP" is kind of short-sighted.
  16. Replace
    Now I'm just getting sick of the "Replace _____". The answer to Linux's problems is not "Out with the old, in with the new" / "Shoot the horse in the head the minute it breaks its leg." I've said it before: I like Microsoft Word 2003 more than OOo, but that doesn't mean I don't love and use OOo. Yes, it is slow, and and uses Java, but still: it's community driven (partly, anyway), and is decently stable. It just. Needs. Tweaking. And I, for one, hope that Oracle makes that a top priority. But the fact that it is a cross-platform means that it allows more of a seamless integration with Windows users, and the fact that it has corporate backing means it has resource not as readily available to a project that is completely community driven. That's not something you just want to toss away.
  17. Mike says...[filesystem layout]
    I've always wondered this because it's so incredibly cryptic for a beginning user (i.e., myself), mostly because instead of full words like "Program Files" or "Application Data", Linux just uses "bin" or "var." OS X does have it going on when it comes to how Applications are built, at least from the viewpoint of simplicity. I think I would be naive to vouch for this point just yet, mostly because if it's so archaic, why has Gobo Linux not really caught on that much? But I guess the best way to get this to change would be to change the mind of the Linux Foundation, since no single distro will take the leap because it's a risk. But if every distro changes, I think it will end out being beneficial for everyone.
    (*Side Note: While searching around quickly, I discovered that it actually was suggested as an Ubuntu brainstorm. I guess I might make a post about that later on.)
  18. Graham says...[unified desktop?]
    Of all the points, this is the one that makes the least he wants to stress that "Gnome, KDE and the rest are free to continue," but their "resources" should be compiled into a "united front"....Seriously, what is he getting at? I think he's saying that there should be one desktop that should be made standard for everything, but the others should be allowed as alternatives. I think that's retarded.
    1. People defend DEs more than distros. Considering that distros are apps on top of the linux kernel, the main difference between distros is the apps; and since the DE largely drives what apps you are likely to get, it is the driving force for distros, in my opinion. To say "Oh just pick one and settle down!" is foolish, in my beginner's opinion.
    2. Honestly, one DE is just not enough for the Linux world. KDE looks nice for a desktop, but maybe Openbox runs better on an older machine. There is no one DE or WM that can run everywhere. In this particular case, I think diversity is ok, especially since -for the most part- it's narrowed down between GNOME and KDE. If we could get down to 1  desktop Linux distro with two different varieties: GNOME and KDE, I would be ecstatic.
  19. Less Screen Clutter
    I have not found this to be a problem. At all. Windows bugs me more than Linux ever has.
  20. Better Organized Settings
    From what I can tell, he's talking about the GNOME menu bar. I'd be inclined to agree, but I'd have to say that LXDE is disgustingly unfriendly and needs to be redone. I haven't used KDE's menu system much, and I honestly think that GNOME Settings/Administration issue is a trivial one, but for god's sake, LXDE, let me put menu items where I want.
  21. Kill Off Dotfiles
    I have no words.
  22. Easier Closed-source Installs
    I honestly thought that this was ok. I'm pretty sure Ubuntu has some of the things he even mentioned like asking the user to install certain closed-source resource, but maybe I'm mistaken.
  23. Standardize use of sudo
    True dat. But if it's one thing that Linux people are afraid of, it's the word "standardize."
  24. No more open core
    This, I am confused by. So he's saying that things should either be completely Open Source, or completely Closed Source....Or just completely Open Source? I think this is more of a philosophical issue than a "this makes Linux difficult/annoying." As much as a world of complete Open Source world sounds wonderful, I don't personally think it sounds effective as a business model, and I'd rather have part of something open source than none of it.

    No comments:

    Post a Comment