Home > Linux > My first blog post: Why I hate Ubuntu now

My first blog post: Why I hate Ubuntu now

Hi my future readers!
I finally decided to start writing a blog (which I registered a year ago, but didn’t update).

This post is about Ubuntu, and how I feel about it, force feeding me its updates.
This is my opinion, I am not trolling, just saying what I think about the ‘we know what is good for you’ routine.

Update: I don’t use Ubuntu any more, I switched back to Kubuntu.

Just like most Linux users I dream about high Linux usage worldwide.
While I do understand very well that the open-source software can’t feed its developers usually, I believe that the core of the computer, its operation system can be free and financially sustained.
All you have to do is to include the OS cost in price of the hardware, and in addition to that make money from closed source user applications.
After all, OEM installs are the biggest source of revenue that Microsoft gets from its Windows division. Users that install this OS on their own, usually just choose to pirate it.

In my opinion, a closed source OS really puts strong handicaps on the hardware and software vendors.

Think about flash cards for example. Why the wear levelling is done inside the card using a puny micro-controller?
You could tap into the main CPU’s almost unlimited power and the vast amounts of system memory to carry out rather clever and generic wear levelling algorithm.
This is not done, because Windows mandates all block devices to expose SCSI interface and has just 2 file systems, FAT and NTFS.
The wear leveling however, is best done in the filesystem or at least in a generic layer between the flash controller and the filesystem.

Yes, you can put the wear levelling into a device driver, but this is dangerous, since each card reader vendor will have to implement it correctly, and you can’t count on all of them to do that right.
Wrong wear leveling doesn’t manifest itself immediately, but rather after a prolonged use, thus users will likely to blame the flash card, rather than, the driver.
Olympus did try to do that, but failed.
Rather that this mess, Windows should have had a generic flash wear levelling mechanism, like UBI for example.

And now lets explore the main thought of this blog post.

Like I said, I dream about high Linux usage worldwide (aka World Domination), thus I choose the distribution I that shows the signs of trying to meet that goal.
So I use Ubuntu.

Once upon a time I used KDE desktop and I was rather happy about it.
The only thing that always bothered me was quite large amount of bugs I have seen, and somewhat lack of progress.

The thing that drove me away from it was KDE4.
I also noticed that KDE developers seemed to inherit the feature removal attitude from Gnome. Also I really hated the Nautilus clone, the Dolphin.
So the major point against Gnome somewhat became a bit blurry.

Also I didn’t like the KDE’s developers decision to develop a compiz competitor, kwin4, since I see compiz as one of the programs that can bring more users to Linux.
I also noticed that KDE4 doesn’t work well with Compiz.

Its sad, that many “hardcore” Linux users consider Compiz, “eye candy” and dismiss it as a toy.
Speaking of which, they a wrong ether way. Regardless of compiz being a ‘toy’, it does bring new users to Linux, and thus should be enabled by default.
Its well-known that looks of a device plays a major role when a customer selects which device to buy.
For example this is a major reason for Apple success.

So, after few tries I switched to Gnome, and soon found few features that made me stay there until now.
Lets look at features of Gnome that I found superior to KDE:

To survive, each OS has to evolve. While Linux has its UNIX roots, it constantly expanding to include support for modern technologies.
To be honest a whole OS isn’t just a kernel and a bunch of libraries. Its much more than that.
Gnome and KDE provide a lot of interfaces that should be considered a part of the OS.
It’s somewhat sad that effort is split. While I don’t mind and actually happy that both KDE and Gnome exits, I think that many core features should be united.

I found out that in many areas Gnome leads in creation of sane new interfaces that are aimed to become a standard part of the Linux.
KDE eventually adopts them as well usually.

Examples:

1. Gsteamer – somewhat buggy, glib based media library. Yet it is the only general purpose media library that supports both decoding and encoding.
Other media playback engines are:

* libxine which only supports playback and isn’t flexible in regard to processing the media stream and it is also buggy (I even fixed few bugs there…).

* VLC that seems to contain libvlc, but I never heard about third-party applications using it. VLC seems to be the least buggy media playback application though.

* mplayer/mencoder is even worse since it is a standalone application, but does support both encoding and playback.

These days almost every application needs sound and video playback of any format, thus gstreamer is the best solution yet.
Also applications need support for encoding for many many reasons.
Not to mention the fact that hardware based playback acceleration becomes the norm, thus a media library more and more becomes a part of the OS.

2. Hotplug & device configuration:
Projects like NetworkManager, Hal (and its replacement udisks and upower + udev) give sane control over the hardware to its users and most importantly, make applications aware of hardware present on the system and its status in generic way.

For example an application might choose to refresh its network connection if it detects (via Networkmanager) that user switched to different wifi network.
Or application might need to present a list of devices currently mounted (like usb sticks) and be aware or addition of new devices.
(udisks does that in generic ways, and hal did that too).
There are many more examples out there. For example an disk recordring (burning) (which needs first to know if disk burner is present in the system…) might ask the OS to inhibit automatic suspend.
One feature that for some reason wasn’t adoped was an ability to access package manager in a generic way thus giving applications a way to install packages, a feature that OS really should provide (Android does btw….)

3. PulseAudio
Yes, that dreadful pulseaudio.
Yes, it broke sound on many systems, including mine once but yet it is a necessity, because these days we need to support sound device hotplug, since more and more headsets that are sold, use USB, and Bluetooth. It is the only way to achieve hotplug support
We absolutely need some sort of common layer that talks to hardware via alsa, and PulseAudio finally seems to become it.
Besides PA has potential to include software sound effects that are norm on Windows.
Kernel developers will never allow that kind of code in it regardless of anything, because it needs floating point support, which kernel lacks.
(And these sound effects are very nice, I tried them many times on windows – sound is just better).
BTW, its latest versions work well on all my systems, including usage of Skype.

4. Compiz
Remember, I switched to Gnome before introduction of dreadful Gnome shell also known as mutter.
Although compiz works fine in KDE, it was developed as Gnome project and fits better there.

In addition to these features, I found replacement for every feature I used in KDE.
Now that I went from 8.4 to 11.4, I finally reached the point when I feel that I must say “Enough!”, and switch to something else.

Lets see the bad things in chronological order:

1. I lost ability to control the alsa mixer directly, so now I can only control it via PulseAudio mixer. (But what if I need to set a not-standard setting?, I have to use alsamixer)

* They made the screen dark. OK, that’s just a theme, and I like it now, but yet, that was forced on users. This is minor gripe though.

* I lost ability to not lock the screen after suspend.
I really don’t need this lock, and no, the gconf key doesn’t work. I had to edit the source to comment that out of indicator-session.
(I never have submitted that upstream, because to have any chance to be accepted I need to make it configurable, and I just commented out the offended code. I also was sure they won’t accept it, so I didn’t bother.)

* I lost ability to configure gdm. about 90% of its settings vanished.
In addition to that fast user switching when KMS is enabled isn’t usually fast anymore, because it throws you to the gdm switcher instead of another user’s screen.

* I got that useless me-menu and they replaced Pidgin which is fine IRC app, with buggy Telepathy, and now I have to remove it, so it doesn’t bother me from indicator session menu.

* I got application close/minimize/maximize buttons on the left.
Sure its trivial to move them back, but I think that was a nasty move.
That copies mac OS which is just bad. I am sure you will say that this is trolling, but thisisn’t.
Sure current Ubuntu copies a lot from Windows look and style, but that has very good and practical excuse/reason, that is the fact that UI that is similar to Windows makes it easier to switch to Linux.
That argument doesn’t work for Mac, so there is no justification for copying it!
Go ahead and tell that is trolling, but that is just a truth you don’t want to hear.

* Now we got ***drum roll please*** Unity.
I admit that it is still better that gnome-shell, as it at least allows me to keep compiz which I like very much, which otherwise could face an extinction cause nether Gnome nor KDE uses it by default.

It’s already pity that Ubuntu, by default enables just small subset of compiz features, especially it doesn’t enable the cube switcher, probably due to an ego of few developers that think it is a useless eye candy.
Even if it were eye candy, the fact that it can bring new users, justifies it being the default. Remember why iPhone and iPad are successful?

But now, you can’t enable the cube and unity at same time. Yes you heard it right.
And that is mostly a political decision, because “it doesn’t fit the user experience”
Lead compiz developer, Sam Spilsbury, confirmed that he can’t do much about it, since this is Canonical decision.
Update:This decision was reverted. Cube can be enabled, though its not a supported configuration.

Not only that of course.
Unity doesn’t have any applets, so say goodbye to your cpu frequency/temperature/load monitors, drawers, timers, etc…
Its date applet doesn’t include weather report, and alternative timezone support is very broken
(But these features I hope will be implemented eventually).

Besides no matter how much I attempt to force myself to like it, I still came to same conclusion that Unity design is very broken. Cramming everything into one big sidebar is really wrong.

But that still isn’t the worst thing.

**** drum roll please ****.

And now please applause to new super original feature of Ubuntu

Meet the global menu!

One of the worst features of the mac, is now directly copied into Ubuntu.
See above why copying Mac is outright nasty, and especially its weak points.
It like cheating on an exam, when you copy the other student’s errors.
Apple keeps the global menu only because they don’t want to hurt its userbase, because unlike Gnome/Canonical, they actually do seem to care about their users.
Also its important to note that usability difference between global and normal menu isn’t that large,
normal menu is better but if you are used to global menu, you might as well not like the normal menu.

But that isn’t all!.
Unlike Mac which had global menu from dawn of its time, Linux never had it widely adopted.
Yes I am aware of global menu applets that did and do exist in both Gnome and KDE, but these aren’t widely used.

So, only GTK+ applications will be able to use that feature.
In KDE version of the global menu probably KDE/QT applications were able to use it, but not in Unity
In addition to that, all applications that use custom widgets, or even use the GTK+ in a non standard way won’t support global menu unless you port them one after another.
Ubuntu did port Firefox so far, but what about others? LibereOffice for example?
This way global menu will just confuse users

Sure, you can disable it, sure, but that still stinks, and besides to disable it you need to remove a specific package, and I don’t even remember which.

I think I know why all this happens.
Since Linux isn’t that usable, that’s just sad truth, I think most Linux developers got themself a Mac, and now they try to make Linux work like it, because they can’t change the Mac and thus to make it easy to dual boot they want Linux to behave like Mac. I know that from my experience. It’s mentally hard to use two different GUI interfaces.
For example I find it mentally hard to use windows now. Ubuntu is already too different from it.
I want (and can’t get) the dual toolbar, I want themes/icons of Ubuntu, etc…

So, they first add a top toolbar (as in mac), then change order of buttons, then add global menu……
Unity itself also isn’t that far from Mac launcher panel.

What next?

Thanks for reading.
Stay tuned for more posts, I have a lot to say.

Advertisements
Categories: Linux Tags: ,
  1. August 18, 2014 at 12:15 pm

    Please let me know if you’re looking for a writer for your weblog.
    You have some really good posts and I believe I would be a good asset.
    If you ever want to take some of the load off, I’d love to write some articles for
    your blog in exchange for a link back to mine.
    Please blast me an e-mail if interested. Thanks!

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: