Hacker News Reader

Comment by zserge | Hacker News Reader

I believe that the major pitfall here is that the feedback you've received is mostly about the changes that people want to see. However if we consider the number of people who want Gnome vs the number of those who want Unity 8 vs the number of conservative users who like Unity 7 as it is now - the results might be different.

I personally am very happy with the current Unity. I find it intuitive and more aesthetically pleasant/polished than Gnome Shell (I've only used that as it comes with Ubuntu Gnome).

So please, don't drop current Unity. Or if you have to switch to Gnome Shell - please keep the user experience as close as possible to the current Unity to help users migrate.

View on HN

In other words: people who like the current status quo are less likely to make themselves heard than those who want something changed. So the trends you tried to extrapolate from the HN discussion may be biased.

As an anecdote, that was exactly my experience. I'm one of those light-weight desktop Linux users. I've only experienced Windwos, OSx and Ubuntu. I can say that I really enjoy the current Ubuntu UI, but due to lack of knowledge of Wayland and whatevery, I'm disinclined to speak up.

I'm pretty sure their decision to drop Unity was already done (internally) even before Dustin started the suggestions thread. So this conclusion isn't necessarily accurate.

To get more unbiased input one might try something like http://www.allourideas.org/ . This method of priorisation and feedback gathering is much better than "more votes = higher priority". The downside is lack of comments/discussion.

I also wonder how representative HN users really are, but maybe that's less of a concern with a product like desktop Linux which hardly has any nontechnical users in the first place.

I think the biggest thing that people want (or at least I do) is for Canonical to drop Mir, and focus on Wayland. Apparently Unity and Mir are so tightly coupled that it wasn't even worth mentioning Mir in Shuttleworth's post.

Noob question: Why? Do you think Wayland is better? Or is it more efficient to have everyone working on and things being written for one protocol?

The second one. As far as I've been told, Mir was conceived because Wayland had some deficiencies, but AFAIK they've been sorted out. So it's really a matter of saving work now.

Those particular deficiencies were essentially fictional in the first place.

I see. What were they?

Is Wayland better than Mir? I have no idea.

Is Wayland better than Xorg? Yes, or at least it will be.

Wayland and Mir are distinct projects made to solve the exact same problem in the exact same problem space.

Usually this is a fine idea, but Wayland could really use the support that Canonical has avaliable, especially since it depends so heavily on graphics driver support, which Canonical can help push for using its association with NVIDIA (the only graphics manufacturer with a completely proprietary driver now).

Notice that Xorg doesn't have any real forks anymore. That is because it a much better idea to focus driver support on one library (graphics drivers are hard enough as it is). Unfortunately, Xorg has some inherent problems that Wayland is designed to fix; so until Wayland is complete and stable enough to replace Xorg, we need all our graphics drivers to target both stacks (Wayland and Xorg). That's already difficult enough without Mir demanding its own attention.

One of the things that has always confused me is that Gnome Shell is really extendible, from a programmer point of view. Even if you don't like Gnome Shell, you can actually start with mutter and build a window manager really easily (or, at least this was the case about 5 years ago when I last looked into this). Building Unity on top of Gnome Sell should be very straight forward. I'm still surprised nobody has tried to do it.

Again, I haven't looked at it lately, but the last time I did, mutter was just fantastic. I was tempted to write my own WM on top of it, but since I have grown disenfranchised with the rest of the Gnome ecosystem (and it's hard to pull that out of mutter), I gave up. This shouldn't be a problem with Unity, though, as it is already entrenched in Gnome.

One potential problem is if Gnome make breaking changes, it's hard for Canonical (or anyone) to use it as a base for their own UI. I've not experienced this first-hand, since I haven't used Gnome or written GTK+ apps since the 2.x days, but judging by what I've seen online they like to break APIs at each release, drop working features if/when their opinions change, and be semi-hostile to any use-case other than their own default settings ("brand"), e.g.

https://igurublog.wordpress.com/2012/11/05/gnome-et-al-rotti...

https://davmac.wordpress.com/2016/07/05/why-do-we-keep-build...

I suppose Canonical are big enough to have their needs taken into account, unlike other app authors.

This is actually the reason I gave up on my idea of working with Gnome Shell. But as far as I can tell, most of the badness happens before you get to Shell. Basically, Canonical has had to deal with it in Unity anyway.

> Building Unity on top of Gnome Sell should be very straight forward. I'm still surprised nobody has tried to do it.

I think this is what ElementaryOS[1] did. It was funny/interesting to me that the "flashiest" features demoed on the website were pretty much just stock Gnome 3 features.

[1] https://elementary.io/

"more aesthetically pleasant/polished than Gnome Shell" I really disagree with that. Are you speaking about Gnome 3? It's a subjective thing, but Gnome 3 is incredibly aesthetically pleasing and polished imho

FWIW, one more liker for Unity here. Too late I guess...

Reply on news.ycombinator.com