I don’t want to sound like a standing-on-the-sidelines complainer, but I do want to express my disappointment.
In theory, what I think is even more important than pushing to launch the Communitheme as the default theme, is realizing that we can’t afford to wait any longer to discharge Ambiance. It looks so old and out of place in the GNOME environment. The icons are long overdue. (Ever since, and even before, the first Unity 8 version came out.)
The question that also arises is: when is a theme - or any work of art, really - 100% ready? KDE, Linux Mint, Budgie, etc. There are many project that seem to be churning out new themes and tweaks every once in a while. While on the part of Canonical, maybe still because of the backlash from launching Unity ‘too quickly’ as a replacement for GNOME 2, they keep on stalling until something reaches a level of utter perfection.
I’m just saying: this is a choice, it is not a law of nature, and it doesn’t have to be this way. A theme, even a default one, can be 75% ‘ready’ and still be launched. If I had a say in it, and the choice were between Ambiance and the unfinished Communitheme, I would go for the theme that already looks a lot better and more polished. Which is the Communitheme.
We can’t agree on everything, obviously (me for one, being an absolute fan of GNOME removing the active desktop), but I hope we can agree on some principles. One of those principles I hope most can agree on is: don’t stall and stall, and wait until it is no longer ‘new’, to launch something. Just like Unity 8, Ubuntu desktop could be at risk of lagging behind constantly.
After launch, I hope the community can continue chipping away the rough edges, and at the same time can start working on a replacement for Radiance!
Whining solves nothing, the time frame has been known. Ambiance serves its purpose well as a very well tested and functionally mature theme which also has carried the Ubuntu brand for years.
A theme is not just a coat of paint, there is a tremendous amount of subtlety involved in helping the user understand the state of their system as they’re using it. Iterating on ideas from two design languages (Ambiance and Suru) to form a new system takes a long time, first to find all the corner cases where the themeing needs bug fixing, and second to arrive at the correct and consistent feel so the theme itself becomes transparent in its function. As is the case with every release, a lot of people will change from the default theme, but that is not a reason to change from what’s worked in the past, and neither is impatience. I use the new theme on all my systems (which are all on 17.10 bar the servers), and can’t imagine using Ambiance again, but it’s not really done yet and the feature freeze is on. It’s life.
Since the beginning we know that to make it we needed a little miracle. We started the project in November, only 5 months before the UI freeze and 5 months of just free time of a small team. I think this is the right choice, Communitheme still has some important bugs, some undefined styling and tons of applications to be tested.
What I like the most is that this Community appreciated so much the work done so far, thanks to all!
Indeed, and you already made a miracle! I’m sincerely really impressed with the current theme state, knowin that real work started mid-November with the first mockups and ideas. Great work to you and the other members of the team, working tirelessly on CommuniTheme! Thanks as well to @andy-k to have summed up well the subtilities of landing a new theme.
However, as I was afraid and stated previously already, we can’t land safely the new CommuniTheme in 18.04. There are multiple reasons:
As @c-lobrano stated, and you will as well if you follow the upstream repositories on GitHub, (I encourage everyone participating to this thread to to do that on the various communitheme-related projects), there are still a lot to be done and important bug reports, a lot of applications to be tested. We just went through and understood completely the GTK2 issues that some people had, but there is Qt and other GTK3 applications that we need to have some testing on and bugs fixed.
This is to totally normal, and those bugs (with new ones discovered everyday) are part of creating a theme. Applications are using the toolkits and styles in very subtle ways, and you need to try to not make the whole theme unmaintainable on the long term with too many application-specific code.
UI Freeze is next week. After that point, it mean that NO graphical/visible changes are allowed apart getting an excepetion. It means in general, no more refinement, no visible element or color changes for the next 5 years on 18.04 and to Canonical to support it. Are you sure the theme is ready for this? Seeing the amount of discussions triggered here, it seems that there isn’t a clear answer, and maybe setting in stone the CommuniTheme choices at this point will definitely decrease the amount of experimentations and changes that would otherwise still happen here. It means basically Freezing CommuniTheme on the base principles choices.
This is to be expected, a theme is hard to make, regressions are easily triggered, issues are constantly coming until we have a good amount of testing. The goal is to push it very early on the 18.10 development cycle (like, within the first month once the archive opens) and iterate from there. We can have regular snapshots of the theme (uploaded manually) and get some feedback.
However, people following and running development release (even more non LTS) are few, and this is why I think we should make it easily available to everyone who wants to opt-in, even on 18.04 LTS, and I have 2 proposals here. The good thing is that people “who can’t stand” anymore Ambiance can continue following this as well
Here is what we can do:
Commit to maintain in parallel than the version in the distribution the ppa with daily builds. It would mean however than every commit (potentially breaking ones) will land on people stable distributions, potentially breaking them even if they are just users who wanted to give a try to the new theme. Also, installing a ppa isn’t for everyone.
Ensure we land basic “double session” support for shipping the theme via a snap that people would have to install. The instruction will basically be snap install <theme> or use GNOME Software to install it, then a new GDM session appears (similarly to today), and people can jump into this.
The benefits are:
The snap isn’t installed by default, so we keep the opt-in as for the PPA
We can use multiple maturity channels, meaning that people who want to follow every new changes upstream can subscribe to the “edge” channel, knowing that they suscribe to the “edge” risks with what it means Then, regularly after community +1 and some testing we can promote a particular revision to the “candidate” or “stable” channel, delivering it to a wider community.
We can even imagine having some snaps built per pull-request, so that people can have a try before things land!
This is a substantial amount of work to be done, but I’m happy handling this if people like that idea. I will need to know this week to land the “in distro” defaults for the new session, appearing only when the snap is installed, and to teach the system to look in that case for the snap theme folder (Note that I don’t know if that will work with GTK2 theming yet).
Finally, I think we should include the great sound work and cursor icons made by @madsrh officially as part of the CommuniTheme project. I would love as well someone stepping up for the plymouth theme work (which is mostly code-work, no css, svg…). Anyone interested?
Great idea. Especially if shipping the theme as a snap means that every snap that at the moment only includes ambiance could somehow use the new theme from that snap. I guess snapping a theme is also a good testing playground, since as far as I can understand this from my user perspective, at the moment only ambiance is packed into the installed snaps.
If that snap would be somehow promoted in the LTS (in the store maybe?) most disappointments could also vanish Oh snap… I wrote “snap” many times in this post.
I would like it to be implemented from Adwaita-ubuntu with the orange color, which would be closer to the gnome-vanilla and the deploy icons of the suru-places folder and the retouching of the Humanity icons.