openSUSE:ALP/Workgroups/Community/Workshops/LeapDesktop
Archive of the Community Desktop Workshop
What was the workshop about?
Identify current strengths and weaknesses of the existing Leap Desktop. Think about the opportunities and Threats that we have ahead of us. What would make a successful desktop?
We've had two timeslots for the workshop one more Eastern friendly and the other more Western friendly. There was about 40-50 participants in total (peak hour in both workshops).
Original annoucements
Meeting minutes including followup discussion
Tuesday's Workshop
https://etherpad.opensuse.org/p/weeklymeeting20220510
https://lucid.app/lucidspark/9491ef81-067a-4b7f-82c6-2a8fc2839705/edit?invitationId=inv_4f8a7922-7323-482d-a487-10f2ede4c3ad
Part 1 - What's wrong with current desktop
(hightest voted topics)
1. hardware enablement on Leap: people who are facing issues with hardware enablement should have an easy way to take kernel from the stable HEAD:
2. software is obsolete, we could provide a better service at the package lookup; relying on 3rd party OBS contributors is not that helpful as you might not trust these persons
3. Audio/video workflows are clumsy card
Simon piperwire will help
Some of these issues such as not good tools will not be resolved without alot of investment by the broader open source community. Others we can probably make a bit better.
Yes, but we can make it better for non FOSS tools as well, so commercial tools work good as in Windows/macOS, users in these platforms often pay for tools, so no reason to have all free in Linux as well (avicenzi) I somewhat agree but this is much less in our control, most of the really good audio tools I use on Windows simply don't work in Linux including under wine (Simon), True, that's why I mentioned DaVinci Resolve and Reaper, I tested, used for a while, works really well on Linux, both can be used "for free", but paid version offers more features
4. forums-o-o vs askubuntu: forums-o-o does not seem to be as easy to use as forums-o-o.
5. Improve SUSE's influence in upstreams (KDE)?
Part 2 - SWOT Analysis
Strengths
(hightest voted topics)
1. Well Tested, good QA
2. OBS is great, anyone can contribute easily, Very easy to get people onboard
3. Security bugfixes come through regularly and often
4. Power-user focus
OBS Marketing - we could improve that. More marketing materials needed
^ it would be good to have a discussion about whether openSUSE is about end-users or devs. In practice, OBS doesn't matter to end users, and devs will have other distros to care for anyway.
Make OBS package maintainence videos/tutorias/blog posts, teach the community to maintain, the wiki is a bit clumsy somethimes, there are hidden things
The official documentation is actually really great: https://openbuildservice.org/help/manuals/obs-user-guide/art.obs.bg.html#sec.obsbg.uc.newprj
https://duncan.codes/tutorials/rpm-packaging/index.html is also a really good beginners guide that's not linked to enough
OBS is a great tool for other distros, marketing that would be nice, Tizen is using an older OBS release
Weakness
(hightest voted topics)
1. Not well known in the FOSS community / bad visibility
People (users and maybe even SUSE employees) have to more vocal on certain platforms (or present at all).
lkocman: perhaps we could create a poll to rate our visibility on individual social networks
brazil region feels underrepresented (we have a heroes groups and a telegram group, some SUSE employees help there, including myself (avicenzi))
Ubuntu community in Brazil is way bigger, but mainly due to community and not Canonical efforts I guess
LATAM in general is a problem, not much investment there
simon: Australia feels similar
2. Not much investment in the final UX
simon: custom branding, more green, create gtk theme, not just limit ourselves to wallpaper, but full toolkit branding to match, particularly having KDE and gtk toolkit branding that matches each other so apps don't look out of place
aleix: being present in the development of the UX is the only way to ensure that you're getting the product you are after. It's both a technical but also a social topic.
Branding is something different than UX. If we talk about branding, be aware, that is hard to maintain. We had two years bugged openSUSE KDE dark theme due to how hard is to watch and keep with upstream changes. It is better to keep it minimal.
3. Lack of interaction with communities
4. Essentialy the same product as any other distro
Opportunities
(hightest voted topics)
1. Linux Usage in public sector / EU seems to be on the rise. Are we on the train?
Brazil is a huge market, and somewhat FLOSS friendly, many examples over the years from the government
Opportunity also in the education sector, universities, schools are more FLOSS friendly due to licensing problems
2. optimize the the software delivery through flatpaks, appimage, etc.
The app delivery through flatpaks is gaining a lot of momentum and this next gen distro will be an optimal time to have the delivery of apps unified in a way that systems are difficult to break. Whether updating through the desktop updates via Discover (KDE)
3. Better support for Arm and RISC-V Desktop and "mini servers"
navigation is not so easy, finding the images for individual devices, the only easy image to find is UEFI image is not easy. https://github.com/openSUSE/get-o-o/ (maybe this one https://github.com/openSUSE/get-o-o/issues/47)
openqa has not many arm boards, but we do have the images, so regression test is often an issue
4. Make sure Leap is very friendly for new users (its already pretty good) and market it as being for new users much much better
Please, don't make openSUSE less power-user focused. If should be friendly, then please do it KDE way (good default, powerfull possibilities)
Threats
(hightest voted topics)
1. ubuntu being the standard
Lack of feeling of ownership (patriotism) Maybe be more active on communities would change that view
LOL, patriotism? openSUSE should be investing in improving the product rather than being overly proud of zypper or yast - I think it's more about making users identify themselves with the distro and the community
2. unpredictable development of ALP
The uncertainty of ALP development direction harms the idea of stable desktop.
3. Fedora being the missing piece in between TW and Leap. Fedora is what openSUSE used to be (before Leap/TW); maybe go back and make Leap the base and SLE the (stable, more tested) spin-off based on Leap? That will be like Ubuntu - "twice a year" and "once per two years". And we will loose "enterprise" benefits of current desktop
4. Packages in OBS are often maintained in a 'best effort' way and end up outdated, or lack a maintainer.
Part 3 - What do we need to do to succeed?
1. Public calendar for events
Don't forget openSUSE heroes (mirrors mtg etc)
2. Central place for documentation
Is it known that Leap inherits all the SLE documentation.
3. All in community
Be more active in non english speaking countries/continents
Be move active in conferences and small events in such countries
It's not just about translation, but engagement with community
4. Keeping the identity
We have good snap support thanks to Apparmor. Have schedule which defines more stable distribution.
Offer multiple Window Managers. Keep power-users focus (powerfull installer, keep traditional package installer, ...)
5. Better UI for websitesâ¨Easier to find stuffâ¨All in one placeâ¨â¨We have too many subdomains and too many tools that do not look alike, just have the same color guides
Single look and feel on websites under opensuse domain
6. Hardware partnerships that are widely available.â¨â¨Lenovo, HP, something that we can tell people "just buy this and you're good to go" rather than "you buy a device, you wipe the OS and then you install our image"
Framework laptop - missing compatibility (fingerprint reader)
What about Cloud companies? I talked to binario.cloud and they included openSUSE in their open stack image list (opportunity)
Better audio/video workflowsâ¨â¨(pipewire maybe)
piperwire as the default.
lkocman: I did use it for two weeks in the row and experienced some issues with bluetooth devices
Yi Fan: I had a better experience with pipewire on BT devices than pulseaudio (with Tumbleweed)
Thursday's Workshop
https://etherpad.opensuse.org/p/weeklymeeting20220512
Part 1 - What's wrong with current desktop
The top ranked cards during voting
1. Newbie onboard (community outreach/building)
co-ordination in between steering commitee and individual communication channels
- there are a lot of project channels (reddit, wiki, mailing lists, discord, etc)
- how are these divided (if at all) by purpose or group?
- relationships between SUSE, devs, heroes, and casual users
2. Codec support clarification for dvd/physical media (copyrights?)
3. KDE not all apps/features are fully enabled (kwallet)
- https://bugzilla.opensuse.org/show_bug.cgi?id=1193348
- Automate testing of defaults, could help
guide to install codecs is outdated
https://github.com/cisco/openh264/issues/3480 - blocked on cisco engineering
3. Some KDE apps / technologies are not fully enabled by default (or at all) sometimes due to security, or other reasons; increased
partnership between SUSE security and KDE desiredHummingbug
4. Software search is missing useful metadata (supported status, devel or stable version, intended for end users or not)â¨Should be equally accessible via web, desktop/YaST, etc
5. Some applications are difficult/impossible to update due to conflicting dependencies (conflicting Qt versions/applications using private APIs)
Part 2 - SWOT Analysis
Strengths
The top ranked cards during voting
1. Robust recovery (snapshots)
2. Enterprise backed distro, quick with patches for SUSE packages
3. openQA
4. Community friendliness & Neutral Support for multiple desktop environments (equal treatment of KDE, Gnome, etc, all included in installer, etc)
Weakness
The top ranked cards during voting
1. Marketing, Recognition, Brand-Image
Community do-acracy
Please join:
openSUSE:ALP/Workgroups/Community
2. PK-zypp integration is wonky/annoying at times
- Whatever the solution is added here, it must be testable and reliable too.
3. Documentation is spread out across too many locations and is confusing
4. Kernel versions can be old (trouble with newer hardware)
5. Do-acracy; Users with no developer skills are second class citizens in suggestions to improve the distro's for everyone
Opportunities
The top ranked cards during voting
1. Add flathub / flatpaks by default
Do we plan to have our own repository?
2. Making codecs easier
3. Immutable OS
4. Improved remote desktop story
5. Application stack could be tested better
- How do we make getting into writting tests easier for the comunity? (poo#109783)
- How do we empower the comunity to contribute with automating tests? (poo#109891)
Threats
The top ranked cards during voting
1. Discouragement of using Leap within the comunity
2. Community (do-acracy) no clear group plan
3. Switching to containerized/upstream packages only may break full stack integration (theming, library swapping, hardware integration) in ALP-like distros
Flatpaks, snaps, other packaged apps sometimes grab user settings (themes, fonts, library config) from environment correctly, sometimes not. Getting this fixed is essential for any distro which plans to recommend them as the first choice for apps. Have to be clear on what is the value added by our distro over others -- more than just the same apps available everywhere else.
4. outdated software driving users away
5. Loss of workflow diversity (by favoring one DE style over another)
Part 3 - What do we need to do to succeed?
The top ranked cards during voting
1. Many applications in the repo (or easily searchable and installable)
2. Get it on employee laptops including non-technical. They'll tell you quickly how good the product is :-)
3. Community distro should have professional like / cool branding. Sometimes I feel we're bit left behind.
4. Codecs
5. Balance of stability and new technology
6.Neutral Support for multiple desktop environments (equal treatment of KDE, Gnome, etc, all included in installer, etc)
7.Successful upgrade path (if desktop changes from 15.5 to containerised applications)
8.Colaborate with business to understand their needs - general programmes for collaboration, sharing, 'Outlook' and specific apps like CAD, etc
- Run user interviews, and generate a list of solutions to workloads for different users/business (e.g poo#109912), creating stories that can be looked at, during the development of the distro.
- Make sure we are looking at solutions, and not looking at applications.