What is Wrong with Linux Software Centers?

Hey all, I've been trying to write a blog post once a week. Anyone who has ever tried to schedule forced writing probably knows that you write best when a novel idea comes to you. Today I knew I wanted to write something, but as I sat and thought about a topic I settled on something I've been complaining about for a few weeks now to my friends and co-workers. Today I'm going to talk about "App Stores", "Software Centers", and "App Centers" that are currently shipping on popular Linux distributions.

Firstly, it makes me very upset that we are no longer the leaders in this type of software (App "Stores"). We were using software repositories to deliver our applications long before Windows or Macintosh, and I remember we did it better for quite a long time. But as time has gone on Apple, specifically, but also Android and even Windows have begun doing quite well with their app stores. My biggest complaint is in regards to software selection and the experience within the apps, and it's time to address it.

How would someone go about solving the aforementioned problems? In my mind, it boils down to three focuses: curation, partnerships, and focus. These tenets are extremely important for an app store as they are, in my mind, the pieces that make the experience of using these applications a positive one.

Curation is where we'll start. What I mean when I talk about curation is a person, or group of people, taking the time to highlight the best applications available for download in the store/repositories. I was extremely frustrated when I found that the "Editor's Picks" in Gnome Software (GS) shipping on Ubuntu (might be this way for all distributions shipping GS) was not Editor's Picks at all, but was instead determined by user ratings on applications. The problem there being that most users making use of the GS were not rating the apps they were using. The result? Weird, esoteric applications being highlighted as the best apps available. We need real curation in the software center(s).

Partnerships are not as hard to accomplish as people make them out to be. I have found, in my own experience, that companies are usually quite open to talking if folks take the time to reach out to them. For instance, I have been discussing distribution of proprietary software from a few companies (will name them when things are finalized) through a software repository - and so far the response has been very good. We need a group of people whose full time job is to go through and get explicit permission to redistribute software from companies and organizations producing the software we use though our repositories. It's frustrating to tell friends and family moving the Linux distributions "just use the app store" and they ask "where is Slack and Chrome?", and you have to reply: "Oh, you have to download that from their websites." I'm going to do my part to solve this problem, but I think that leaders like Canonical should work much harder on these partnerships to expand their partner repo - and make it an option as setup/installation to enable this repo.

As a side note in regards to partnerships, let's solve payment once and for all in each of the app centers. I use mostly open source software, but dammit, we need to allow folks to get paid for their software if its proprietary. I, fortunately, don't have to use proprietary software for my job - but there are many people who do and not having it be easy to install the software that they are forced to use doesn't advance open source and Linux. If the platform (Linux) can be shown to be viable, that opens the door to more open source software adoption on the desktop (IMO).

Focus. Wow, this one is really upsetting these days. We have industry leaders like Canonical and Red Hat who have displayed virtually zero focus on their desktop software centers. I don't care if you plan to replace what you are shipping or make it better. What you ship today is what your users experience, that is you. Best intentions don't count in this game. If you don't give a shit about making desktop Linux a viable replacement for other proprietary operating systems, then just get out of the game. Some have heard me vent at how much more polished the Patheon desktop that ships with Elementary OS feels vs Unity. That shouldn't happen. Who has more resources? Last I checked Canonical has an entire division focused on desktop, Elementary has like one or two full-time employees total. I can't stress how strange it is that they are comparable given the difference in available resources. So I ask the big boys to ship something good today, and focus on making it a great experience - today.

This is a negative post, so I apologize for anyone I've offended. But I'm very upset at the story we tell our users around software installation. Whether it's debs, Snaps, RPMs, or some other format - I don't care. These are all solid technologies, let's figure out our distribution methods and really nail them.

I'm sure I'll get some crap about how I should get off my high horse and write something myself. My answer, I've dug into Gnome Software and Elementary's App Center this past week. I am considering throwing my weight (support, evangelism, and code) behind something in the coming weeks. I want to hear feedback, and see if we can make something that I feel content telling friends and family to use. Something that gives them current versions of the software they need.

That's the end of my venting session. As always, feel free to Email me at [email protected] or message me on Twitter: @ryanleesipes.