Old (2020) but gold : Alan Pope explains the added value of snaps

Media Coverage
1

Snaps are an ongoing debate.

I am pretty neutral about it because I see both pros and cons of using them.
Personally I think it is well worth to use them for internet facing applications although I would avoid them in latency critical message based setups like for instance multi process studio grade audio setups.

Though I don't use any snaps at the moment, I do see the benefits using them in certain circumstances. For instance how it encapsulate the apps for better security (jailing) and the benefit of being less dependent on external libraries.

But Alan describes in a video something valuable I've never thought of and I thought it might be interesting to share it.

I should be ashamed of myself because it took me so long before I discovered this video, it is almost 5 years old but very interesting nevertheless and still very relevant today.

The video covers more topics than snaps but the part I want to share is this:

7 Likes
2

Yeah, interesting to listen what he had to say and have his post from a bit later after the interview in context:
https://www.reddit.com/r/linux/comments/orz1lr/alan_pope_left_canonical_because_of_snaps/

And now we have:

I personally don't mind snap and do use it for some core Ubuntu apps but in the end I wonder - how would it work out if he didn't leave Canonical

3 Likes
3

Wow....that is an amazingly good discovery. :+1:

Who would have thought that he would make a 180° turn about it. :smile:

Thanks for the heads up Alarik, and also for the link to his github page to make migration from snap to flatpak easier :smiley:

3 Likes
4

You're welcome ! Hope it was an interesting read haha :blush:

Sad to see such sharp turns after you've seen such positive thoughts about the project from the same person because one can deduce it was mostly because good ideas kept being ignored - I've seen it many times in my field of research as well, a shame really

5 Likes
5

Oh yes it was absolutely interesting :smiling_face:

Here is what Alan said about leaving Canonical:

5 Likes
6

Thank you very much for the video - I really appreciate it - it just confirmed my thoughts about the whole topic :smiley:

2 Likes
7

The fact that a developer highlights community behaviour as a reason for leaving their job is bananas. Shocking.

@Alarik: Sad to see such sharp turns after you've seen such positive thoughts about the project from the same person because one can deduce it was mostly because good ideas kept being ignored - I've seen it many times in my field of research as well, a shame really

You might have missed this comment in the repository itself:

Note: The existence of unsnap is merely a tool to enable users to switch from snapped applications to flatpak applications. This is not intended as a commentary or slight against any software. It's just a utility.

My guess is his position is more along the lines of: he wanted more resources to focus on desktop users but the CTO was not on board with that. Clashes with leadership and their priorities are different from "good ideas kept being ignored". I know this frustration all too well, but I find it's better to try and understand what their priorities are, rather than try to impose my priorities.

Regarding the reddit commentary, in particular about Canonical losing their desktop focus: I still find Ubuntu MATE to be a solid desktop, but is Ubuntu really keeping up with other distributions in terms of desktop delivery? Maybe not; hence why I am (finally) spending a bit of time this year looking around.

2 Likes
8

So ... if there was a difference of opinion regarding the willingness to dedicate resources, or the scope that such resources should encompass, could it be because of the possible expressed desire by M$ that such focus should shift to Server, away from Desktop, to fill the strategic trench of M$'s needs/weekness in that area, assuming M$'s own resources (however mis-guided or mis-informed they be) would shift their focus and skills (or un-skills) to Linux-oriented Multi-Platform Desktop/GUI/Apps end of things.

  • Is the M$ takeover of Canonical in the cards?

  • Is Canonical re-defining its "strategic" focus towards the eventual abandoning of the stable of Multiple Desktops in favour of a single "streamlined" Desktop that is, in their view, a "viable" compromise (shades of Unity)?

  • If such is true, should we all be concerned that the extent of Canonical support for coordinated development resources and activities could shrink or even disappear altogether?

3 Likes
9

You can look at Canonical's financial statements if you like: CANONICAL GROUP LIMITED filing history - Find and update company information - GOV.UK. The word 'desktop' doesn't appear anywhere in their strategy statement, so it tells me that's not where they make money and hence not where their focus is. Fair enough.

Kudos to Canonical because a) we wouldn't have the Linux desktops that we have today without them and b) they have cemented themselves as leaders in the cloud/infrastructure realm (shocking that they haven't made more money out of it, I guess).

But I'm getting ready for someone else to take the lead on desktop.

4 Likes
10

Well in regards to MATE as a desktop combined with Ubuntu I don't see much will change for the worse since I think it's just not the same people and MATE itself is a desktop oriented project so if anything happens with their pairing with Ubuntu I guess one can just find a different OS that supports this DE.

A larger problem altogether that I see is lack of people in development that seems very real and more dangerous since if there isn't enough people working on it there will be less and less people using it since the project will have more and more delays and delays do deter some users since we live in an age of "I want it all and I want it now" mindset - problem with good projects is that they need time and community to build them and if there is imbalance and constant pressure to rush things then things will start to fall apart

I mean - all those questions about "Is MATE gonna be supported in the next Ubuntu release" really only put people in a state of worry and I think that is the main point to take from the reasoning about "leaving the job because of community" - fearmongering/abrasive discourse/toxicity towards people who use it - they are all valid reasons to not feel well being on that project and try and move away from it - I mean life is too short to be bothered with toxicity and it's not like he sold his soul to corporate world - and in that last video @tkn sent you can really feel that sentiment that he most of all works to help people and cherishes helpfulness above all

4 Likes
11

@Stephen_Wade , @Alarik , since the Corporate Filing states

Ubuntu is the preferred platform for developers and is widely used for Linux deployment across both public clouds, private data centres and embedded infrastructures.

  • doesn't that imply nurturing ease-of-use,

  • availability of tools, and

  • coordinated evolution of ecosystem for stable enabling of both end-user (developer/user) and business continuity at the "Desktop" ?

Just wondering about anyone else's interpretation! :slight_smile:

3 Likes
12

Hmm I focus on the word "developers" - I mean when they point out developers as those whom prefer Ubuntu as a platform it does not seem to me that they imply desktop ease-of-use as a main focus

I see it more geared towards people like my friend that is working in robotics who uses Ubuntu at university since the Ubuntu is the only platform their robots are compatible with - and they do not really use GUI all that much since most of the things they deal with is based in terminal for programs and so on

Availability of tools here is crucial but is not tied to GUI ease-of-use for a casual non-tech user like myself or at least I don't read it with that being implied but that just may be my professional deformation since I read texts with literary formalism lens :blush:

3 Likes
13

My interpretation is closer to @Alarik's. The sheer volume of their reporting that is specifically about servers, cloud, IoT, or enterprise ... that tells me what they are paying attention to most.

On the one hand I could be proven totally wrong. Thats great - it means I'll continue to have a solid experience with Ubuntu. On the other hand, what can I do to make sure I still have a solid Linux desktop?

We can stargaze all day but really what matters is what we are going to do. I'm not really the person to figure out how Canonical can improve their ROI for desktop. So if you are the type to plan ahead; that might include exploring other options when you can. And of course; make sure we keep building community so wherever any of us land we have the skills and knowledge to make it a great experience.

3 Likes
14

Personally, about Ubuntu and desktop, I don't see anything particularly worthy of their time and attention, outside from what they already do at this point. Nor do I seem to have a good grasp of whatever it is that the community seems to ask more from them.

Ubuntu as a base has personally been quite reliable to me for daily desktop use, even the default Ubuntu Gnome desktop. I usually interpret it in a way that Ubuntu and Canonical matured and regained their focus after multiple mishaps back in the time of Mir and Unity. Ubuntu 18.04 is quite the welcome release for me, for the similar reasons. If Gnome's not anyone's poison, we have flavours for that (just like Ubuntu Mate). Ubuntu works closely with these projects to ensure delivering a well-rounded experience based on the same criteria that drives the default.

On Canonical's end, I think it would be pretty difficult to do more than what they do right now. What people generally seem not to understand is that Canonical and Ubuntu is always beyond the desktop. You have IoT, Core, and Server. From what I heard, there's even the Robot Operating System. That's what their focus is on; the industry. Industry gives them economic input, and that's great for a for-profit company. That's what they're supposed to do anyways. There's only so much that they could allot for something that they want to work on, and I think those people are doing a good job.

From my perspective, things about the desktop should be the worries of the project team behind that desktop environment primarily. Distributions are responsible for packing them, and making sure they work well on their end. So, in that same point of view, it becomes particularly difficult for me to grasp what users ask for from Ubuntu that they couldn't from the DE teams.

If it's really about that, there's hardly ever any constructive criticism that you'd hear from these very people. As from earlier, Ubuntu works closely with the teams behind the flavours. Additionally, they work on some useful extensions (and Yaru) on top of Gnome to provide that very experience. Yet, it seems to be always 'parroting' about Snaps. People, particularly relatively inexperienced ones, echo the same nonconstructive criticism about how Snaps are bad, often for reasons that are just really mind-numbing to hear. That is despite the fact that they hardly know anything about Snaps really, about what its goals are, and what is it used for.

5 Likes
15

To follow — not to nitpick on other distributions — most of them deliver the upstream desktop anyways.

  1. Debian is pretty stable, but in exchange, you get older versions of packages. The same goes for every RHEL clone and SLED.
  2. Fedora delivers the same thing too, in a faster cycle, which of course affects stability at some point. I believe the same goes for Arch and Gentoo as well.
  3. OpenSUSE, on the other hand, has a faster cycle and adds some few modifications on top of the desktop, but delivering nothing of major deviation.

Those things are for the better, or for the worse. If people are asking whether Ubuntu is keeping up with other distributions in desktop delivery, I might as well ask — what is special from those other distributions that seem to make Ubuntu lag behind? Are the changes that Ubuntu introduces really that detrimental to the core software it's built upon?

7 Likes
16

Interesting thoughts and I would agree on pretty much most of them - only thing I would maybe say otherwise is that I still do think some emphasis should be on GUI users since most of the people coming from other OS-es are coming to Ubuntu or Ubuntu-based (Mint) Linux and I mean sure if your focus is Tech and industry go for it

But in the same time - you make snaps that have potential as we could hear in a 2020 interview, and then you don't encourage that ease-of-use development and team members leave (if we deduce what happened in later comments/videos) - that also sends a message to the community at large and even though your focus may be big business I still think most of the users use it as a private/home daily OS

Sure DE developers are here mostly to develop DE but they also need a base that is easy to mold and I think snaps did help with some apps with that (discord which I rarely use for example don't know why but it always crashed after a while and after one point I could literally count down when it will crash on Mint - snaps package never failed me like that)

If they don't focus mostly on us GUI users as it seems by the wording of Corporate Filing one just needs to imagine how much more we could have if it was one of the main goals - that is mainly my thinking on the whole thing since I really really love Ubuntu + MATE and other options just seem as second best because this community cannot be replaced, for me Linux is not just terminal and skill with commands it is also the people that teach you along the way

And about some of the snap hate - I don't really understand what people get from criticizing snaps and repeating other peoples thoughts like parrots when they know nothing or too little of it to make sound judgements - I mean when people start arguing against snaps and then drop the "I use arch btw" it is just toxicity for toxicity's sake in my book

5 Likes
17

Since you mentioned that...

I still do think some emphasis should be on GUI users since most of the people coming from other OS-es are coming to Ubuntu or Ubuntu-based (Mint) Linux

I think it's pretty difficult for them to deduce what exactly does the community want or ask for. What do you think should they do? Moreover, what can we do as a community to help in that regard?

3 Likes
18

I offer the following approach as a means to "distill" just what it is that GUI-based end Users want:

  • compile a list of the "leading" DE/Distro bundles,

  • create the list of pairings to go "head-to-head" (i.e. MATE on Ubuntu/UbuntuMATE vs Cinnamon on Mint or MATE on Ubuntu/UbuntuMATE vs MATE on RHEL).

then, offer a questionnaire for each single pairing of those configurations that, for each "feature" grades (radio-button scale)

  • which of the pair is perceived more "desirable",

  • which of the pair is perceived more "robust" technically,

  • which of the pair is perceived more "intuitive",

  • which of the pair is perceived more "performant", and

  • (limited only to a pool of confirmed developers) which of the pair's source code is perceived more "clean" (I do not mean more terse, or more compact, but more maintainable because of modularization, logical compartmentalization, and choice of strings assigned for naming variables, functions, APIs).

I fully realize that the scope of such an endeavour would be rather huge, and would be very much dependent upon the willingness of those end Users to take the time to be objective and forthcoming with their (as objective as posssible) opinions for each of the many individual Questionnaires covering those pairings. But I don't know how else anyone could reasonably "get to the bottom" of the key attractants offered by the various Desktop Environments (along with features of associated integrated toolsets)! :slight_smile:

3 Likes
19

My couple of coppers here...
The driving force behind GNU/OSS is the idea of free information and technology sharing among IT professionals, i.e. scientists and engineers, is not it?
FOSS is intended to have somewhat broader audience. Great. Why not? Well, the competitive high quality software is produced due to industrial production process which needs industrial funding.
In brief, that is why I feel extremely thankful for Linux, Debian, Ubuntu, Mate, Ubuntu Mate and many other developers for what they give to mere mortal end users as their outright charity.

3 Likes
20

I appreciate where your comment is coming from. I wouldn't want anyone to mis-interpret what I said above as an attempt to set demands upon the very generous developers who share their bounty with the world at large!

But do you acknowledge that such skillful donors do turn their focus away from their "job" (loosely defined as such) to look around them to see a perceived need ... and choose/want to make the effort/attempt to dedicate some of their skills (and time) to improving the world around them, as a kind of lasting legacy?

And if such is the case, wouldn't it be "attractive" to those developers to know that someone went to the effort of trying to "clarify" and distinguish the "desirables" from the "un-desirables" ... and the intensity with which such labels are applicable to the various current/emerging features (or combinations thereof)?

Wouldn't having that information make it easier for them to make their informed choice as to which "charity" would be "worthy" of their support and to which they would want to contribute?

Yes, I acknowledge that popularity does not equate to merit/quality!

But ... that is why the structuring of the questionnaire is critical to ensure the subjective is classified using a method that attempts to be objective, by breaking down the components into separate attributes of the various contributing elements! :slight_smile:

... AND ... since we do want to avoid "herd mentality" that could lead to Lemming-like extinction, if someone else were to examine the trending in terms of the adoption of various technologies/capabilities to address the specific needs, would such highlight the emergence of awareness regarding the evolution or migration of focus where such efforts could explore or be dedicated to? :slight_smile:

3 Likes