Talk:Snap (software)

Request help from neutral editor

I spent a bit of time trying to edit this to be more neutral, removing wrong or unsubstantiated elements on both "sides" and all my edits were summarily reverted by some guy who just vaguely alluded to the criticisms being "valid". One of the things I removed was, for example, a quote from a random Red Hat employee, a competitor of Ubuntu. It has no place in an encyclopedia entry, which should just be fact-based. Another thing I removed was a completely unsubstantiated or cited claim that other package managers involve more scrutiny of software (which is ludicrous, as all package managers allow publishing by developers directly, not just snap.) I'm not going to waste my time getting into some petty edit war with some misinformed zealot with some weird dog in this fight, but this is why people often don't trust Wikipedia; it's dominated by people who have the time to fight these stupid battles, so I think we need some editor to step in on this article and get it fixed. It reads like a debate transcript, not an objective article. JonBirge (talk) 01:09, 26 January 2022 (UTC)

I undid those edits, please refrain from personal attacks, see wikipedia pages on civility, assume good faith and Wikipedia is not a battleground, I wrote large parts of this articles (including some positive stuff), am second in the number of edits made to it and try to keep up with information about snap so i could update it, i don't revert stuff just because if feel like it, the Red hat employee is the type of source you would see in an article about FOSS (where sources are of lower quality, the wall street journal is probably not going to write about snap). An encyclopedia does not have to be just fact based , just try to maintain npov (There are entire articles dedicated to cricism, some of them linked on this talk page, having opinions is OK), It is reasonable to argue that repositories add validation, for example if a developer stops updating a package it can be handed off to another package maintainer, I also definitely don't want edit warring (A person that does this can probably get banned) and it is against wikipedia policy, so if you insist we can start a dispute resolution process. I also don't think it's unreasonable to want people to participate in a dispute resolution process and argue for why his edits are correct (especially for a page that can get over about 8k views per month) Zash 15 (talk) 20:23, 26 January 2022 (UTC)

I apologize for getting personal; I figured you were some random vandal, not somebody who cared about the article. I get that articles can *cite* opinion, but that opinion has to be worthy of an encyclopedia article and not clearly from sources with a conflict of interest. That's why I left the negative opinion from the prominent founder of Mint Linux, but took out the criticism of a random Red Hat employee, where that person was neither prominent nor without conflict (Red Hat and Ubuntu are staunch competitors). In general, the rest of my problem with the criticism you assert is valid is that it's either completely unsubstantiated or flat-out wrong. First, there is no evidence an Ubuntu-packaged apt package has any more scrutiny than an Ubuntu-packaged snap. If anything the snap is safer because it's sand-boxed. Which brings me to by biggest point: any assertion that snap is really all that different from yum or apt because of the distribution model is flat out wrong. Second, ALL package managers offer the option for developers to directly release packages OR have package maintainers from the distribution do so. You can download software directly from publishers using apt and snap, and Ubuntu packages software for both apt and snap. The only real difference between snap and apt (or yum) is that snap packages are sandboxed and run inside a container. So, to assert that snap is somehow less safe than apt is ludicrous on its face and speaks more of the bias of the person writing than anything. For example, the lack of any such criticism in the "flatpak" article is conspicuous, even though flatpak and snap essentially work the same way. So, it's clear to me that a lot of the criticism in the snap article is just people with a problem with Ubuntu and snap, and most likely, the closed nature of the snap store. In fact, the only valid criticism in this article is on the closed-form nature of the server software, but even that is slanted and makes it sound (incorrectly) that snaps can't be distributed or used without the use of closed-source software. Nonetheless, I left that criticism in because it at least had some truth to it and wasn't just blatant misinformation. If you want to go through and refute any of this I'm happy to listen with an open mind. I probably should've had a talk page discussion before I did my deletions, for that matter, but some of this was so blatant I didn't think it needed it. JonBirge (talk) 19:11, 5 February 2022 (UTC)

I concur this article could really use more neutral contributors. I refrain from further reworking the article since Zash 15 claims I am too biased because I create snap packages. I do not have much faith in Zach 15's neutrality, however, as most of their contributions have centered around adding or re-adding criticism. Therefor, I would really like a more neutral party to take a look at this page. Merlijn-sebrechts (talk) 00:28, 3 December 2022 (UTC)

Potential conflict of interest and the article is structured in a way that criticism is somewhat hidden

one of the users here (Galgalesh/Merlijn-sebrechts) disclosed he is affiliated with canonical and made some substantial edits to this article seemingly removing or hiding criticism of snap . the complaint about the contributor license agreement was removed, it’s inability to use third party stores was changed to calling it “currently” the only store (I see no evidence of this, a citation for a bug making this request that was closed as a “won’t fix” got removed from the article, another quote from canonical staff saying it is not worth while for them to open source the server was also removed).

another problem is that some of the content removed from the criticism section got moved to under the “functionality” section, that means that when looking at the headlines it is easy to miss that snap is criticized (I would not expect criticism to be under “functionality”). I agree that the criticism section got too big and this type of laundry list of problems does not belong in a encyclopedia but criticism should be under a “reception” section at least so a reader with a general interest in snap could jump to it to get a overview of what are the pros and cons of snap without having to read a detailed description of it’s functionality in order to discover that.

Here is the state of the article before the user made these edits.

we should assume good faith but as the wikipedia conflict of interest policy says:

"Editors with a COI are sometimes unaware of whether or how much it has influenced their editing"

So I think we should be diligent and consider if we are still maintaining NPOV. Zash 15 (talk) 23:05, 9 March 2021 (UTC)

This article reads like pure marketing. I was quite shocked how skewed it is, in stark contrast with the constant criticism in the Linux circles. It reads very obvious as a PR piece by canonical. Favorable comments are cherry-picked, the little bit of critical voices are somewhat devaluated e.g. with statements like "while acknowledging his own bias". Where is the full disclosure of the bias of the editor, working for the company that develops snap? The decision of the Mint team was maybe announced by Clement Lefebvre but certainly supported by the whole team. The way it is written reduces the conflict of "major ubuntu-based distro will take out anything snap-related" to "some dev has issues with snap". The adoption section is also needlessly inflated. I hope that big parts of the original section can be restored. I understand the problems with criticism sections as state below, but this piece of software if one of the most controversial things that came out in the last decade certainly up there with systemd. And it is still running hot. Whereas this article reads like PR-speech after valid criticism has been deleted or undermined. Not a good look. --Bernhard Livermore (talk) 20:48, 30 August 2021 (UTC)

I agree sections of this read like pure marketing, but other sections read like petty biased sniping by somebody who doesn't like snap. All in all, this article almost needs to be rewritten to just state the facts. Do any other article on a software distribution package even have a "Reception" section??? I took a crack at removing a few obvious bias issues, but a lot of works needs to be done to just make this a neutral statement of fact, not a debate transcript. JonBirge (talk) 03:12, 25 January 2022 (UTC)

Hey, I don't usually participate in wikipedia beyond a few light formatting fixes and corrections to a translation or two but yes this article seriously needs to pick up on the criticism, as a daily linux user for the last decade Snap is, at the best of times, offputting, and I've yet to talk to any ubuntu user who doesn't have a visceral distaste for it. It's the kind of bad that makes it hard to write about how bad it clearly is because it evoques very strong feelings on people who are acknowledgable and motivated enough to go write wikipedia articles.

And even if I didn't have criticism, some of its issues are so glaring they're not criticisms as much as just pointing out how it functions. It's engineered to be invasive, and canonical has spent the last several years being more and more overbearing with it; It has a purposefully complicated process to uninstall it requiring at least five separate commands for a task that could be trivially automated (As you have to manually remove it's core packages in a very specific order before being able to apt remove) and reinstalls itself automatically with every dist-upgrade, with the 22.00 version making firefox automatically install the snap package without any way to opt out of it that doesn't risk messing up the entire dist-upgrade, a process that can take hours.

It's worse than terribly insecure, it presents itself as secure and may lull a less experienced user (The ones who need that built in security the most, for obvious reasons) into a false sense of safety. What few security features it has seem, at best, to be hurdles for developers who then have to deal with bug reports because of someone else's shoddy coding, as the concern was clearly more on advertising the product and making it border on trashware than making it worth using.

It's priorities are just too obviously skewed and more focused on constantly trying to force itself down your throat until you give up and stop uninstalling it than on delivering something anyone would want to use, to a degree that makes Microsoft look like cherubs. There's maintaining a neutral point of view when writing an article, and there's ignoring obvious malintent in software design. 45.173.17.114 (talk) 01:42, 14 August 2022 (UTC)

I am not affiliated with Canonical and I have never said that. You can see the discussion on my talk page.

I do have a connection to Snap: I occasionally contribute to Snapcraft and I maintain a bunch of snaps. Recently, I've been elected to the Ubuntu Community Council, although that has very little to do with Snap. If this is enough of a "relation" to trigger Wikipedia:Conflict of interest, then by all means add a banner. I would be surprised, however, if maintaining A Debian or Fedora package is a good reason not to contribute to those respective pages.

It seems this thread mainly attracts people explaining their dislike for Snap instead of an actual discussion about a possible conflict of interest. Merlijn-sebrechts (talk) 01:03, 3 December 2022 (UTC)

Criticism

The criticism section is a mere irrelevant opinion of one unknown person. It's not encyclopedic at all and should be removed. 47.62.157.23 (talk) 14:11, 24 October 2018 (UTC)

The removed material can be found here: https://en.wikipedia.org/w/index.php?title=Snappy_(package_manager)&oldid=864903343#Criticism — Preceding unsigned comment added by 174.240.131.71 (talk) 22:39, 22 January 2019 (UTC)

The screen recording application has problems due to the snap confinement. the Peek developer sentence is not as reported in wikipedia:

"Using snap is more time-consuming than Flatpak or AppImage for developers"  

but he says in reddit:

"The current Snap sandboxing does not allow Peek to access any screen recording capability in current Wayland implementations. Screen recording on Wayland still is a difficult topic anyway, but there are some compositor specific solutions available. But Snap does not allow me as an app developer to punch the necessary holes into the sandbox that would allow Peek to utilize those solutions" 

and

"Setting up and maintaining the Snap build system took me way more time then the related work on Flatpak or AppImage. There are still unresolved Snap packaging related issues in Peek I am not willing to spend any more time on debugging" 

sentence that was then wrongly reported from OMGUbuntu in the way you can see in wikipedia. this still remain a criticism of one single developer for a single specific application with special authorization/accessibility needs ...so why was it reintroduced as a relevant opinion? why don't you talk also about Visual Studio Code statement :

“The automatic update is the biggest benefit and we like the way they run seamlessly in the background.” 

or GitKraken:

"We’ve saved a lot of development time. Not only by coming to the Snapcraft Summit to accelerate our progress, but looking forward, the aim is to eliminate the need to target all the different platforms thanks to snaps’ cross platform approach"

or Plex

"The biggest appeal of Snaps is the simple installation mechanism”"

or JetBrains

"Snap packages seemed exactly what we need, and we’re happy that now our Ubuntu users can easily install an IDE from a desired channel and forget about updating the builds as the updates come in the background automatically"

or many other but only the peek criticism? --151.20.97.45 (talk) 13:58, 25 May 2019 (UTC)

I did a big refactor and cleanup of the page. I merged the (properly-sourced) content of the "criticism" and "security issues" the main article. See the entire refactor.

All encyclopedic content on Wikipedia must be written from a neutral point of view (NPOV), which means representing fairly, proportionately, and, as far as possible, without editorial bias, all the significant views that have been published by reliable sources on a topic.

Criticism sections are discouraged because they focus on the negative viewpoint. Wikipedia pages should be written from a neutral point of view. Please stop adding a criticism section! Instead, add properly sourced criticism in the body of the article. A number of tips for when you want to add criticism:

• Do not give undue weight to the comments and opinions of individuals or of a minority group. Indicate the relative prominence of opposing views. For example "Developer X says that ..."
• Make sure criticism is properly sourced. reliable source are especially important for controversial things. You should not add original research. If no reliable sources can be found, it should not be on Wikipedia. In other words: it's not enough for something to be a fact; it must also have appeared in a reliable source.
• Use nonjudgmental language.

Wikipedia is an encyclopedia, not a rant-o-pedia. ;)

For transparency; I have made some small contributions to the snapcraft build tool and snapcraft documentation, and I maintain a couple of snaps. I am not affiliated with Canonical and don't have financial ties to Canonical or Snap.

Galgalesh (talk) 18:35, 7 August 2020 (UTC)

Snap is a pretty controversial technology, in particular the way it uses the $HOME directory, as is evident from the very long discussions on a number of bug-tracker discussions, such as the nearly 300 comments on bug 1575053. In the light of the severe criticism, the complete absence of a carefully written “Criticism” section, where contributors have an opportunity to elaborate on the fundamental architectural problems with the snap approach, looks highly biased and is certainly not in line with the practice of other Wikipedia articles about controversial topics. The current article with the criticism section removed looks suspiciously white-washed by someone with commercial interests in the topic. E.g., there is no mention that one of the main problems with snap is that its approach appears to be fundamentally incompatible with e.g. NFS-mounted home directories, and therefore excludes a lot of users in organizations that use such a setup from accessing software that is increasingly only available in snap form. Markus Kuhn (talk) 18:05, 23 October 2020 (UTC)

 Disagree I am the one who added the "Criticism" section for the first time in 5 April 2018 and re-added it with little improvements in 23 January 2019 when it was removed by IPs/other users, but currently I believe that such section should not be added because as user:Galgalesh said "these are discouraged because they do not present a neutral point of view", also the article of the competing technology (Flatpak) does not have such section while it has some/many problems same as Snap. -- Editor-1 (talk) 04:47, 22 November 2020 (UTC)

there is no mention that one of the main problems with snap is that its approach appears to be fundamentally incompatible with e.g. NFS-mounted home directories
— User:Markus Kuhn

If there is no reliable source stating this is "one of the main problems with snap", then it should not be on Wikipedia. If there is, then it can be added to the body of the article.

I agree that Snap is pretty controversial but only credibly-sourced criticism should be added, with appropriate weight and in the body of the article. As comparison, Microsoft is a very controversial company, but its article doesn't have a criticism section either.

Galgalesh (talk) 11:09, 19 January 2021 (UTC)

micrsosoft does have a mention of it's criticism, in fact an entire article is linked to under one of it's sections, google and amazon articles do have sections dedicated to criticism that appear under their table of content and also appear at the start of the article (the lead section) and also link to entire articles (1 2 ) dedicated to criticizing them. regarding criticism sections being "discouraged" the only thing i found was this article but it says on the start it is just a article and not a policy or guideline (besides one of the policies of wikipedia is that there are no firm rules) . I also left a message on your talk page. Zash 15 (talk) — Preceding undated comment added 22:53, 19 January 2021 (UTC)

I take back my comments about the Microsoft article, I was not aware criticism articles were so prevalent.

WP:Criticism is indeed the main source for me saying "criticism sections are discouraged" but as you say, it is not a policy. The "Neutral point of view" policy, however, also talks about it in WP:STRUCTURE:

"Segregation of text or other content into different regions or subsections, based solely on the apparent POV of the content itself, may result in an unencyclopedic structure, such as a back-and-forth dialogue between proponents and opponents"

The policy also says

"Try to achieve a more neutral text by folding debates into the narrative, rather than isolating them into sections that ignore or fight against each other."

Finally, the policy links to Template:Criticism_section which says

"This article's Criticism or Controversy section may compromise the article's neutral point of view of the subject. Please integrate the section's contents into the article as a whole, or rewrite the material."

In summary, criticism sections are not prohibited but they are a potentially problematic structure in terms of NPOV.

Regardless of the rules, I think this article is improved by ordering the content based on subject instead of Point Of View. The criticism sections kept attracting unencyclopedic content with unreliable or no sources. This hurt the quality of this article. Some of the content was written with editorial bias and without proper context hurting the Neutral point of view of this article. Merging those sections into the body fixed those issues.

Galgalesh (talk) 15:05, 20 January 2021 (UTC)

What's the point of using GNU/Linux, if you install snap?

Snap gives users somewhat quicker access to new versions than the package managers like apt. That seems to be the main point of using it.

On the other hand, the server side isn't Free Software in the FSF sense; it's closed and proprietary. Also, updates are installed automatically without explicit user consent (or even user awareness) - especially annoying when updates introduce new bugs.

GNU/Linux is all about Free (non-proprietary) software, and giving the end-user more control. If these don't matter to some users, one has to wonder why they're using GNU/Linux in the first place.

These are obvious points, but probably inappropriate for the Wikipedia article because I can't find sources acceptable under Wikipedia norms. Can anyone point to a good source for them? If so, please update the Criticisms section. — Preceding unsigned comment added by Longitude2 (talk • contribs) 22:50, 7 July 2020 (UTC)

If there are no reliable sources which show this is a significant view then it should not be on Wikipedia. Also, please don't re-add a criticism section; these are discouraged because they do not present a neutral point of view.

Galgalesh (talk) 18:41, 7 August 2020 (UTC)

Change of heading structure to include a 'Support' section

I restructured/renamed the heading 'Universal Linux Package' as it seemed to be contradicted by its own content and the top level sections didn't mirror comparable topic pages. I struggled to reason about associated image and found the move necessary to promote it to a top level heading, left a nasty gap of text around the Snap Craft widget (the first instance of a second instance I have ever witnessed), but found pushing the 'Snap store' section down as well fixed this and may even improve the logical order.

There is a lot more that should be done, but this simple change does a lot. James Bateaux (talk) 20:52, 28 March 2024 (UTC)