Linux HATES Me – Daily Driver CHALLENGE Pt.1

[Ashleyyyy]

9 hours ago, Jeppes said:

Mistake number 1. People are stupid and popularity is rarely a good way of making choices.

idk about you but tbh manjaro is a nice beginner distro and it's much better for gaming than Ubuntu LTS is, simply because it's more up to date. furthermore there are gui ways for everything, even for using the AUR. 

[RanAwaySuccessfully]

I'm going to waste more of my time again by trying to see if maybe if I re-word myself a bit I'll be better understood.

 

Issues happen on every OS. But when I think about "an issue has happened" I'm thinking of the way three different people groups would try to troubleshoot it: 1. techie users who not only will try to fix the issue but also learn how their system is working behind the scenes, 2. regular users who when they stumble upon an issue resort to a google search about it to try and fix it, 3. regular users who when they stumble upon an issue just call someone else who knows better about it or completely give up and try to switch something without any issues.

 

I think I'd go in group 1. I'm fine with using Linux to this day and I've learned a bit about how the system itself works. Despite the false assumptions against me (of which there were many), I think I'm actually more used to Linux than Windows at this point, simply because on Linux I've actually bothered to read up documents about how some components work under the hood, like systemd, PulseAudio, dconf, etc. That's not to say I actually know fully how they work but it's a learning process. Now, there are different degrees of techie users, some of them like things to mostly work by default and like to learn their system gradually as they can, and some of them use Arch or something more extreme. I've never screwed up my system like Linus did, simply because I've already taken time to know what I'm doing before I do something, even in the command line.

 

Most people at my workplace probably fall in group 2. They know how their system works, but cannot be bothered to know how things work "under the hood", even on Windows. They can really only be bothered to fix the issues they encounter and just roll with it. They know the GUI, but only in so far as they need to use it, otherwise they just want their own computer to stand out of the way of what they need it to do. This is me when it comes to Windows. I don't know how Windows works beyond its GUI, I just trust that for the most part it'll work, and to its credit it usually does. Except when it breaks, I'm looking at you Windows 10. Meanwhile, Windows 7 was the most stable experience I had with any operating system to this day, and only because I took care to install a billion updates right after installing it, and then never again. Windows 7 is dead though, so there's not much I can do about that.

 

There's one guy at work in group 3. After spending a lot of time on Windows and having to deal with its issues (and after having Windows completely break), and after I installed Linux Mint on his computer just so he could recover his files, he eventually decided to just buy a Mac just so he could get away from the madness of it all.

 

If I, in group 1, who I already use Linux as my daily driver for 3 years, already run into headaches trying to learn how the system works, and I already have to deal with ALL OF THIS THREAD, then how am I supposed to recommend Linux to anyone in group 2, or let alone group 3? And if you think that the answer is "people either need be willing to git gud and switch to group 1 OR they shouldn't use Linux", then you're gonna have to accept that Linux will never surpass macOS's popularity, ever. You cannot blame these issues on the end user AND then complain that the end user doesn't want to use Linux because of it.

 

edit: just got hit with nostalgia from the Windows XP days, then the Windows 7 days. good times.

Edited November 12, 2021 by RanAwaySuccessfully

[gaesgesa]

30 minutes ago, RanAwaySuccessfully said:

I'm going to waste more of my time again by trying to see if maybe if I re-word myself a bit I'll be better understood.

 

Issues happen on every OS. But when I think about "an issue has happened" I'm thinking of the way three different people groups would try to troubleshoot it: 1. techie users who not only will try to fix the issue but also learn how their system is working behind the scenes, 2. regular users who when they stumble upon an issue resort to a google search about it to try and fix it, 3. regular users who when they stumble upon an issue just call someone else who knows better about it or completely give up and try to switch something without any issues.

 

I think I'd go in group 1. I'm fine with using Linux to this day and I've learned a bit about how the system itself works. Despite the false assumptions against me (of which there were many), I think I'm actually more used to Linux than Windows at this point, simply because on Linux I've actually bothered to read up documents about how some components work under the hood, like systemd, PulseAudio, dconf, etc. That's not to say I actually know fully how they work but it's a learning process. Now, there are different degrees of techie users, some of them like things to mostly work by default and like to learn their system gradually as they can, and some of them use Arch or something more extreme. I've never screwed up my system like Linus did, simply because I've already taken time to know what I'm doing before I do something, even in the command line.

 

Most people at my workplace probably fall in group 2. They know how their system works, but cannot be bothered to know how things work "under the hood", even on Windows. They can really only be bothered to fix the issues they encounter and just roll with it. They know the GUI, but only in so far as they need to use it, otherwise they just want their own computer to stand out of the way of what they need it to do. This is me when it comes to Windows. I don't know how Windows works beyond its GUI, I just trust that for the most part it'll work, and to its credit it usually does. Except when it breaks, I'm looking at you Windows 10. Meanwhile, Windows 7 was the most stable experience I had with any operating system to this day, and only because I took care to install a billion updates right after installing it, and then never again. Windows 7 is dead though, so there's not much I can do about that.

 

There's one guy at work in group 3. After spending a lot of time on Windows and having to deal with its issues (and after having Windows completely break), and after I installed Linux Mint on his computer just so he could recover his files, he eventually decided to just buy a Mac just so he could get away from the madness of it all.

 

If I, in group 1, who I already use Linux as my daily driver for 3 years, already run into headaches trying to learn how the system works, and I already have to deal with ALL OF THIS THREAD, then how am I supposed to recommend Linux to anyone in group 2, or let alone group 3? And if you think that the answer is "people either need be willing to git gud and switch to group 1 OR they shouldn't use Linux", then you're gonna have to accept that Linux will never surpass macOS's popularity, ever. You cannot blame these issues on the end user AND then complain that the end user doesn't want to use Linux because of it.

(In case you're wondering, I'm not making an argument. just interjecting.)

I will recommend Linux to people in group 1 and 2 but not 3. And I don't complain if any of them don't want to try it out.

I do take issue when people spread around misinformation.

Like how a lot people are saying "Linux sucks because It can break by just installing a third party software (Steam)" while implying this does not also happen on Windows.

Those people will never try out Linux anyways so they will never be a potential Linux user ever. 

However, when they go spread around that misinformation, this WILL scare away the actual potential users.

It is the same as FUD.

 

 

[RanAwaySuccessfully]

1 minute ago, gaesgesa said:

I will recommend Linux to people in group 1 and 2 but not 3. And I don't complain if people don't want to try Linux.

I do take issue when people spread around misinformation. Like how a lot people are implying "Linux can break by just installing a third party software (Steam), Windows doesn't do this!"

Those people will never try out Linux anyways so they will never be a potential Linux user ever. 

However, when they they go spread around that misinformation, this WILL scare away the actual potential users.

It is the same as FUD.

Fair point. This issue with Steam is a very very difficult thing for someone to stumble upon naturally, even with other packages. I still think it's just a small side effect of a bigger flaw of putting too much pressure on smaller distros to verify the integrity and consistency of all their packages (specially on a complex scenario like this where Ubuntu updates something that only broke on Pop OS, at least that's what I assume happened), when instead an AppImage/Flatpak would be more convenient for most non-essential software (Linux Mint already does this to an extent by having some applications available on the Software Manager only via Flatpak), but it is true that it's extremely unlikely for this specific issue to happen through the Pop Shop (or any other software manager) of all places, and even more so for the end result to be the entire desktop environment being uninstalled.

 

Also, if it was only broken for 53 minutes, then I wouldn't be surprised if only Linus himself was the one who managed to uninstall his desktop environment during that period of time, or maybe one or two other unfortunate souls. It's not great that it took them almost an hour to fix it, but I bet Microsoft (a giant corporation, mind you) would take a comparable amount of time to take down a potentially problematic download on their own store, if that ever happened.

[Gimmick21]

1 hour ago, RanAwaySuccessfully said:

There's one guy at work in group 3. After spending a lot of time on Windows and having to deal with its issues (and after having Windows completely break), and after I installed Linux Mint on his computer just so he could recover his files, he eventually decided to just buy a Mac just so he could get away from the madness of it all.

 

If I, in group 1, who I already use Linux as my daily driver for 3 years, already run into headaches trying to learn how the system works, and I already have to deal with ALL OF THIS THREAD, then how am I supposed to recommend Linux to anyone in group 2, or let alone group 3? And if you think that the answer is "people either need be willing to git gud and switch to group 1 OR they shouldn't use Linux", then you're gonna have to accept that Linux will never surpass macOS's popularity, ever. You cannot blame these issues on the end user AND then complain that the end user doesn't want to use Linux because of it.

 

edit: just got hit with nostalgia from the Windows XP days, then the Windows 7 days. good times.

lol

 

The one guy, that ran away, is a wise man.

There is another group:

They use the pc just for writing letters, browsing, YT, LTT forums, ... Install a small distro, that runs on their old system, the system never changes again -> happy people

 

Edit: People like Linus are born for software testing!

[leadeater]

5 hours ago, Gimmick21 said:

From my perspective: A PC is still a PC. it's an unmanageable complex thing and you should be able to do with it what ever you want. If no Linux distro fits 'the average joe' that's a totally fine result. But that doesn't mean Linux is bad and has to change. Maybe the average joe has to change

True, I never actually said Linux is bad. However ownership and acceptance of problems leads to solutions, anything other than that does not.

 

Thing is there's a lot of distro presented as "for the average joe" and for they most part the succeed in that, yet they still fall down and that's totally fine (as long as progress and improvement is made). Too often efforts get fragmented with "Yet another average Joe distro" heh.

 

To be honest I think Valve is going to have the best shot at that last bit of polish that is so hard to achieve, now that they have more skin in the game with an actual computer product.

[Rauten]

4 hours ago, LAwLz said:

By the way, for everyone going "the Linux community is so toxic", the Windows community is not exactly showing how friendly and loving it is in this thread either. 

The Pop_OS! Dev has put his Twitter on private so that he stops getting harassed by LTT viewers. 

 

3 hours ago, Arika S said:

Wtf is wrong with people............ raise your concerns publicly with generality if you have an issue, don't go after the devs. 

For context (and no, it's no excuse for their behaviour), it appears the main sticking point for many wasn't the error itself, but rather the dev's response.

During a web-search-turned-rabbit-hole I ended up in the LinuxGaming subreddit, in a thread about this video, and it seems like the dev's response was something akin to "well, it's HIS fault for not filing a proper ticket and not checking the required documentation etc etc etc" - more or less victim-blaming. https://www.reddit.com/r/linux_gaming/comments/qq9ei2/ltt_linux_hates_me_daily_driver_challenge_pt1/

 

The thread itself has quite a few people condemning that one dev/System76 not for the error (they just say that was horrible luck/timing), but for the response, for being completely out of touch as well as unreasonable for a distro aimed not just at the general public, but specifically at a gaming general public.

[leadeater]

6 hours ago, finest feck fips said:

-snip-

Seems like TL;DR blame shifting to me and lack of willingness to own an issue. All you're doing nitpicking and it's clear to me you understood what I said so what are you actually trying to say? We can skip over everything you just said because it's quite unnecessary. 

 

As to my point about Red Hat Satellite, I hope you actually understand why I said that, it's about owning risk responsibility. You cannot outsource risk, literally impossible, so many choose to mitigate risk(s) by controlling software package sources and versions. Pop!_OS COULD do the same but they cannot for multiple different types of resourcing reasons. That entire comment was nothing more than a tangential point about why such measures are done, to prevent the exact issue Linus had because that point in time situation would have never made it through to a production system, or in his context himself the user.

 

It doesn't matter that it was only possible to encounter the problem for 53 minutes, somebody somewhere at all times could be affected by it and it makes you (Pop!_OS) look bad. That sucks but it's the reality of the situation.

 

That is quite obviously the difference between Red Hat who entirely maintain every single official repo and also patches and builds software, it's all end to end Red Hat and that is supported by the licensing you pay to use it. That is the resourcing divide between those like Red Hat and those like Pop!_OS.

[leadeater]

5 hours ago, finest feck fips said:

I so badly wish that were the case. But blame shifting, blindly referring people to other departments, saying that things that are possible but require deep change are simply impossible, etc., are in fact extremely common in corporate environments.

Commonly tried yes, rarely work and lead to looking any good at the end of it. Not once ever has blaming an external vendor for releasing a bad patch you installed and did not test property work.

 

There's a difference between up management accepting an excuse or dropping the subject and putting a "black mark" against you or your department. You become a commonality around multiple issues and you'll know how they feel about that real quick.

[gaesgesa]

21 minutes ago, Rauten said:

 

For context (and no, it's no excuse for their behaviour), it appears the main sticking point for many wasn't the error itself, but rather the dev's response.

During a web-search-turned-rabbit-hole I ended up in the LinuxGaming subreddit, in a thread about this video, and it seems like the dev's response was something akin to "well, it's HIS fault for not filing a proper ticket and not checking the required documentation etc etc etc" - more or less victim-blaming. https://www.reddit.com/r/linux_gaming/comments/qq9ei2/ltt_linux_hates_me_daily_driver_challenge_pt1/

 

The thread itself has quite a few people condemning that one dev/System76 not for the error (they just say that was horrible luck/timing), but for the response, for being completely out of touch as well as unreasonable for a distro aimed not just at the general public, but specifically at a gaming general public.

Linus was not a victim. He suffered nothing.

Nobody is forcing him to install PopOS. He did it out of his own.

He accepted the license agreement.

Quote

ABSOLUTE NO WARRANTY

This is also true for him deleting his own DE due to a bug in software.

 

He was a user that encountered a bug. And chose to publish his findings which inevitably damaged PopOS's (and Linux in general) brand.

This is one hundred percent justified,  After all he is allowed to voice his opinions.

The response from PopOS devs of "Instead of your action that damaged our brand you should've done this thing instead which is more constructive."

This is also one hundred percent justified. After all he, the dev is also allowed to voice his opinions

[Leslieann]

13 hours ago, Arika S said:

LOL, are you serious?

 

how else would a newcomer to Linux have any idea where to start?

You're both right on this.

You have to start somewhere and it's a logical place to start, unfortunately these days too many "articles" are click bait written by people without a clue and in the case of Linux, a lot of tribalism as well.

 

 

16 hours ago, The89Lunder said:

And I partly agree with that.

Some skills and knowledge carry over. Like the basic understanding of a desktop UI and the navigation of it. Depending on usecase, propably some more skills.

Once your apps is installed, using it is not more difficult than Windows, depending on what your use is of course. Surfing and doing e-mail, music etc. is not gonna be way different. It may require you to use some other apps. (These are my findings, YMMV)

Absolutely.

This is why I say it's easy to convert "average users" (not gamers), they just need a browser, text editor, printer etc... Works fantastic for parents and grandparents once setup. My mom is running an Arch based distro and she loves it, it's been so much less hassle for her than Windows ever was but don't ask her how to install it.

It's good to try various OS, learn them find what is good and bad about them, then use what works best for you.
Even if Linus goes back to Windows (and he will for some things) his perception of Windows will never be the same.

[Leslieann]

3 hours ago, RanAwaySuccessfully said:

I've never screwed up my system like Linus did, simply because I've already taken time to know what I'm doing before I do something, even in the command line.

Never?

If you play you're probably going to nuke an OS at some point and it's probably one of the fastest ways to learn.  I've nuked pretty much every OS I've ever used, multiple times, and it's a very long list.

 

For the record, what happened to Linus is actually relatively minor and easy to fix.

To a Windows user it looked destroyed and to be fair on Windows or Mac it pretty much would be but if you've messed with Linux for a little while it's a simple command to fix it, just drop to console and reinstall the desktop environment, less than 2 minutes if you know the command or have something like Timeshift installed. That's part of the beauty of Linux, you can install multiple desktops or even go without a desktop entirely so losing one is not a big deal really. However, I'm not blaming Linus for not knowing that and with it being a fresh install it would be easier to just re-install and start over fresh.

[DrJankenstein]

8 minutes ago, Leslieann said:

Never?

If you play you're probably going to nuke an OS at some point and it's probably one of the fastest ways to learn.  I've nuked pretty much every OS I've ever used, multiple times, and it's a very long list.

 

For the record, what happened to Linus is actually relatively minor and easy to fix.

To a Windows user it looked destroyed and to be fair on Windows or Mac it pretty much would be but if you've messed with Linux for a little while it's a simple command to fix it, just drop to console and reinstall the desktop environment, less than 2 minutes if you know the command or have something like Timeshift installed. That's part of the beauty of Linux, you can install multiple desktops or even go without a desktop entirely so losing one is not a big deal really. However, I'm not blaming Linus for not knowing that and with it being a fresh install it would be easier to just re-install and start over fresh.

dd - the fastest way to end a system

[Helpful Tech Witch]

2 hours ago, Leslieann said:

If you play you're probably going to nuke an OS at some point and it's probably one of the fastest ways to learn.  I've nuked pretty much every OS I've ever used, multiple times, and it's a very long list.

 

yep. thats a gaurentee on every os.

Windows has nuked itself for no reason. Hell, sometimes windows installs will just nuke themselves on the setup.
Ive nuked ubuntu on my pi like 4 times, lubuntu once, xubuntu a couple times. 

On 11/11/2021 at 8:18 AM, InstantNewt said:

forum insists on telling the user to "just use the terminal" and distros don't bury it deep in the closet like Apple and Microsoft have done, Linux will never be a mainstream desktop OS.

it takes just as long for me to get to the terminal on lubuntu as on windows.

win key (on both) then terminal (on both) will get me there.

 

many microsoft forum posts suggest using regedits, cmd, powsh, and all sorts of hidden menus

[Granular]

Linux doesn't hate you. Linux respects you. It respects you enough to give you all the rope you need to hang yourself with.

Props for doing the series though.

[Spindel]

3 hours ago, Leslieann said:

Never?

If you play you're probably going to nuke an OS at some point and it's probably one of the fastest ways to learn.  I've nuked pretty much every OS I've ever used, multiple times, and it's a very long list.

I've never nuked MacOS/OSX.

 

I've never nuked windows, not Win95, Win98, Win2000, WinXP, WinVista, Win7, Win10. But I have had Windows nuke itself without any involvement from my side other than clicking OK on an update. 

 

I've never nuked Commodore Basic (OK I know it's a stretch to call that an OS) 

 

I've never nuked Amiga Workbench (again a stretch since it was loading off a floppy)

 

I've never nuked Linux on specialized devices like routers or NAS etc that includes having to do device specific settings after a DD-WRT flash (that had a high probability of nuking it). 

 

I've nuked desktop Linux at least 2 times (probably more times because of ignorance) back when I was maining Linux. 

 

But the times I remember nuking linux one was related to Mesa and the other was some package install that screwed up being able to boot (didn't even get to a CLI) and that one didn't even give a warning on install like the one Linus got.

[Lenovich]

10 hours ago, LAwLz said:

By the way, for everyone going "the Linux community is so toxic", the Windows community is not exactly showing how friendly and loving it is in this thread either. 

The Pop_OS! Dev has put his Twitter on private so that he stops getting harassed by LTT viewers. 

No you din not get point. Windows users REACTING to nerditity of linux user blaming people instead of weak operating system. No on in whole world whould say to you: windows users are more advanced and superior than rest of world users.

[finest feck fips]

5 hours ago, leadeater said:

Seems like TL;DR blame shifting to me

Seems like reading as careless as the writing to me

 

(More seriously, I think we just have different communication styles. I think maybe I'm too literal/hyperspecific for you and you're too contextual/vague for me. Care to see if we can overcome that?)

 

5 hours ago, leadeater said:

lack of willingness to own an issue

I think the idea that ‘blame shifting’ in ‘the Linux community’ is a serious problem for the viability of, for example, Pop!_OS, is a bit confused. To use myself as an example:

 

1. I'm not a Debian or Ubuntu or Pop!_OS developer
2. It's actually not a problem for me if the tools I like and use are not agreeable to someone like Linus
3. The people who are meaningfully responsible and do care about whether the tools they use are agreeable to someone like Linus (Pop!_OS developers, the APT maintainer in Debian) have in fact owned the issue by adding several new processes:
   • a new openQA implementation for some integration testing on Pop!_OS releases
   • patching the Pop Shop to give better messaging in case of dependency conflicts like this
   • additional gating in APT on the Debian side, with a prototype implemented already by the Pop!_OS developers

To look at this situation and say that one of the central issues is ‘blame shifting in the Linux community’ is to treat an incidental pet peeve (people on YouTube or Twitter or Reddit or here in these forums being insufficiently sympathetic to Linus' plight in the video) as somehow a causal factor in the issue Linus encountered, which is undermined by the fact that all of the ‘blame shifting’ on the web hasn't actually inhibited developers from taking steps to solve the issue, as though it were indeed, according to them, their problem.

 

5 hours ago, leadeater said:

That is quite obviously the difference between Red Hat who entirely maintain every single official repo and also patches and builds software, it's all end to end Red Hat and that is supported by the licensing you pay to use it. That is the resourcing divide between those like Red Hat and those like Pop!_OS.

Right, so some Red Hat users go to extreme lengths to control their software environments, and Red Hat has some tools on offer for helping them to do that. But this is also not a problem that Red Hat is, in principle, immune to. For example, it's quite common for customers to purchase some instrumentation product for debugging and performance monitoring, like New Relic, that they need to install in production. And the supported method from that vendor is adding their third-party repos. Red Hat isn't really in a position to tell customers ‘you can't buy New Relic’.

RHEL users are not in a position that makes them somehow immune to this kind of thing, and trying to micromanage every RPM package a RHEL server is ever exposed to is actually an insane way to try to approximate that. A sane way, if you wanted to stay on RHEL, would be to Dockerize your production applications and anything you might otherwise install from a third-party repository. Unlike any kind of repository management, which can only make it unlikely that mistakes or version mismatches across repositories made by different entities will result in conflicts, containerizing the application actually would make such kinds of dependency conflicts with the base system impossible. (It would also be way less work and money.)

Regarding corporate environments (quotation below enclosed in a spoiler tag, for my own convenience as I reply):

 

Spoiler

5 hours ago, leadeater said:

11 hours ago, finest feck fips said:

I so badly wish that were the case. But blame shifting, blindly referring people to other departments, saying that things that are possible but require deep change are simply impossible, etc., are in fact extremely common in corporate environments.

Commonly tried yes, rarely work and lead to looking any good at the end of it. Not once ever has blaming an external vendor for releasing a bad patch you installed and did not test property work.

 

There's a difference between up management accepting an excuse or dropping the subject and putting a "black mark" against you or your department. You become a commonality around multiple issues and you'll know how they feel about that real quick.

 

 

 Maybe we can make this metaphor a little more useful. Imagine that you're my boss (a fantasy you already seem a little bit engaged in ), and I work in IT supporting Linus, among other people, and he runs into this issue. If you come to me and say

 

Quote

This is unacceptable, we have to switch our staff to some desktop other than Pop!_OS, because we need our users in an environment where installing a new app cannot possibly affect the operating system in a way that could make it unusable for them.

my answer to you would probably be something like this:

Quote

Changing distros is probably not worth doing, because it won't really achieve what you've said you want anyway. While this particular issue only affected this particular distribution, this kind of issue is not distribution-specific. Making sure that such issues are in principle impossible cannot generally be achieved by switching distros. (There are some distros that can do this, but our users would probably hate them, and they would make it harder for them to get their work done. The tools exist, and we use some of them on our servers, but they're not ready for use by anyone outside of devops.)

 

One thing that we could do is keep our users away from the system package managers altogether, and tell them to use containerized app bundles instead. Pop!_OS actually comes with one such system designed for desktop apps, including Steam. If we set a corporate policy (and remove sudo rights for apt-get) that Flatpak is the only way users are allowed to install apps, they'll never be able to do this again.

If you're worried that may be too restrictive, we have another option, depending on how strictly you mean ‘should not be possible’. If you are willing to settle for ‘no one could do that exact thing and have that exact outcome, and similar issues will be much less likely’, we can add some restrictions on the use of the package manager that require more manual work to override, we can add flashing red text, and we can set up some kind of system for rollbacks.

Let me know which you'd like to do, or if you're interested in a more technical discussion of how switching distros would (and wouldn't) affect our susceptibility to this issue.

In that scenario, am I shifting blame away from Pop!_OS? I'm certainly not invested in doing so. What I'm interested in doing is replacing an unproductive analysis with a productive one, and devising a plan that can actually meet the boss's stated needs.

[finest feck fips]

6 hours ago, Rauten said:

 

For context (and no, it's no excuse for their behaviour), it appears the main sticking point for many wasn't the error itself, but rather the dev's response.

During a web-search-turned-rabbit-hole I ended up in the LinuxGaming subreddit, in a thread about this video, and it seems like the dev's response was something akin to "well, it's HIS fault for not filing a proper ticket and not checking the required documentation etc etc etc" - more or less victim-blaming. https://www.reddit.com/r/linux_gaming/comments/qq9ei2/ltt_linux_hates_me_daily_driver_challenge_pt1/

 

The thread itself has quite a few people condemning that one dev/System76 not for the error (they just say that was horrible luck/timing), but for the response, for being completely out of touch as well as unreasonable for a distro aimed not just at the general public, but specifically at a gaming general public.

Who is the Pop!_OS dev who makes an appearance in the discussion on that Reddit post that you linked? I don't see any, but I've not opened every single link where Reddit asks you to click through to continue.

 

There is something sort of similar in the discussion on the Pop!_OS sub, with Pop developer Michael Murphy (mmstick on Reddit).

 

He wrote the following comment and pinned it at the top:

Quote

I understand that this video may be upsetting some people. But remember that Linux and open source at large is a collective effort that needs your help to be better. It's okay to be upset that an issue occurred, but harassing people who work on open source software that are already doing the best they can is more likely to cause people to burn out from stress. So please just help point out things we can improve, and we'll add them to our todo list.

 

The thread you are thinking of is probably the one in which the same engineer wrote the following:

Quote

To be fair, this was because he used the terminal. The GUI correctly doesn't allow you to remove essential packages. [That our attention was directed to this issue] was a good thing though. It would have been better if he had reported the issue to us directly, but I patched apt

‘It would have been nice if he had filed a bug report before (or as well as) just putting us on blast with the video and letting us discover the issue through the harassment of his Twitter followers’ is pretty fucking different from ‘it's his fault because he didn't file a bug ticket’.

It's also relevant that he's replying there to a comment about how this started because Linus tried to install Steam through the GUI, which is why the difference in outcomes between attempting the install with the GUI and with the terminal gets a mention at all.

[leadeater]

43 minutes ago, finest feck fips said:

I think maybe I'm too literal/hyperspecific

Well yes exactly because you were nitpicking where it makes no difference at all. Anybody releasing in to for example Ubuntu 20.04 main line repos is responsible for ensuring what they put in is compatible with the entire Ubuntu 20.04 main line repos and it's Canonical who are responsible for creating the process to make sure that happens correctly. If someone manages to put garbage in then you get garbage back out therefore it was broken. A package shall be considered broken if it's released in to a channel it was not compatible with.

 

However if this was only an issue with Pop!_OS specifically then all the above still applies except replace Canonical with System76. Is System76 mirroring all the main Ubuntu Repos and version locking them so they can validate releases? No. Would that have prevented the issue? Yes. Is it feasible for System76 to do this? Likely not.

 

In no way am I being too vague, anyone releasing something in to a channel which breaks the system has released a broken package. I literally couldn't care less if it worked on their system with their builds of the dependencies because that does not reflect reality nor the end result. A perfectly intact brick through a window is still a broken window, nobody cares what the state of that specific brick actually is, the window is still broken 

[finest feck fips]

3 minutes ago, leadeater said:

nobody cares what the state of that specific brick actually is, the window is still broken 

Yes, and every native English speaker would still look at you sideways for trying to explain the situation in terms of a ‘broken brick’, and they'd be right to doubt the accuracy and usefulness of your analysis for use of such terms.

[leadeater]

1 hour ago, finest feck fips said:

RHEL users are not in a position that makes them somehow immune to this kind of thing, and trying to micromanage every RPM package a RHEL server is ever exposed to is actually an insane way to try to approximate that.

No it isn't, it's what we do, many others do and it's extremely simple to do so. You can further increase this by using Puppet or w/e other flavor of Configuration Management you want but you can easily lock a system to specific and tested known package versions and you can extremely easily version control repos in Satellite, it's one of the core purposes of it's existence.

 

You are trying to make it sound much harder than it is and something that has been successfully done for many many years for many many organizations. 

 

You still missed the point, I'm not at all saying Pop!_OS or any home user should have to do this or try to, all I'm pointing to is the issue encountered by Linus is why RHEL/RHS is used where it's used for the reasons I gave. Immunity isn't the goal, it's control and accountability and taking ownership of as many risks factors as you can. I am well aware none of this is realistic anywhere other than where it is already being done.

[finest feck fips]

  

9 minutes ago, leadeater said:

You are trying to make it sound much harder than it is and something that has been successfully done for many many years for many many organizations.

Nope. I'm actually taking the concept of possibility seriously, and you are not. Keeping dependency conflicts out of your production environment is just ensuring that dependency conflicts are not actual, not ensuring that they are not possible. The latter can be done, but it's not a matter of being more careful with your packages, even with the help of automation.

 

9 minutes ago, leadeater said:

Immunity isn't the goal, it's control and accountability and taking ownership of as many risks factors as you can.

Okay. In that case, the fact that

 

12 hours ago, finest feck fips said:

none of the measures [the Pop!_OS developers have] taken can really ensure that ‘installing X requires removing Y’ is impossible.

is not a problem for you, and acknowledging it is not a problem for anyone who wants to ‘accept responsibility’ for Linus' shitty experience.

[leadeater]

1 hour ago, finest feck fips said:

Nope. I'm actually taking the concept of possibility seriously, and you are not. Keeping dependency conflicts out of your production environment is just ensuring that dependency conflicts are not actual, not ensuring that they are not possible. The latter can be done, but it's not a matter of being more careful with your packages, even with the help of automation.

 

2 hours ago, finest feck fips said:

But this is also not a problem that Red Hat is, in principle, immune to. For example, it's quite common for customers to purchase some instrumentation product for debugging and performance monitoring, like New Relic, that they need to install in production. And the supported method from that vendor is adding their third-party repos. Red Hat isn't really in a position to tell customers ‘you can't buy New Relic’.

 

2 hours ago, finest feck fips said:

RHEL users are not in a position that makes them somehow immune to this kind of thing, and trying to micromanage every RPM package a RHEL server is ever exposed to is actually an insane way to try to approximate that

 

All these do to me is demonstrate your lack of experience with RHEL, RHS and Configuration Managed environments. All that would be done is the 3rd party vendor Repo added to the RHS list of sync'd repos and then it would be version controlled and locked and Puppet classes created that ensures only the specific allowed and tested version of the packages can be installed. None of our RHEL servers connect directly to any external repos, the only source location is RHS itself and nothing else.

 

So in fear of having to repeat myself again, this is not has hard as what you are trying to make out and not "insane". But you do realize you are now having a irrelevant debate of a process and system tool chain I never said Pop!_OS or anyone else has to or should implement.?

 

The existence of these tools and processes is the proof of the problems I highlighted existence. 

 

If you want to get back on to discussing something actually relevant then we can go back to the issue I originally highlighted around package release processes. To my very original point the only counter that really needed to be given was that the Steam package released in to the Ubuntu repos worked fine (far as I understand the situation) and only had issues with Pop!_OS because of their own Pop!_OS specific repos so a broken package really wasn't actually released, and that is a fair enough point. One of the problems I highlighted still exists however and that's the responsibility and accountability for the repos you include in your distro,  here that is System76.

 

Trying to raise that certain distros, here Pop!_OS, have upstream repos they are not in control of doesn't at all change the situation or the issue and in fear of repeating another thing again, is blame shifting. That is the risk that is taken on board by doing that and nothing actually has to be done about that, acknowledging that it exists and is not feasible to do anything about and the degree of risk from it is low so you choose to do nothing about it is actually an acceptable decision to make. Largely what System76 and many others in those situations are doing. System76 has done some new additional things which you have pointed out which is great however, and I don't know if you noticed so now would be a good time to quote my past self

 

15 hours ago, leadeater said:

Note this is largely a community issue not a developer or maintainer issue who do understand and accept these problems, they just aren't easily solved long term.

 

So your, not quoted here point, about blame shifting not preventing the developers from introducing some new measures to help doesn't go against what I said in the first place. Right now you are the literal example of a problem I highlighted, you are in "the Linux user community right"? All you've been trying to do is blame shift this entire time. You don't have to agree or like what I originally said but quite literally every point you raised is the very issue I said exists.

 

Not every problem can be solved and that is ok.

[leadeater]

2 hours ago, finest feck fips said:

Maybe we can make this metaphor a little more useful. Imagine that you're my boss (a fantasy you already seem a little bit engaged in ), and I work in IT supporting Linus, among other people, and he runs into this issue. If you come to me and say

Not at all, also your example/scenario is silly and pointless and entirely unrelated to what I said. I have literally no interest in you personally, as such this and my above will close our discussion because I can't possibly envision anything more worthwhile coming out of it.

 

P.S. I think you have a "youism" problem where you to you means the literal real you even when the context is a generalized "you", the hypothetical or non specific. I'll leave you to decipher the you word problem, yes I wrote this in the most confusing way on purpose with as many usage of the word you as possible.