On caring for user data: NeoVim caused Vim undo files to be deleted
unsung.aresluna.org318 points by jandeboevrie 6 hours ago
318 points by jandeboevrie 6 hours ago
This story has no references that support the author's version of events, but it does appear to be substantially true that:
1. The change would break undo history, for both Neovim and Vim [see edit: this is not the really the case]
2. This means Neovim would delete data created by a different program, on another user's computer.
3. This was known before the feature was released.
4. They did it anyway.
I don't think there can really be any post-hoc justification of this.
https://github.com/neovim/neovim/pull/13973#issuecomment-789...
edit: I missed an important detail. The user specified the same path for undodir in both nvim and vim. Vim requires a path to enable the feature - there is no shared default path. The user sharing a path changes the story considerably in my view, because now this is a case of nvim deleting data created by nvim as an alternative to writing a data migration for it.
I could still disagree with that, but it makes alternatives like "just use a different path" more complicated at a minimum and really changes my read of this situation completely. I think Neovim's decisions are justfiable in this context. Maybe they could have saved the contents of the old undo folder somewhere and notified the user - arguably that would be more empathic I don't really agree they had a moral duty to do this.
The feature that was broken is called persistent undo. The docs unambiguously state that edit history will be preserved unless Vim/Neovim detects that the "undo file is no longer synchronized with the file it was written for."
https://neovim.io/doc/user/undo/#_5.-undo-persistence
So even putting aside the Vim/Neovim interoperability angle, that contract was broken. Undo files created by old versions of Neovim will be deleted by more recent versions.
The interoperability issue can't be ignored either. Neovim was touted as a drop-in Vim replacement, and that needs to be taken into account when considering how changes affect users.
So many Neovim users have started by reusing their existing vimrc. I'm sure many still have ~/.config/nvim/init.vim symlinked to ~/.vimrc to this very day. With Vim, persistent undo is opt-in. Vim users who cared enough to opt in are highly likely to have explicitly set the path for undo files too. That's because Vim defaults it to the same directory as the file being edited, which clutters the filesystem. As a result, there are many users who shares undo files between Vim and Neovim simply because they reused their vimrc, not because they made the conscious choice to do so.
The data is stored in ~/.cache which has the contract that it user-wide cached data, which also means it can be deleted without severely impacting programs.
It seems the author has too high expectations of this feature, or vim is using an incorrect path to store this is they want to make it available more reliably
That's some weird technicality that has nothing to do with the issue while also being completely false.
https://github.com/neovim/neovim/blob/fc3f0041fbe01c96f38224...
The feature that was broken is called persistent undo. The feature was inherited from Vim. In case it's not obvious from the name, nowhere does it state in the docs for persistent undo that data may be deleted at any time.
https://neovim.io/doc/user/undo/#persistent-undo
That Neovim changed the default storage path for undo files is completely irrelevant to whether the contract for persistence should be broken.
NeoVim docs say it defaults to "$XDG_STATE_HOME/nvim/undo//". Do you have a source for ~/.cache?
This reads to me like a case of blame on both sides. Putting data you don't want to lose in ~/.cache is PEBKAC, but if that fact is incidental and NeoVim would have removed the undo history regardless of its filesystem path, then the point remains valid that NeoVim is deleting user data that isn't its place to delete.
A lot of the comments here are getting caught up in legal arguments which may or may not be valid. Those are irrelevant. No one is taking the NeoVim developers to court; the post is merely warning that they knowingly accepted behavior which can cause harm. The question is whether or not that choice was responsible, not whether it's legally actionable.
So if a program decided it owned .cache and deleted it every startup you’d be fine with it?
Honestly, kinda? It'd be rude and cause a performance regression, but it shouldn't break anything.
What a weird straw-man. The comparison here is a program breaking compatibility with a file in ~/.cache/... and deleting it instead of providing compatibility. I'd do that without thinking about it twice.
I don't see it as a strawman. The comparison here is a program breaking compatibility with a file _used by another program_. If vim and neovim undo files are incompatible, why does neovim delete vim's file instead of using its own file name/extension and not caring about vim's undo? And why do they use the same subdirectory even? To me it is a case of a program deleting cached files of another program, which, as a user, I see as deleting user data that said program should have touched.
I switched from Vim to NeoVim this year, after about 15 years on Vim.
Not withstanding this incident, I would say that on the whole it has been a good experience with Neovim. (I didn’t actually ever use the persistent undo functionality in vim. Guess because of VCS it’s less needed for my use-case).
Same timeline for me, but this was my 3rd or 4th attempt at switching. Not sure if something materially improved or maybe LLMs finally just got good enough at helping me convert.
But you still have to share the undodir for both vim and neovim right? If your undo data was that precious why would you gamble the interoperability? You could just have them in different folders and it would be fine.
So the case is, you really need you undo history and want to try out neovim so you just copy paste you vimrc to the new place and accidentally delete your data? I don’t think it’s that bad personally. It’s bound to happen but it’s not automatic.
I have a far less sensationalist headline for this which is:
Two programs that save their cache in the same folder causes issues.
It was ostensibly necessary for other, arguably more important features. Persistent undo maybe just isn't that important? I certainly don't use it nearly as much as the author seems to.
Why not give it a different name, so that it would not break vim?
Just editing the primary file without corresponding update to persistent undo record already breaks persistent undo. Do you prevent users from editing their own files in your text editor?
As a Neovim user, this stopped me dead with painful realization: I may have suffered the same thing but didn't realize it. There was a time when I could not undo something, and it was after a Neovim upgrade.
Unlike Dr. Chisnall, I started my editor journey on Neovim, so it wasn't a transition that bit me. However, if the format of the persistent undo file is unstable, and Neovim just deletes it when it doesn't recognize the previous format, then it seems conceivable (to me) that an upgrade after changing the format would delete the file too.
Ouch. This is making me think about getting off of Neovim. Yes, FOSS comes as-is, but if there's an alternative...
It’s such an odd design decision.
I could see ignoring the old undo data or warning before discarding it but just silently wiping it is so user hostile it’s hard to comprehend.
It's some sort of industry standard, though, isn't it? Google Chrome on Android auto-deletes the history database silently if it deems it corrupted (e.g., due to a bad write on a system with low storage). You just open your browser one day and it's all gone.
If not, I wonder how Google justifies it.
Google is the epitome of design that doesn't care at all about the user and user experience.
If you ignore it, do you continue allowing edits to the file? Now undo history is corrupt and useless anyway. Or do you disallow editing until the user manually deletes some hidden, implementation-detail file for a feature they probably never use?
If making a breaking change to the format of the file, why in root's name would neovim not just make its own new file!? Import the vim version if that's desired into persistent-undo.neovim and then just go from there. I don't understand their choice here at all nor why you're acting like this is complicated or there was some need for them to reuse the exact same file as vim?
To support you further, why would neovim users have any illusion there would be cross compatibility between undo files? Separate undo files seems the obvious and expected choice.
Sorry for my ignorance, but what is the usefulness of undo persistence after a IDE restart? Don't you normally save your changes and finish a piece of work before exiting an DE?
Edit: some great examples in the replies here, thanks! Perhaps I should start using it in editors that support it, never gave it a thought before.
The whole point of undo is that you realize you made a mistake later. If you've done that after you've saved your changes and exited the program, you'd have no expectation of undo still working unless you've enabled this feature.
It doesn't matter whether or not you can think of a reason you would want to enable it, but it should be pretty trivial to do so. Once you've enabled it, it should work.
The reason why this feature is particularly useful in vi-likes is because they are not IDEs. I hop in and out of vi all day long. Depending on the task I might create a new tab, run a command in vi, or exit vi to run it. Vi’s appeal is that it is “just” an editor and your terminal and workstation form your un-integrated development environment.
The article gave 2 examples.
I can give some as well. IDE crashes. Computer reboots at an undesired time. Even if my work is saved, I am not “finished”. I would like my undo history to extend before file / open time.
I often use it to edit system config files, often on embedded devices where I don't have my git credentials setup, or on servers. In these cases we're already not in an ideal world, I don't have good reliable version control or change management in place. In those less-than-ideal worlds, having proper undo history is nice.
Most apps don't have it, so you might not be used to it, but have you never wished that you could undo the changes you made after you restarted the program?
Moreover, the vim history tracking is amazingly advanced. It's almost like a mini version control system. I highly recommend getting to know it.
I've used vim since the first time it appeared in Debian repos. I have been told a thousand times that NeoVim is better, more modern and solves many (conveniently never cited) issues. I just ignored it and carried on. Feeling terribly vindicated at this point as it's a feature I use regularly and have no idea that it would be an issue in NeoVim.
I've used vi and nvi and (now) vim, and tried neovim, but :! is broken and “Conform to POSIX vi” is one of their “Non-goals”.
Nvim's `:!` behaviour was a deliberate design decision, which is one of the reasons I just stick with Vim9.
What's the issue? I use nvim and regularly use :! and haven't noticed an issue.
It doesn't support interactive commands. For example if you run a command that asks for input before finishing, it gets swallowed.
It does something using neovim's internal terminal mode rather than use the file descriptors it was called with. Most commonly this breaks when a command produces an interactive prompt or uses a pager.
I've never used NeoVim because Vim has been good enough.
Vim is good enough now. The reason I switched to NeoVim in like 2015 is because it had async plugin support. With vanilla Vim at the time, if you had a lot of plugins it could get pretty slow and stuttery pretty quickly.
Vanilla Vim added it in 2016 I think, but by that time I had already fully switched to NeoVim and it didn't seem worth it to me to switch.
I used neovim because it felt way faster by default. I think treesitter runs circles around what classic vim is using to highlight syntax. That said, I never used both beyond basics, thus no horror stories either.
I haven't noticed any performance issues in vim to start with. It was fast on a 75MHz pentium for me. The syntax highlighting, same. No issues. I don't always use that though because I can't be bothered to set up anything much past the basic config.
At one point vim lacked asynchronous plugins. If a plugin was running a builder or linter it locked the editor up (from what I recall).
That fell apart when people wanted vim to do some more modern IDE kind of things like all the “… on save” stuff (build on save, test, lint, etc). I think LSP support is native in neovim as well.
I believe vim merged asynchronous plugin support a while back though, so I’m not sure how different they really are anymore.
That seems like solving the wrong problem.
Just run make from another terminal...
It was a real problem, but as GP already mentioned, Vim has integrated some version of it since Vim 8 (2016), probably at least somewhat motivated by NeoVim (forked 2014).
IIRC vim became really slow with the combination of large source files (thousands of lines of code) and language server plugins (e.g. code completion and live error squiggles).
I had a C program that was a single file 140,000 lines long back then. No issues.
I was using vim-enhanced as red hat-based distro installed it. No customization. No customization of nvim either.
I often use vim or nvim on disposable machines, so persistent undo is something very low on my list of things I want.
Tree-sitter is more performant for sure, and its potential for easier text objects is nice. However, Vim's syntax engine is more than good enough for non-complex highlighting (and many people don't want complex highlighting) and, while I don't have enough tree-sitter knowledge to explain this myself, I've seen examples of Lua versions of Vim plugins where tree-sitter is unable to handle some of the regex-based text objects present in the vimscript version.
As far as really cool uses for tree-sitter go, ast-grep is a fantastic tool. Of course it's a standalone program, so it doesn't need nvim to make use of it.
If one really needs an undo in today's day and age, use git.
It’s two different dimensions in this case. VCS doesn’t preserve your local history (in JetBrains’ terms) of changes (and it wasn’t meant to by design)
Am I missing something? Are people using persistent undo as backup?
This seems however more like of a documentation and UX problem. Neovim should warn and ask before deleting old undo files, or at least back them up, but it's not neovim's fault if people don't use reliable backup and versioning systems. Relying on persistent undo for this is kind of a self inflicted wound.
Use the proper tools for the job. Saying that neovim developers "had no concept of a duty of care to their users" is really disrespectful. Neovim's Lua API is overflowing with care, you just need to go look.
I've always assumed that undo history is ephemeral in any application I use, other than maybe an application where history is first-class in a timeline like Fusion360. I'm surprised to hear that people rely on it being persistent.
Persistent undo is useful, but when it's unavailable or corrupted for me it's like "oh too bad", because it's just a convenience. And then I read things like this I get horrified by the workflows people adapt in order to avoid learning how to use the available tools properly. Seriously, it's just vim, borg and git...
> I'm surprised to hear that people rely on it being persistent.
The feature's name has "persistent" in it.
Not using it as a backup strategy. But it doesn’t matter. Vim has a feature, persistent undo. People are using it for whatever reason. Under your blessing or not. A different app, neovim, broke vim’s feature.
They don't get a pass because you disagree with an imagined use case.
Yeah, they could have done better, but to say they "had no concept of a duty of care to their users"... Isn't it a bit much?
The entire story is a setup for a remark about a Neovim maintainer's refusal to remove a David Heinemeier Hansson quote from the Neovim homepage [0]. So with that in mind the rhetoric makes sense.
[0]: In spite of the author at Unsung saying that Neovim is "in news" for other reasons. Chisnall's account is being given in the immediate context of that news.
So you say the authors accusing neovim of being careless are themselves careless with the readers when they lure us into reading a piece that has second political intentions?
Good question...
Maybe the author of the actual blog post had different intentions in sharing the Mastodon post than the author of the Mastodon post did. But the Mastodon post was made in response to the event that the blog post attempts to separate itself from. It's a reasonable thing to do if you're more interested in the technical issue and not the political one that the technical issue is being compared to.
But the context of the original post is meant to respond to a (techno-)political issue. This can explain why the post's rhetoric comes across as hyperbolic.
It doesn't help that UX seems to be talked about in ways that borrow the rhetoric of ethics from other disciplines like law and healthcare.
Seems so. As smart as keeping tabs open in browser instead of using bookmarks, but "workflow".
Btw zen browser's folders are many streets ahead of Firefox's traditional bookmarking.
According to the rev history for VIM, persistent undo arrived in version 7.3, released in 2010. So Chisnall may have used it since 2000, but he didn't have persistent undo for at least two of the books he wrote. And it means it wasn't "maintained for almost 20 years", it's at best 16.
But it is a nice feature.
I do that by using version control. I have it hooked to my editor so that "save" is "check in". Now I have persistent, versioned, copies of all my states independent of whatever tools I happen to be using.
Type A. Type B. Save. Type C. Save. How do you get back to state “A” using git?
You do not have all your states available.
Maybe through autosave? If your editor makes a commit every time you're inactive for longer than X seconds that should solve that, which I'm pretty sure is how most editor undo stacks work at some level.
I just tried it in Nextcloud Notes on Android (FlorisBoard has undo/redo buttons), and it seems to work this way. If I type a bunch of words quickly and hit undo, it removes all of them. If I stop for a couple seconds between each word, it undos them one-by-one.
I was also a very early user of Neovim.
The way I personally remember it being positioned was “Vim, but with breaking changes.”
Yeah, this hasn't been a problem since forever (or at least I haven't noticed). They had to break the undo file format a while ago to actually work correctly for byte-level buffer change events (very important for tree sitter and other plugins listening for buffer changes). The undo file format is versioned, too, so it's not just YOLO'd; Neovim has since forever returned an error saying there's a version incompatibility. The outrage here doesn't make a lot of sense to me.
The original author of this post (on Mastodon) appears to have a bone to pick with Neovim since it's "platforming white supremacists"? I couldn't figure out what this meant. Some sort of internet drama I guess.
The “drama” was linked in TFA:
Gotta be honest, this seems like making a mountain out of a molehill. Yeah, DHH is an awful person. His endorsement was removed from the page pretty quickly, but of course not fast enough to avoid the social media mob handing down their verdict upon the thankless maintainers. I sympathize with the Neovim developer in that thread, the driveby Github commenters with seemingly no other involvement in the project are some of the least well-socialized people alive.
To be fair, it’s not hard to say “white supremacy is bad and I’m not cool with it”
It took me only 10 seconds to write.
I still don’t understand how quoting something that someone said about Neovim means that you support Nazism, just because said person happened to be a Nazi.
> only 10 seconds to write
I'm White (99.44% pure Celtic-British Isles), and I type 100wpm. Isn't it ironic.
It's also hard to stay cool and agree with the mob dogpiling you over (tbh) a triviality. Anyone who's had to have difficult conversations in the real world knows that. I'm not going to expect people to respond in the best way to such things (easy to project this expectation when you're an anonymous name in the crowd, though).
tl;dr: the website showed the quote “Full-screen Neovim looks cool as hell!” by some programmer with political views. After initial pushback it eventually got removed
> programmer with political views
It's dishonest to downplay what his "political views" entail. He compared the removal of minorities to pest extermination.
> When wolves get out of control, you shoot them. When gypsies take over public spaces, you deport them.
It doesn’t matter what his political views are, as they’re totally unrelated to what he said about Neovim. Quoting what he said about Neovim does not in any way indicate support of his political views.
In my view the core of the "drama" is about whether quoting an unpolitical statement of a person endorses the person as a whole, including their political views.
The maintainer who originally responded to the github issue clearly thinks that it does not. The reporter and Twitter mob think that it does
Whether we disagree with his political views (I hope so) and how political a project like neovim is would be questions to answer once agreement on the first question is reached
When you put an endorsement quote on your project page, you're addressing it to a large range of unrelated people who majoritarly won't even be users of your project (yet).
It's basically advertisement, we won't ever reach some universal conclusion on what any specific aspect of it means for your project, it's all in the eye of the beholder.
It's also all social norms, the maintainer doesn't get a choice on how it will be taken.
Accepting a PR from a racist is a very different thing to using a quote from a racist as advertising for your product.
If you don't want reactions to your advertising, don't advertise.
I would be very mindful about advertising endorsements from anyone at all myself.
[flagged]
The people who say that he is racist or fascist don’t believe that immigration can be illegal. That is, they believe that all migration should be allowed all of the time, and that successful countries should not be allowed to keep people out.
Feels like opal drama all over again, more than a decade later. Why is the ruby community so uniquely toxic, whichever side you’re on. (Yes I’m aware neovim isn’t a ruby community, but the drama is once again centered on a ruby personality.)
> the attitude that just because something is a persistent file on your filesystem that contains data that you might want is no reason for their program not to delete it meant they had no concept of a duty of care to their users.
That's the wrong way to look at it. NeoVIM has a different concept of care for their users. They're optimizing for another kind of care, more in line with modern expectations, which VIM did not care about (hence the fork).
It's not better or worse, just different.
This same article could've been written about how VIM has no native LSP integration or autocomplete and they don't have duty or care for their users.
Deleting important data without warning is not the same as not providing a feature users might want.
There is a flip side to this - is vim encrypting this data it keeps forever? If I delete something I consider sensitive, I might unwittingly be recording that presumed-deleted piece of data to my user profile somewhere to be deleted never.
That doesn't justify its deletion, it might justify warning the user, but unconditionally deleting data not owned by the app shared with other apps?
Vim's persist undo is disabled by default, perhaps for privacy reasons. edit: I see sebzim4500 got there first.
Stop looking for post-hoc justifications for behavior that is clearly wrong.
Second time in this thread you’re telling other people how to think and behave because they aren’t acting like you
How could you possibly encrypt it in a way that holds water? Any attacker who cares enough to get your undo file can also get your undo keys, unless you want to wire a whole system of undo-now-requires-password-auth-with-MFA
It's a feature most software doesn't even have, and it is off-by-default. If you rely on your undo history to store "important data", you are doing something horribly wrong - to the level of storing your critical emails in the Trash folder.
Persistent undo exists to recover from an accidental write-and-quit mid-session nuking some stuff you really didn't intend to delete. If you care about its contents beyond a handful of hours, you either need to adopt proper version management, or start making backups.
Reading the PR the change was needed because the old undofile format was fundamentally broken. They considered making an undofile-upgrade mechanism, but it would've caused more issues that it would've solved. In other words: stuck between a rock and a hard place.
A duty of care also means occasionally having to break things to make it better, or else you end up being stuck with spacebar heating[0] forever. As a user it does suck, but that's the price you have to pay for using actively-developed software.
User data shouldn't be silently deleted just because you think they aren't worthy enough. Especially when the name of the feature, persistent undo, and the docs explicitly promise that the edit history will be preserved.
The user data in question lies in ~/.cache, which is a directory for files that aren’t meant to last long.
> Persistent undo exists to recover from an accidental write-and-quit mid-session nuking some stuff you really didn't intend to delete.
That's one use case, sure, but not the only possible one.
> Reading the PR the change was needed because the old undofile format was fundamentally broken.
That's a good reason to have a new undo file that's completely separate from the old one, and use the new one instead, and tell users "Hey, whatever undo information you had in your old undo file isn't accessible any more through neovim, you'll have to use vim if you need to get to it."
It's not a good reason for just deleting the old undo file with no warning. All the new version needs to do is ignore it, not nuke it.
> It's a feature most software doesn't even have
"All software should be shit because most software is shit"
It is insane that people are defending behavior that is clearly abhorrent just because they like the software. Are the claims even true? If they are, you should be updating your opinions about the person writing this software.
Insane? Abhorrent?
Relax for a second. Get some perspective.
My perspective is that the Neovim maintainer implemented a change that would delete data created by another program on other people's computers. They knew this would happen in advance, and it was done anyway.
From Cambridge Dictionary:
```markdown
abhorrent
/əbˈhɒr(ə)nt/
Abhorrent is an adjective that means morally very bad, hateful, or causing strong disgust and loathing.
```
I realise that this causes strong disgust and loathing in you, but can you please expand upon how it's morally very bad to prioritise speed and ease of use over undelete capability in Neovim? Are you unable to make backups or use Git?
You are presenting a false dilemma: Neovim could have easily done both without the deletion. Willful negligence is morally very bad when it can cause user harm.
> Are you unable to make backups or use Git?
Ah yes, the old perty theft justification as applied to intentionally-inflicted data loss; "They won't mind if I take/delete this, they (should) have insurance".
Sci-fi premise. Somewhere [1] there is a message board where developers openly celebrate those feelings of dislike and disdain toward users.
On that board, silently deleting a user's files is obviously justified. What would be out of line is complaining about software silently deleting user files.
[1] Long ago. Galaxy far away.
All they had to do was pick a different file name
True. But it’s not like they planned some vengeance against the VIM backup file.
People are making it sound like this is a huge company with a product and not a grouping of curious people with a couple hours of free time on a Sunday.
Most likely they never thought about it, then it broke, then it was not impactful enough to fix on their free time compared to other stuff.
They knew about it in advance. They decided people shouldn't care about that data.
https://github.com/neovim/neovim/pull/13973#issuecomment-789...
But they hold themselves out as being a normal software project, suitable for use by casual users. They even talk about having a "stable release" and advertise various sponsors.
If it's really just a toy with no expectation of not unexpectedly destroying user data then they ought to make that more clear.
Undo files and persistent undo are not meant to work that way. They are just persistent across process restarts. That's all.
They live in ~/.cache which is defined as "user-specific non-essential (cached) data".
Despite this user's impressive résumé, they simply misunderstood the feature.
What did OP misunderstand? I am not sure how and where their understanding of "persistent undo" is different from "persistent across process restarts"?
In what way is it not a process restart if I stop the process, install/upgrade NeoVim and start a new process?
It's the part where you upgraded and apparently failed to read the release notes.
Or in the OPs case, installed a totally different application and was surprised by an incompatibility.
Nobody should have to read release notes to avoid data loss; there should never be a change such that someone who hasn't read the release notes (or missed a detail in reading them) to avoid such a problem.
The lifetime of files in ~/.cache/ is the same as what the FHS documents for /var/cache [0]:
> Application cache data. Such data are locally generated as a result of time-consuming I/O or calculation. The application must be able to regenerate or restore the data. The cached files can be deleted without loss of data.
Meaning: the persistence of such files is not guaranteed across application restarts. If vim (and also neovim) had intended for the undo files to outlive the program, the files should have been put in ~/.local/state instead -- as also explicitly documented by the XDG [1]:
> [XDG_STATE_HOME] may contain: [..] current state of the application that can be reused on a restart (view, layout, open files, undo history, …)
[0] https://en.wikipedia.org/wiki/Filesystem_Hierarchy_Standard#...
[1] https://specifications.freedesktop.org/basedir/latest/#varia...
That doesn't work here because the data was not wiped by a generic cache flush from a third party script or whatever, but a targeted hit from the application itself on files known to be persistent undo.
You just can't point the finger at file system standards, or shoulda-read-release notes or whatever.
There is no such thing as "by mistake, we historically stored persistent files in a directory with 'cache' in its name, contrary to a popular standard, so that makes it okay to trash them now".
> The application must be able to restore or regenerate the data
That's a break of the contract then, right? The application was not able to regenerate or restore.
Cache is the wrong place for a persistent undo file.
Software developers do sometimes make promises to their users, but I think these promises ought to be explicit rather than assumed. You can't simply assume a "duty of care" and expect that other people will understand them the same way you do.
(Or rather, you can, but you will likely be disappointed.)
They are open source developers, giving away free software as a gift. There is no duty here.
We can speak, respectfully, of how important backwards compatibility is to us, and ask nicely for them to give more of their time to support it when their free sodftware isn't backwards compatible. Perhaps we can even offer to help implement it.
But they have no duty to do so or to "care" for their users. (They have already demonstrated they care for their users, btw, by giving them free software.)
EDIT: I missed an important part of the story, which is that they mutated existing files in a non-backwards compatible manner. That should have been avoided, I understand where the (secondary) author is coming from now.
The argument is that developers have a duty not to arbitrarily delete data on your system that you may have relied on that was produced by another program just because they’d like to replace it with something in their own format.
I don’t care if you’re volunteering your time; if you build something whose implied purpose and success depends on other people using it, you emphatically have a duty to those people not to silently destroy stuff that was there before your software was.
It's worth pointing out that vim is not "another program" as far as Neovim is concerned. It's an earlier version of the same program. Neovim is a fork of vim, not a totally different editor that happens to share part of a name.
Forking creates new programs.
If I fork chromium, I do not get to claim that I am a chromium maintainer.
Vim and Neovim are two seperate programs.
So maybe it should have two different undo files so both can be run side by side?
It absolutely is another program, I can tell this because it has a different name. To run it I type "nvim" instead of "vim". I use different config files to control its behavior. I install a different package on my system when I want to use it on a new computer.
If it really is another program with reference to vim, it should leave undo files made by vim alone.
If it's going to nuke vim's undo files, it can't hide behind "another program".
That is the point.
I don't understand. Are you saying neovim does not nuke vim's undo files? The whole point of the article is that it does.
If that's not your point, then I have no idea what you're trying to say.
> I use different config files to control its behavior.
From what I can gather, you can use the same vimrc file to control both vim and neovim.
> giving away free software as a gift
Even a gift comes with an implicit promise that it will do no harm. Deleting important data of yours without warning is harm.
If you’re storing your important data in ~/.cache, you’re eventually going to suffer even if you don’t install Neovim.
This gift comes with the following clauses listed plainly in the license file:
7. Disclaimer of Warranty. Unless required by applicable law or
agreed to in writing, Licensor provides the Work (and each
Contributor provides its Contributions) on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
implied, including, without limitation, any warranties or conditions
of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
PARTICULAR PURPOSE. You are solely responsible for determining the
appropriateness of using or redistributing the Work and assume any
risks associated with Your exercise of permissions under this License.
8. Limitation of Liability. In no event and under no legal theory,
whether in tort (including negligence), contract, or otherwise,
unless required by applicable law (such as deliberate and grossly
negligent acts) or agreed to in writing, shall any Contributor be
liable to You for damages, including any direct, indirect, special,
incidental, or consequential damages of any character arising as a
result of this License or out of the use or inability to use the
Work (including but not limited to damages for loss of goodwill,
work stoppage, computer failure or malfunction, or any and all
other commercial damages or losses), even if such Contributor
has been advised of the possibility of such damages.It's genuinely mind-blowing to me that software can do something obviously bad, someone can point it out, and then someone will link to the license file to say they have the right to do it.
That's such an obvious category mistake that I'm not sure how to respond. It almost feels like a bad-faith interpretation of Wichary's original point.
Right. It is weird that this has to be explained, but let's make it clear: in the Before Times, probably anytime up to at least the late '00s, probably approximately no-one in or around free or open source software would have endorsed the idea that the legal disclaimers completely free even the most mainstream, self-publicising, broad-userbase open source projects from any moral or ethical obligation to have even the slightest concern to ensure that their software doesn't hurt or betray its non-paying users, even in the most harmful ways. And if one of the many and often vocal FOSS opponents of the time had started claiming that this is what open-source developers really believe they'd have rightly been seen as having veered off into the lunatic fringe. And that's because it's a, frankly, bonkers idea which is radically detached from normal human understanding of the social and moral role and obligations of volunteers, voluntary organisations, charitable givers or gift-givers. And also because it's a wildly counterproductive idea to put out there if you're hoping to increase FOSS adoption.
That does leave the question of why this idea has started to take off more recently. Part of the answer is certainly that Rich Hickey, disgracefully, set the ball rolling in this direction, and that many others have welcomed it as one weird trick and one pat answer for all the worsening problems of developer burnout. Unfortunately it seems hard to dismiss the idea that it's also social breakdown driven by a broader trend, as over time we move further and further from the pre-'60s "neurotic society" of people obsessed with duty and social conformity (often with oppressive or destructive results, to be sure) into the "psychopathic society" in which even people who don't themselves merit a Cluster B diagnosis have internalised narcissistic and psychopathic attitudes.
Equally mindblowing that folks feel someone hacking on open source software has an obligation to do anything the way they feel it needs to be done when the whole point is that anyone gets to do more or less what they want with the code.
You can't have both.
They get to do more or less what they want with the software, and others get to say more or less what they want about that. What's wrong with deciding they've crossed a line, being bothered by that, and warning others? Nobody's trying to get the law involved, or maliciously retaliate against them, or anything like that.
I think you are objecting to OP’s “duty of care” wording which could be interpreted as suggesting a legal obligation. Nobody in this thread is really arguing that open source developers have a legal obligation to do this and to not do that. We are just saying they should do this and should not do that.
> has an obligation to do anything the way they feel it needs to be done
> anyone gets to do more or less what they want with the code
so if the software had explicitly installed a root kit, you'd say the same?
If you give out free soup, you get to poison it too?
I don't think anyone is looking for legal remedies, this is not the right layer.
Is some part of:
"Licensor provides the Work (and each Contributor provides its Contributions) on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied, including, without limitation, any warranties or conditions of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A PARTICULAR PURPOSE."
Unclear? The fact that it's part of a license does not make it a legal-system-only warning.
The software might eat your dog, and feed your homework. Run it at your own risk, and be prepared to submit patches or fork it to make it behave the way one prefers.
This is one of the most exhausting (and frankly, exhausted) recurring arguments that comes up over and over again on HN.
Some bit of open source software does something bad or unwanted which causes people point out that it shouldn't do that bad thing. In this case, not even "the authors of this software should be held liable for the software doing the bad thing", just, "Hey, the right thing to do would be to update the software to not do the bad thing."
Why does this always lead a zillion people to come out of the woodwork to point at licenses and warnings or whatever? Like, yes, there's a warning. Your software having a warning doesn't mean people can't criticize you and your software for doing bad thing. Your software license does not give you immunity from criticism or from people saying you should change your software.
No, you don't have to change the software. Yes, other people are within their rights to fork the software themselves. But they can also point out that they told you that your software was doing the bad thing and you didn't fix it or change it, and that as a result they don't like you or your software or both — whatever. Nothing at all wrong with that.
I actually think you framed it really well. No parties involved have any particular obligations to each other. Nor should there be expectations otherwise without support. It sucks to lose data, everyone knows that, and no one wants it. And identifying things which can be improved is important. I stop at expecting someone else to do something because I want them to. I might hope they would, and try to convince them.
Is the fact that we are not discussing legal remedies in this thread unclear?
The software might eat your dog, and feed your homework
Sure, and when it does we can say “this piece of shit ate my dog, and the authors of the software have no concept of a duty of care to their users”. And no amount of “well, axually…” is going to make any difference. I have a hard believing someone is copy-pasting a license file in good faith in response.
Most open source software is written by individuals who aren't paid for the effort, and are solving their own problems. Presuming that they feel an obligation to the folks who download and use their work for free seems... bold.
If my neighbor mows my lawn for free, I'm not going to complain about his workmanship. If I want the job done a particular way, the solution is to do it myself, or pay someone to do it the way I like.
Suppose that you build a childrens' slide in your front garden. You put up a big sign saying "Consider using my slide! https://neovim.io/ Here are several wonderful things about it. It's free for everyone!" in your yard. Underneath in smaller letters you add "[No liability]". You also put up noticeboard ads for your free kids' slide in neighbourhood shopping malls https://launchpad.net/ubuntu/+source/neovim https://wiki.archlinux.org/title/Neovim . Unfortunately, when you built the slide, you left sharp metal edges and corners sticking far up on the inside, reaching into the path of the user. No reasonably competent and diligent metalworker or slide-maker would have failed to notice these major flaws or failed to understand the serious danger they represented. Several neighbourhood children use the slide and receive serious gashes to the legs, arms or face, and have to go to the hospital. Even assuming that your no-liability small print somehow had you free and clear legally, do you believe that your behaviour would have been ethically and morally above all criticism? Do you think that "should have read the small print!" or "can't I build what I like in my own front yard?!" would have you covered? Do you think that your family and friends would agree?
> If my neighbor mows my lawn for free, I'm not going to complain about his workmanship. If I want the job done a particular way, the solution is to do it myself, or pay someone to do it the way I like.
I think this is a decent analogy, but it works better the other way. If my neighbour offers to mow my lawn, I accept, and then he destroys the flower bed adjacent to the lawn, I will be upset, and I will have every right to complain about what he did. If he reacts by blithely dismissing my concern, then I certainly have the right (and arguably the obligation) to warn others that they should think twice about accepting his offers of gardening assistance.
Some expectations reasonably go without saying; "don't destroy my flower bed when mowing my lawn" is one, and IMO "don't destroy my data without a clear warning and a chance to back out" is another, though of course we might disagree about exactly where this does and doesn't apply.
If my neighbor mows my lawn and ruins my whole garden, I'd definitely complain. (And if I were in the neighbor's shoes, I'd feel awful about it and try to fix things.)
If my neighbor mows my lawn for free, and in the process mows my flower garden down, things change a bit though don't they. That is a closer analogy. In that case I am going to complain, and maybe also tell everybody he's careless and not to let him near their lawns.
There's a long literary tradition of representing contracts as a tool of villainy. Signing them is generally treated as a Faustian bargain.
This is a great example of why. Most humans have a sense, deep down, that contracts often exist to bridge the gulf between the ethically defensible and the legally defensible.
It's hard to imagine that any sane person who is just looking to use a popular editor would read some broad limitation of liability language like the above, and interpret it to mean, "By the way, we intend to quietly delete certain files created by a competing fork of this project whenever we find them."
It's true that contracts with liability limitation clauses like this are an absolute necessity in this day and age. But there's also a non-legal principle of mutual respect that is absolutely necessary to a healthy open source community.
> It's hard to imagine that any sane person who is just looking to use a popular editor would read some broad limitation of liability language like the above, and interpret it to mean, "By the way, we intend to quietly delete certain files created by a competing fork of this project whenever we find them."
Forks of projects trodding all over each others files is one of the more common problems that has happened, historically. Prior to the major efforts around freedesktop.org around configuration standardization, it was quite common. It'd be one of the first things I looked for when switching to a fork.
I’m not saying it doesn’t happen. I’m saying it shouldn’t be defended as good. It’s a defect. And in this particular incarnation it’s a defect that directly clashes with fundamental Free Software principles such as personal digital sovereignty.
I also suspect that few people actually believe it’s ok and these legalistic defenses are more about circling the wagons. How many people would defend Microsoft if a new Office version automatically and quietly stripped edit history from documents that were originally created by other versions? Would we be hunting for limited liability clauses in their EULA to defend the design decision?
> I’m saying it shouldn’t be defended as good.
Thankfully, that's not a thing I ever did.
> I also suspect that few people actually believe it’s ok and these legalistic defenses are more about circling the wagons.
For me it's more about healthy boundaries and expectations. If I'm somehow paying for a project's development, I have higher expectations. If I'm not, I understand that I've chosen the dev/test track and there will be bugs and issues. The developer may choose to run off in an odd direction coughGnome3cough and my only recourse is to fork or hope someone else does. Disagreements as to how things should work happen pretty often.
I'm not defending anything or anyone. Just describing the system as it exists.
“Here, take this car, it’s free!”
“Ow, the gas pedal is actually a bear trap.”
“Why are you complaining? It’s free!”
This is a very apt analogy. If you give someone a car (or sell it cheap) and you know there is a problem with the car, then it is your duty to tell them. "Hey, I've been having trouble with the breaks they need to be pumped. Be careful about that."
Then you don't take the car. Why is this confusing?
Because you didn't find out it was a bear trap until after you stepped on it?
Well I would feel slightly irate if I had to find out about the bear trap before I decide I don’t want the car.
That’s fine, you can leave the car, but you’ve already lost your foot. Why do you find this confusing?
You discover it’s a bear trap not before you try to pull out from the parking lot.
A duty, like a responsibility, can be something you feel or choose for yourself. The fact that the maintainers of some software don’t feel it is a perfectly good argument for not using that software. There is no necessary connection to compensation or transactional thinking.
If I give away food I have a duty not to poison you. It doesn’t matter you didn’t pay for it.
Others are eating it and are fine, chances are you’re allergic. Fortunately all ingredients are listed, you can check - or even modify the recipe yourself.
> Fortunately all ingredients are listed
How does this translate to the vim situation?
The point would be that it's open source. And "chances are you're allergic" translates to "your use case / expectations are unusual". But in my opinion the analogy doesn't hold well at all, because destroying a persistent undo file without warning is pretty obviously incorrect and unexpected behavior.
It's not poison, it's a broken feature. And you have the source code. Fix it if you like.
Sure, and then have it break again the next time upstream pushes a change. Or have something else break because upstream doesn't care about breaking things that I'm relying on.
The only way to "fix" this from a source code perspective would be to fork Neovim.
Software isn’t food.
So software that deletes important data of yours without any warning is ok because at least it isn't poisoning you?
If you have to twist things so far that intent doesn’t matter, something’s wrong.
According to TFA, it was completely intentional that they deleted someone’s data that had been created with a different program. Otherwise they would have acknowledged it was a bug. I do not know if this is true, only that it is what TFA claims.
It’s not great that Neovim destroyed/replaced files that aren’t clearly under its purview. They should have probably also made the consideration that a lot of Neovim users are going to be migrating from Vim.
Maybe if they’re making their own persistent undo standard, use different file naming conventions.
That said, as a fork of Vim, maybe the Neovim authors are at least partially reasonable in assuming that you’re not running multiple forks of Vim that could then potentially conflict with each other. It just sucks that they never really thought of a migration path for this particular feature.
Yes, everyone is aware of that. The parent poster was bringing up the example of food to demonstrate that our society does not agree that giving something away for free doesn't mean you have no obligation as to potential adverse effects of the thing.
Analogies like this don’t work when comparing against heavily regulated industry.
Neovim being free software is a lot different than United Airlines giving me a free flight or my doctor giving me a free consultation.
It’s very true that you can go to civil court and make a claim that a business or individual harmed you. If you do that with Neovim the judge is going to pretty quickly read bullet point #7 of the Apache 2.0 license, you’re going to struggle to prove any kind of mal-intent, you’re going to tell the judge that you paid $0 for the software, and you’ll be pretty quickly sent along your way with no damages awarded.
Only in the sense there's no explicit legal responsibility.
Do you really believe developers have exactly zero moral responsibility to their users? None at all?
Really?
Would you still believe that if you found the developers of ssh deliberately included a backdoor for hackers, or neovim was uploading everyone's code to the NSA?
I never said developers have no responsibility. Let’s not escalate. Now we are talking about actively malicious software versus a bug or design flaw.
This issue with Neovim sounds like less than ideal design, or a bug, or both, not some kind of intentional affront to users.
That is why software should have very large liability for anything software they release do? Say your software due to bug destroys some document. Surely you should pay damages caused by such bug even if quite high?
You must feel clever completely missing the point on purpose like that.
"Your analogy is false because the two things you are comparing are not literally the same, dear sir!" is such a classic Internet discussion trope that it must have a name. If it doesn't, can I please name it? Maybe Perfect Analogy Fallacy?
But they couldn’t be further from different. You have to have a health inspector visit your kitchen just to sell food at all. You need affirmative permission from the government to sell food.
The original analogy was not about selling food:
> If I give away food I have a duty not to poison you. It doesn’t matter you didn’t pay for it.
If I invite you to eat at my house, there is no health inspector and no government. I'm just giving you free food. Would you agree that I have a duty not to poison you in that case?
True, yes, I agree.
If we aren’t making the business analogy though, then Neovim has even less obligation to produce a professional quality product. When you come to my cookout the expectations are different than when you go to a restaurant.
What do you mean by "obligation"? Nobody is claiming they have some kind of legal obligation. Nobody is saying anything analogous to your "health department" analogy. I think you're arguing against something people by and large are not claiming.