[godelski]

  > Lets not warn anyone about that AT ALL, lets keep the feature activated and let them users lose their data

How did you conclude this?

IDK why the assumption is that safety measures haven't been created. You wouldn't mark the bug as resolved if you put in safety features, right? You *ONLY MARK AS RESOLVED* after reproducing the bug and *VERIFYING* that it won't happen again. Right? Dear god I hope this is what you do, because otherwise you are prematurely closing bugs.

[autoexec]

I'd agree with you that the fact that the bug is still open after 17 years isn't the problem, but the issue is that people are still (as of 10 months ago) running into the issue of their mail being deleted. If they'd secretly implemented "safety measures" as you suggest that wouldn't be happening.

Looking at the timeline, it's possible that they've addressed a few of the bugs that result in data loss several years ago, and it's possible that the latest guy who ran into the problem within the last year triggered it in new ways or under new conditions but it's clear that the problem of thunderbird deleting messages from the server when copies haven't successfully been saved during a move operation wasn't solved by any "safety measures" 9 months ago and it's doubtful that it's been solved now.

My guess is that because thunderbird ultimately doesn't bother to make sure that messages are successfully and accurately copied before it removes them from the server it'll only be a matter of time before someone else stumbles on some other set of circumstances which results in data loss when messages are being moved.

[godelski]

Reading the bug report it is unclear to me if the newest one is the same bug. There are also other bugs referenced that look to be fixed.

  > If they'd secretly implemented "safety measures" as you suggest that wouldn't be happening.

But we can *VERIFY* that measures were taken. In fact, easily! We can look at the references at the very top of the bug report!

  - Title: move/copying multiple imap messages to local folder bypasses offline store and redownloads messages. Need to preflight the move/copy.
    Status: RESOLVED FIXED 
    https://bugzilla.mozilla.org/show_bug.cgi?id=505456

There are others that need to be hunted for but this one was trivial to find and was implemented pretty quickly. There are also other similar bugs that weren't marked as dupes. Some of these have been marked as resolved and fixed. That leads me to believe that they just don't know what exactly this bug is because they can't reproduce. It may very well have been resolved and new issues might be completely new bugs. I mean... it has been 17 years... and TB has undergone significant rewriting. Don't you think that the software changed quite a bit in that time?

Which all I'm trying to argue is that they didn't just sit on their asses and do nothing for 17 years

[thaumasiotes]

> But we can *VERIFY* that measures were taken. In fact, easily! We can look at the references at the very top of the bug report!

>> Title: move/copying multiple imap messages to local folder bypasses offline store and redownloads messages. Need to preflight the move/copy.

>> Status: RESOLVED FIXED

This might be a more compelling observation if the bug was related to data loss. This just says "if you have a local copy of something, read from that instead of reading from the remote server".

It addresses the specific observation made in the thread that you can encounter the data loss bug even if you already have local copies of the messages, because Thunderbird ignores those, redownloads from the server, fails, and then deletes everything. Now, if you have local copies, Thunderbird won't try to redownload them from the server and the fact that the data loss bug isn't fixed won't matter to you.

You could apply this same approach to the entire bug, by guarding the "delete all of the user's emails" action instead of the "move emails that already exist locally" action. But they don't.

[redeeman]

are you being for real? did you see anything as such in the bug listing? but even IF they did put safeguards in place, the fact that this is SEVENTEEN YEARS, no warning, functionality still enabled without ANY WARNING losing people data. unforgivable.

How can you possibly justify this behavior? I understand they dont owe the world any software, fine, but dont knowingly publish stuff that KILLS PEOPLES DATA without atleast a warning

[godelski]

  > are you being for real?

Yes

  > did you see anything as such in the bug listing?

Yes

  > but even IF they did put safeguards in place, the fact that this is SEVENTEEN YEARS, no warning, functionality still enabled without ANY WARNING losing people data. unforgivable.

The software has change a ton in 17 years. Right? We can agree on this? (I mean it underwent a major revision in 2018, getting a lot of the codebase rewritten (like Firefox Quantum).

So let's consider a hypothetical situation. Suppose the problem was resolved in the almost 2 decades of rewriting BUT you still do not know what caused the bug in the first place and, consequently, can't reproduce it.

Do you mark the bug as resolved?

Now let's not sit in the hypothetical setting and act as developers. Some safeguards have been put in place (you can verify by looking at referenced issues). You've solved similar, but are unable to determine if these are the same problems or different problems (again, see referenced or use the search).

Do you mark the bug as resolved?

Your sibling commenter implied they would. Personally, I wouldn't. Marking as resolved is a promise to the user that it is fixed. But I can't make such a promise. I can't make any strong statement until I can reproduce. So yeah, it seems appropriate to me that it is marked as "unresolved" with steps "needs reproduction." That is an entirely appropriate status to me. You try as hard as you can and you implement as many safety features as you can, but you don't mark as resolved until you can verify. Unfortunately, this means issues go stale. Hell, there'll even be some noise like if a hacker or even just your dog deleted everything. We wouldn't want to assume the user is dumb and lull ourselves into a false sense of security, right? But you can only do so much.

*YOU CANNOT CLOSE A BUG REPORT IF YOU CANNOT VERIFY THE BUG*. That's the policy they are using. You may use a different policy, but that's the one they are using.

[redeeman]

> YOU CANNOT CLOSE A BUG REPORT IF YOU CANNOT VERIFY THE BUG. That's the policy they are using. You may use a different policy, but that's the one they are using.

and then I would say: YOU DO NOT STOP WARNING PEOPLE UNLESS YOU CAN VERIFY ITS FIXED

(which ofc assume you bothered warning people to begin with)

[godelski]

  > UNLESS YOU CAN VERIFY ITS FIXED

Ummm...

Yes?

Aren't we saying the same thing then?

We must be talking past one another. I'm (and others) are assuming they can't reproduce the bug. Assuming they aren't lying when they say so and assuming they've tried.

I mean let's take the trivial case. Assume user is dumb, deleted the files, made a bug report. Devs will never be able to reproduce unless user tells them they deleted everything 'on purpose'. That ends up with a permanently opened bug report no matter how much time you spend trying to fix the issue and no matter how many safety features you build in, right?

[redeeman]

no, I am saying that I agree that you shouldnt mark it fixed until you know its fixed.

I am also saying that if they dont know for a fact its fixed, they should WARN PEOPLE IT MAY EAT DATA.

its 17 years!

[godelski]

Okay, then yes I misunderstood you. I mostly agree but it's also been 17 years and what are the odds that the offending code still exists? What are the odds that it's TB's fault?

I know people report the issue but googling I can find similar complaints across all major mail clients.

I just don't think there's enough information to make strong conclusions and I don't think California cancer warning labels work. I think they teach people to ignore warnings instead.