Runbox7 zombie emails and other strangeness

Most of the time I access my emails via POP3 and/or IMAP, but occasionally I find it useful to use the web interface for operations like bulk-moving emails between folders. Normally that means Runbox6, but I have tried R7 a few times to future-proof myself in case you decide to retire R6 in favour of R7. So I am not a typical R7 user. However, I am computer-savvy, and I do find some aspects of the R7 interface completely baffling. I know it’s in Beta, but unless I’m being completely stupid, some of this stuff should surely have been sorted out by now?

My client environment is Linux (debian stable), and my browsers are Firefox and Chromium.

INDEX SYNCHRONIZATION

If I say “don’t sync” when I login, then the interface behaves predictably - ie I can click on a folder in the left hand panel, and I see the list of emails in that folder in the central panel. This is consistent with what I see in R6.

However, if I say “sync”, the weirdness starts. Now if I click on a folder, the central panel shows an apparently random list of emails. Some folders which I know to contain emails show up as being empty, and others which I know to be empty show up with random contents - including emails that are actually in different folders, and other emails that used to exist but have been deleted.

SEARCHING

I don’t know where to start with this. Firstly, as of today’s date (2020-02-22) my folders contain in total about 800 emails. However, if I sync the index, the middle panel says “Start typing to search 4985 messages”. What??? I suppose at least that goes some way to explaining why searches tend to uncover a variety of zombie results (long-deleted emails).

Anyway, let’s go back to the non-sync option, which seems to be slightly (!) less insane. A few of my emails have a From address of (eg) donotreply@XYZ, so that might be a useful thing to try searching for. If I search from:“donotreply” (as set up by clicking the spanner icon in the search field), then I get no results after 5 minutes - just a progress bar suggesting that something is happening. That’s what happens if Inbox is my current folder … Inbox doesn’t in fact contain any donotreply emails, but as I haven’t told the search to operate just on the current folder I expect to see some results from other folders. OK, 10 minutes elapsed now with no results, so let’s go to a folder where I know there are some donotreply emails. Now I’m searching for folder:"" AND from:“donotreply”; I know I should get some results, I can see a couple of donotreply emails with yesterday’s date, near the top of the displayed folder contents - however, all I’m getting is that pesky progress bar again, with no sign that it’s ever going to stop.

Well, I don’t believe that’s ever going to complete, so let’s abandon it and try something different. Back to Inbox, and try just typing something in the search field instead of going to the advanced (spanner) option. Just typing “a” is sufficient to get some results, which as far as I can see do all contain the letter “a” … however, all the results are within a narrow date window (2020-02-06 - 2020-01-25) although I didn’t set any date criteria. Also, while some of the results are genuine, quite a lot of them are zombies (emails that used to exist but have been deleted).

Does anyone have any thoughts on this? Has “search” not really been implemented yet? Why does it say “search 4985 messages” when I have only 800? Does your roadmap have any thoughts on how long R7 is likely to remain in Beta?

Hello @BillW and welcome to the forum.

You seem to have more than your fair share of indexing problems there. The most obvious thing we should do it re-index your account so that we can try a fresh index and see if that behaves better.

To do that we would need a support ticket opened by you at https://support.runbox.com so that we could get your account details from you. Could you do that please?

1 Like

I’m pleased to say that following Dave’s suggestion to re-index my account, I have not seen any further instances of the problems I described in my original post. The question of how the index got into the bad state in the first place is another matter …

1 Like

@BillW Thanks for your detailed feedback above.

It’s likely that the server-side index of your account was outdated, possibly due to an issue that has since been resolved but that still left the index in an outdated state.

If it happens again after the re-index we can investigate the index more closely and check whether there is indeed an issue that still needs to be resolved.

– Geir