| Index > Scribe > mail downloaded multiple times | |
|---|---|
| Author/Date | mail downloaded multiple times |
| Scott 31/08/2012 3:19am | I'm using filters to move new mail to separate mail stores (v3). Occasionally, Scribe gets confused and doesn't recognize that the mail on the server has already been downloaded and it re-downloads the mail from the server.
I had set the keep mail on server to 60 days, so over 900 duplicate emails were downloaded and dispersed over multiple mail stores and folders. I gradually trimmed this setting back to 10 days, but periodically, duplicate mail still gets downloaded. How can I prevent this from happening? I am near my bandwidth limits and unnecessary downloading exacerbates this issue. How does Scribe know which mail has been downloaded? Also, cleaning up the mess is a chore. Some of the duplicate mail (the oldest copy) is set as "read" and I don't wish to lose that setting. The "Remove duplicate mail" script doesn't preferentially delete the most recent duplicate, so often the resultant Trash will contain the "read" mail and the "unread" mail remains in the original folder. I've tried sorting by Received date, but I have yet to find a simple way to "Remove the most recent duplicate mail". Is there a fix for this script or a procedure that could be employed to ensure that the newest duplicates are removed? |
| fret 31/08/2012 7:52am | Occasionally, Scribe gets confused and doesn't recognize that the mail on the server has already been downloaded and it re-downloads the mail from the server. There are 2 possibilities that cause that to happen. One is the server decides to reindex all the unique ID's of the email it's storing so Scribe can't recognise what it has already downloaded (yes it DOES happen on some servers). The 2nd is that the Scribe options file gets corrupted somehow. And unfortunately that does happen too. I've been seeing a few cases lately of the entire file getting nuked and I can't figure out why. However in that case you'd lose all your account settings as well. Which doesn't sound like what you describe. How can I prevent this from happening? Can you switch on connection logging in File -> Options -> Other -> Debugging: Bytes Only + a path to a log file. I want to see if the server is changing the UID's or not. And that log will tell us. Wait for it to re-download mail and then zip and send the log file. I'll check the UID's before and after. Also, cleaning up the mess is a chore. Some of the duplicate mail (the oldest copy) is set as "read" and I don't wish to lose that setting. The "Remove duplicate mail" script doesn't preferentially delete the most recent duplicate I'll look into fixing that. It's a reasonable thing for it to do. |
31/08/2012 8:49am | > However in that case you'd lose all your account settings as well. Which doesn't sound like what you describe. The account settings aren't lost, so I suspect that you're right about the server re-indexing. > then zip and send the log file. ok. I've enabled this logging. > I'll look into fixing that. It's a reasonable thing for it to do. Excellent!! Thank you. |
| fret 31/08/2012 10:13am | [quote]I'll look into fixing that. It's a reasonable thing for it to do. Excellent!! Thank you.[/quote] That part's done. It was surprisingly easy. You can download the updated script here. Just save into your Scripts folder. |
| Scott 04/09/2012 11:22am | Thanks, the script works like a charm. |
| Scott 29/09/2012 2:46am | I'm pretty sure that you're right about the server reindexing the unique ID's. My wife is using the default mail client on the Mac and it got hit at the same time as my Scribe with duplicate emails being downloaded. It looks like the Mac client doesn't even support removing duplicate emails. Hats off to Scribe for this useful feature!
I snagged the log file containing the duplicate entries but it's huge (17 MB). It also contains passwords and stuff that I'd prefer to strip out before sending. What exactly were you going to look for? Maybe I could save you the trouble by looking myself? |
| fret 29/09/2012 8:20am | The command and reply you're looking looks like this:
UIDL What you are looking for is the values (e.g. UID192542-1093141749) all changing between two POP sessions. Most of the time they'll be almost the same lists. Maybe plus a few new ones, and minus a few old ID's. But if the list suddenly completely changes thats going to cause everything to be re-downloaded. I don't know what causes that, but it's a server side issue. Most likely the company hosting the email server did something to the email software to fix a problem. |
| Reply | |