Slashdot Mirror
A look at Thunderbird 2.0 Beta
lisah writes "Linux.com has reviewed Mozilla's first beta release of the Thunderbird 2.0 email client and says that, while it 'won't knock your socks off,' there are plenty of reasons to try it out or upgrade from previous versions. The new Thunderbird does away with the limitations of labels and instead allows users to tag emails to their heart's content, in the same vein as Google's GMail. Developers also tossed in a bunch of other useful features like customizable pop-up notification of new email, better search capabilities, and a neat way to navigate through the history of recently read emails. Mozilla developers didn't get everything right, however, since the account setup continues to be something of a headache."
When I switched from Eudora to Thunderbird the thing I missed the most was the ability to edit ANY message. Including incomming ones. For example if somebody mailed me something that was unclear I could edit it to add a sentence from myself clearifying. I really like this freedom.
Is it able to import data from other fucking Thunderbird installs yet? I'm tired of having to fiddle around with the profile folder whenever I do a fresh Windows install and need to put my e-mail back.
Come on, guys. How hard can it be to add support for that to the import wizard? It just needs to be a frontend for copying the files! That feature has been lacking from Thunderbird and its ancestors for, like, ever.
The Yasashii Syndicate ||
Does any one know what happened to message tabs. Its a feature I would really like as I become sick of having to re-find a message if I want to check another at the same time. I saw this proposed at some stage and thought it was going to be a 2.0 feature but there is no comments on it in the review. Did it get pushed back to 3.0?
:(){
If you've ever used SpamBayes for MS Outlook, you'll understand why bayesian analysis engines need to have some sort of grey area instead of just a binary spam / ham bit. With SpamBayes in MS Outlook, I have (3) results after spam processing:
"ham" - Messages which scored below a rather low value (10?) and are considered non-spam. Those messages get left alone in whatever folder they were found in.
"unsure" - Anything that falls in the middle gets moved to a "Maybe Junk" folder. For the most part, this stuff is spam, but the bayesian engine isn't quite sure. So it's worth checking for false positives (which are rare, but can happen until the engine is trained).
"spam" - Stuff in the spam folder scored so high on the bayesian value that it's almost certainly spam. The odds of finding a false positive in this folder are extremely low so I never bother looking.
Now for the real magic of SpamBayes... it remembers where a message was when it was flagged as "unsure" or "spam". If you find a message that was mis-tagged, you can tell SpamBayes that it made a mistake and it will add the message to its ham corpus and move the message back where it belongs.
(That and intelligent message notification are the two things that drive me nutz with Thunderbird 1.5 and prevent me from switching over entirely.)
Wolde you bothe eate your cake, and have your cake?
Still hard to beat mutt--I can use it over an SSH connection and it's much more responsive than any GUI or web-based mail client. It's also insanely configurable.
When I first started using it at the office, I used to joke that when it came to Word document attachments, in the time that it took an Outlook/Netscape user to open the document in Word, I was able to open the document in catdoc, skim through, confirm that the document was not worth reading and delete the message.
I'm with you on message notification, but it's not a huge issue for me -- I just check it frequently on my (Linux) desktop, and watch the Dock icon on my Powerbook. But if you know of a better Linux email client I could be using, I'm listening -- my solution works with ANY IMAP client. Read on...
I use bogofilter, but it's the same thing: spam, unsure, and everywhere else.
Advantages:
- Filters on the server, as messages come in.
- All the features you're talking about, from any IMAP client.
Disadvantages:
- Linux only (uses inotify to detect messages dropped to the "spam" folder)
- Any other filters (LKML goes here, girlfriend's stuff goes here) must be implemented on the server, currently Maildrop only.
- Filters on the server. If you have limited server resources, this may be a problem.
- Could conceivably lose mail during a retrain, due to the (admittedly stupid) way in which I handle retrains.
- While it works on any client, no client that I know of has decent keyboard shortcuts to help me out.
Basically, when a message comes in, it goes through bogofilter first, then maildrop. Maildrop looks for a bogofilter header, and drops it in the "spam" or "unsure" folder when it finds it. Otherwise, it goes to the rest of my maildrop rules, which mostly sort things into folders by mailing list.
There are also retrain folders: Retrain as spam/innocent. When a message is dragged to retrain/spam, it's retrained as spam and dumped in the spam folder. When it's dragged to retrain/innocent, it's retrained as innocent, with an extra header added (I think it's a bogofilter commandline option) to specify that it was reclassified (as it still might have a score of spam), and then is sent through the maildrop filters again. The maildrop filters look for that retrain flag, so it's guaranteed not to end up in spam this time, and gets sorted according to mailing list rules, etc.
This is where inotify comes in -- which means it MUST be a Linux server for this to work. As soon as the message appears in that folder, maildir structure guarantees it's just been rename'd in, so it's complete and safe to touch. Therefore, I can immediately retrain stuff, meaning the wait is less than a second, but I don't have to poll.
Boring implementation details follow:
The one major design bug is that I don't really know how to deal with maildir folders, so when I see a message appear in one of the retrain folders, I immediately open it, then unlink the file, to prevent Thunderbird or my own script from touching it until I finish piping it through the retrain process. I probably should be putting it in some temporary place, and indeed, maildir folders do have a "tmp" dir, but I simply don't know how to use it properly -- and I would have to rename it where I'm unlinking now. Basically, if the rename/unlink succeeds, it means I've beat the client to it. But if it fails, it means the client has done that stupid thing it does where the message is "delivered" to the folder as new, then the client marks it as read, which moves it from the "new" to the "cur" dir within that maildir.
I suppose I could build this into the IMAP server, but I like how this solution has already been ported from Exim/Courier-IMAP to Postfix/BincIMAP. I could write it as an IMAP proxy, but that's both more complicated and potentially slows down operation other than retrains -- an IMAP proxy would have to intercept and parse every line, whereas I only get to notice when an actual file is created in the retrain dir, and until that happens, my script does absolutely nothing.
Don't thank God, thank a doctor!
I've used Thunderbird for years before I found out that if you delete a message, it doesn't get deleted at all, but is just made invisible. It doesn't delete it until you 'compact your folders'. There's an option in the settings to have it do that automatically. I find this behaviour annoying, because if you don't know that you have to compact your folders (and which non-computer-savvy user does know that?) you will be left with an ever-growing, huge mail folder. I didn't discover this until my backup script started to take a long time copying my mail folders. I haven't seen this 'feature' in any other email client I've used. I hope the Mozilla team will correct this in the future.
-- Cheers!