news…

I’ve used careful folder / mbox structure to learn the entire organization of a large natural sciences school at a major university. This has allowed me to bootstrap an academic job out of thin air.

BUT now I am in a quandary, I also use search when looking for emails.

I have recently switched from PC EUDORA to mail.app (along with a fujitsu to macbook pro switch). I have used Eudora Mail Cleaner and imported all my 400+ filters that dutifully filed my messages for me AFTER having read them (it isn’t a problem to create a filter or two a day in EUDORA).

But the Rules based method in mail.app assumes you’ll want to automatically apply rules upon receipt of a message. When unchecked, rules don’t apply (as far as I can tell) and even manually applying them doesn’t give consistent results. Rule copy-to targets are lost and there are no tools for organizing or automatically creating rules (and mbox’s to target from within the rule pane).

So although I would like to preserve the way I work (filing via filters manually augmented by search) I cannot easily do this in mail.app.

What do you think the recent “Eudora goes open source” will mean for “filtering” in Thunderbird?

:)

DDK

We’ve set up a survey to allow product direction from the general public to add or subtract features you think are important. One of the points mentioned here is hierarchical tagging so ya’ll can vote on that.

Thanks Tim for allowing me to post. I’ll go and spam my own blog now with the invitation open to the people here for a spam jam.

Mark

tags that are accessible across a system in many different applications is a holy grail.

No kidding! Boundary-dissolving tagging, mmm.

Truth of the matter is that no folder-wise setup is perfect, things aren’t always where you think they might be, information content of certain items may change over time, sometimes things may want to fit many places at once…

That well summarizes the main usability problems I have with traditional nested file/folder hierarchies.

Seems to me the hierarchical filesystem is responsible for impinging its location-dependent structure on UI design as a primary data organization component even in cases where it’s clearly not an ideal approach. I was surprised that the UW paper made no mention of the filesystem considering how significant its influence is that way.

The file/folder/desktop metaphor is so pervasive and taken for granted that it’s hard for many people to consider viable alternatives (excluding obviously irrelevant computing contexts). And most “Finder sucks” discussions get stuck on tedious, redundant file/folder management aspects supporting that metaphor, with PathFinder possibly the current king of the Yet Another OS X File Manager castle.

The trouble with “files” is they can’t easily exist in multiple folders. The trouble with “folders” is they can’t easily have multiple parents.

Will certain kinds/amounts of (meta)data eventually be more effectively managed/organized using UIs/tools operating with a unified abstraction for creating arbitrary relationships unconstrained by the limited dimension of files/folder hierarchies?

Sorry for the length… I always get carried away when there’s a chance to discuss this topic.

If performance of Search Folders were adequate, I suppose I would not be as reliant on physical folders — but this is what I mentioned before, when cross-referencing your approach and Google’s “Don’t file, search” motto.

Truth of the matter is that no folder-wise setup is perfect, things aren’t always where you think they might be, information content of certain items may change over time, sometimes things may want to fit many places at once…

This is a touchy subject for many. Please understand that I’m not saying folders are “wrong” or attacking anyone’s method of organization, I just have yet to understand the value they provide over not using them. For all of the reasons I stated in the tutorial, it mostly appears to be the “clean desk” feeling that people like. It is a major paradigm shift to not feel uncomfortable with a MASSIVE pile of stuff to sift through. Before I started AntiFiling, I must admit it did seem overwhelming. It took a little while to get over that initial “fear” for lack of a better word, but eventually it went away as I came to realize how easy it is. It’s this uncomfortable feeling that I really want to understand as we build our email tagger.

Oh, and all of my sent items are in one place — that is where we sure do match! :))

Search folders (like Smart Folders that Mail.app and friends had for some time, and Linux’ Evolution had for a bit longer) — these are fine, but I suspect (though cannot be sure — have not tried OL implementation of them) that there is a performance impact, compared to plain folders.