Feature request: missing "one-of" tag operator

Michael,

I love EF - you know I will buy it in a few days.

Meanwhile - something I think is missing (but it may be me not knowing how to do something) …

The “Untagged” pre-provided smart folder isn’t useful as it is because new files automatically get the unread tag which means they don’t appear in Untagged until you have read them - since you have to find them to read them this is catch 22 ;-).

So I was playing with smart folders to try to work around it by selecting records that have no tags or exactly one tag that is that one - and as far as I can see it can’t be done with the logical functions provided. My logical thinking tells me why its not possible too - if you look at every expression language used for saying things about data every one of them has three functions

none-of
one-of
one-or-more-of

(though some of them use zero-or-more-of for the last one above). Examples are UML, talking about instances in DTD editors, grep/sed/general-unix regular expressions and so on. This is because you need all three functionalities in order to be able to express any possibility - with only two of them it can’t be done (imho). Now looking at what one can do with tag expressions in smart folders I can’t find a way to do “exactly-one-of” nor any other way to express “no tags or exactly one tag which is the unread tag”.

If you or somone else can find a flaw in this argument I’ll be very happy to be corrected but as it stands I think not everything one might want to do can be done for the reasons I’ve stated.

What do you think ?

PS: not on this issue but something else - I’m finding one or two minor bugs in tag creation particularly. They won’t put me off EF but should I presume you know about them or do you want them reported? (they are so obvious I’m sure you’ve discovered them). Is there a list of known bugs you are working on somewhere ?

andy

ps2. I would never post on a forum without copying the text first in case of time-out while editing but I love the autosave on this forum. All forums should do this :-).

My thinking here was that one would click on “unread” to find the “unprocessed” records, and then if you marked it as read it was appear in Untagged. So it doesn’t seem to me to be a Catch-22 unless there are records that you intend to tag before you read them and you don’t want to tag them at the time of import.

Yes, I think you are right that there is currently no way to make an “Only unread” smart folder, except to manually make a list of all the tags that you want to exclude. (You could do this by selecting all the tags in the source list, copying, pasting into a text editor, replacing all the newlines with spaces, copying, and pasting into the smart folder window—a pain, but it works.)

What would you suggest as the most elegant new operator to solve this (and enable other types of interesting smart folders)? I’m thinking maybe “Tags exactly match” and then you would provide a list of one or more tags.

There are no known bugs with tag creation, so please send an e-mail.

Thanks for responding. Sorry the response quoting below is messed up.

Ah - well here you have the guy with a use case that doesn’t fit your assumption. The reason is I often have documents I don’t read immediately that need tagging (I don’t even read every document I have to keep - I can’t, I get too many - but I have to have them available). I also sometimes download documents in batches, maybe all the ones to do with one particular standards group - then I might not read those particular ones until I go to some meeting and need to but meanwhile I fetch others for other standards groups that I do read immediately. Sometimes I read documents with my imap email set-up on say my ipad then they come in to my pop mail on my laptop and just need to be tagged and stored not read. For me personally the unread tag isn’t useful (would be for some people) but the untagged tag would be very useful. First thing I did was look for a way to disable the unread tag but couldn’t find a way (obviously I could “ununread” every file but but of a pain).

That would work fine. Another way would be to provide a tag count (but I think that wouldn’t easily solve every case - it would solve it but only with complicated expressions for some things).

ok - I have some definite repeatable features so I’ll mail you.

Thanks again for your excellent responsiveness.

andy

You can do this with the MarkNewRecordsAsUnread esoteric preference.

Thanks. Whilst my argument about power of expressions still stands (the 3 kinds are still needed) this does fix my requirement to find untagged files whether read or not. Thankyou.

I’ve emailed you a bug report but it may be in a spam box.

Another small new feature follows

andy