Eagle Filer from Michael Tsai

[Peter_Gallagher]

Interesting new competition for DT from the programmer that made the invaluable SpamSieve: EagleFiler.

Another file database but based on Apple’s CoreData (I’m not sure if DT uses CoreData, but I guess not). It does many of the things that DT does. One obvious point of comparison with DT is EagleFiler’s simpler import facility (including a ‘print service’ to import PDFs from Preview or elsewhere). It seems to be pretty zippy, too, at imports, indexing and searching.

It’s biggest advantage, in my view, is that it’s nicely scriptable which is the main advantage I can see in DT Pro at present… although not one worth an extra $US45 to me.

EF appears to use ‘spotlight-style’ indexing rather than a concordance. A concordance may not be needed because ‘Libraries’ are not concatenated in the Database… Documents are stored in the regular file tree and can be opened in the Finder. I’m pretty certain that this will make synchronization between computers a much simpler task than in the case of DT’s database. I use Synk (decimus.net) for backup and synchronization; it uses the Spotlight database to determine which files have changed and need to be synch’d or backed-up. It looks like that will do the trick nicely with EF.

Personally, I find classification in DT is more a distraction than an aid to data structuring since I can’t use it to weight searches. I find auto-classification is too ‘hit and miss’ to be of much use. Classification may speed-up DT (I’m not sure) but CoreData seems to be un-fussed by one big glob.

EF does not have the same number of window configurations as DT (is that a limitation?) and does not claim to be ‘intelligent’ (I was never sure what that meant in DT).

It offers fewer search configurations (no ‘fuzzy’) but I find I use only one or two of the DT combos (not ‘fuzzy’) so this isn’t a problem for me. Boolean, phrase-based and wildcard searches etc.

The current version (1.0.2) doesn’t export to IPods or replicate instances (it will prevent the import of duplicates). It opens in default applications rather than an internal version (of Webkit or PDFKit)… one advantage of which is that a search for a term in a PDf file opens Preview with the search window already populated showing all instances.

I guess from the docs and general form that MT build EF as a mail archive database (it imports the valuable ‘MailTags’ tags).

Overall it looks like good value. At the same price as DT, I’m hoping that EF will force D-T to make the ‘personal’ version scriptable soon and adopt the print-services approach to importing PDFs

[ndouglas]

Tried it and didn’t really care for it.

For one, the UI isn’t good for handling large numbers of folders. The icons are larger and, instead of resizing (which would be bad enough), the folder list starts scrolling. I much prefer the smaller icons in the DT group panel, since I have a lot of base level groups. The tags, as they appear in the panel, don’t have icons but they still take up just as much vertical space. It seems like that you couldn’t have more than 25-30 tags without severely impacting the usability of the program.

For two, I was kinda excited that the screenshots showed some smart folders, until I learned that I was apparently unable to create them (?). I suppose I would be stuck with the ones already there, which aren’t really that helpful.

I’m not sure if this was subjective or objective or due to my specs, since I’m running this on a 1.2GHz iBook, but it wasn’t snappy for me. DT isn’t that snappy either, but I have a lot of stuff in it and it has better functionality. For the nearly-empty database I had in Eagle Filer, I would have expected lightning speed. It took several bounces to start up, beachballed repeatedly, and in general was laggy and unresponsive. I ran it four times, for about a half hour each time, and performance never improved.

There’s some behaviors that are either unintuitive or buggy. I created a rich text document, which was then tagged as “unread.” I clicked on the file and read it, then clicked on the “unread” tag – which listed the file as unread. Then I clicked back on the file, clicked in the text editing field, and clicked back on the “unread” tag – which still listed the file. I’d think that this wouldn’t be a particularly difficult feature to implement correctly, but I might be missing something.

There are no automator actions, unless I’m somehow making a large mistake. I like my Automator, especially when I’m dealing with thousands of files.

Clicking on the “Library” icon shows me all of the folders in the Library… even those that are contained inside of other folders (?).

I dunno. In general, I’m not too crazy about this being a 1.02 version. 0.* would be fine, but it seems too buggy and unusable for me to take it too seriously.

[mjt]

Thanks for the feedback. Let me reply to a few points:

@kalisphoenix

The icons are larger and, instead of resizing (which would be bad enough), the folder list starts scrolling.

There’s an option (in the preferences and action menu) to use small icons, as with Mail.

@kalisphoenix

The tags, as they appear in the panel, don’t have icons but they still take up just as much vertical space.

Are you saying that you want the tag rows to be thinners than the others?

@kalisphoenix

For two, I was kinda excited that the screenshots showed some smart folders, until I learned that I was apparently unable to create them (?).

Sorry, it’s not possible to create your own smart folders in the current version.

@kalisphoenix

I’m not sure if this was subjective or objective or due to my specs, since I’m running this on a 1.2GHz iBook, but it wasn’t snappy for me.

I’m working on improving performance for future releases. It isn’t as snappy on older hardware as I’d like, however I think you’ll find that adding a lot of data doesn’t slow it down much once it’s open.

@kalisphoenix

There’s some behaviors that are either unintuitive or buggy. I created a rich text document, which was then tagged as “unread.” I clicked on the file and read it, then clicked on the “unread” tag – which listed the file as unread. Then I clicked back on the file, clicked in the text editing field, and clicked back on the “unread” tag – which still listed the file. I’d think that this wouldn’t be a particularly difficult feature to implement correctly, but I might be missing something.

Marking items as read automatically makes sense in a mail program, but I don’t think it’s necessarily what people want when viewing potentially much larger documents. So it won’t disappear from “unread” until you remove the “unread” tag or click “Mark as Read.”

@kalisphoenix

Clicking on the “Library” icon shows me all of the folders in the Library… even those that are contained inside of other folders (?).

Well, that’s how iTunes, iPhoto, and Xcode work, too. It makes it easy to view and search the whole library.

[luix]

i think it’s nice for all dt-user to have a look over the fence from time to time. i’m not very near to database-stuff, but i spent a lot of time to try out several apps that could make data handling and storing comfortable and secure (from notebook, notetaker, sticky brain, yojimbo etc.).

all of them are nice and the most of them show features i really liked. the ideal app would be a mixture of all of them. and i’m sure, everyone of you would take other features. so what we need is a gigantic toolbox with an endless amount of tools and that all for not more than 50$, right?

apple seems to make it easy to build organizer and database applications and i’m confident that we will see a lot of better and better apps, that we really like to have. but least the question is »is that new app so much better that i really decide to leave my old working environment with thousands and thousands of data entries?«

may a good competition in that »database scene« make the devonian have a sharp look on the schedule for the dtp 2 version.

[ndouglas]

@mjt

Thanks for the feedback. Let me reply to a few points:

Thanks for responding, and sorry for not submitting it to you directly

@mjt

There’s an option (in the preferences and action menu) to use small icons, as with Mail.

Hunh. I never saw that, in either application. It doesn’t bug me in Mail, because I generally stay with the stock number of folders + a couple smart folders. It did bug me here. Quite an idiotic oversight, on my part, and I apologize.

@mjt

Are you saying that you want the tag rows to be thinners than the others?

If this were KalisphoenixFiler, I’d pick a different interface for the tag rows. A longer discussion is at the end, where I was talking about the collapsed folder hierarchy.

@mjt

Sorry, it’s not possible to create your own smart folders in the current version.

Okay. (More on that below, too)

@mjt

I’m working on improving performance for future releases. It isn’t as snappy on older hardware as I’d like, however I think you’ll find that adding a lot of data doesn’t slow it down much once it’s open.

Glad to hear it

@mjt

Marking items as read automatically makes sense in a mail program, but I don’t think it’s necessarily what people want when viewing potentially much larger documents. So it won’t disappear from “unread” until you remove the “unread” tag or click “Mark as Read.”

Very true, and that makes good design sense to me now that I think about it. It’s not intuitive to me, but I’ve missed some pretty damned obvious things before (as displayed already in this post) :-/

@mjt

Well, that’s how iTunes, iPhoto, and Xcode work, too. It makes it easy to view and search the whole library.

I have 1135 folders in DEVONthink right now, and I’ve been using it less than a week. I figure that it’ll probably get to at least 8000-10000, possibly more. So I’m horrified with the prospect of clicking on the library icon and seeing 1135 folders in the list pane that I’d have to scroll through

But I have this massive number of folders largely because DEVONthink’s “smart groups” are somewhat difficult to set up and I lost my criteria (in the file comments) when I rebuilt the database while stress-testing. I’d much rather have Honest-to-$DEITY smart folders and a nice flat iTunes/iPhoto/etc system. One thing I’ve learned in this big long project is that the “everything in its right place” motto is silly: everything has about a thousand different “right places.” I could put my Characters folder in my Manufactured Reality folder, several folders really should be in at least three different places, and so forth. You know how it goes.

And that’s where the tags come in. I can’t see myself ever scrolling down a left pane that has leagues and leagues of tags. I suppose MediaWiki categories might be translatable to tags, and I had 459 of them (and surely to expand to more). So the UI isn’t appropriate for me. Not that I’m knocking it in an objective sense, but I’m insane For me, a completely tags-are-invisible-unless-specified system like MediaWiki is better, or the one of DEVONthink, even though it relies on Applescripts to work. (The obvious and easy solution would be for me to keep the Tags folder closed at all times… but if I did that, you’d never know that I did it, or why)

Again, thanks for responding to my concerns. You’ve cleared a lot up for me. Take care!

[mjt]

@kalisphoenix

I have 1135 folders in DEVONthink right now, and I’ve been using it less than a week. I figure that it’ll probably get to at least 8000-10000, possibly more. So I’m horrified with the prospect of clicking on the library icon and seeing 1135 folders in the list pane that I’d have to scroll through

The library is an exception. For the lower-level folders, it just shows the immediate contents of the folder. So you could, for example, put a single folder a the top level of your library and then click on that to browse in a more Finder-style manner.

Also, you can type the name of the folder that you want into the search box for the library, and then click the Select Source button to show just the contents of that folder.

@kalisphoenix

I’d much rather have Honest-to-$DEITY smart folders and a nice flat iTunes/iPhoto/etc system.

Understood.

@kalisphoenix

And that’s where the tags come in. I can’t see myself ever scrolling down a left pane that has leagues and leagues of tags. I suppose MediaWiki categories might be translatable to tags, and I had 459 of them (and surely to expand to more). So the UI isn’t appropriate for me.

Perhaps it will be when tags are searchable?