Info panel

It’d be nice if the Info panel had collapsable sections. That way it could more effectively use screen real estate. For instance, let’s say I mostly just want to keep it open all the time to add Comments, I could collapse all the other info and maximize my screen space.

Good idea.

A bit thinner and like the Finder Info would be great too. And how about the URL and Path buttons in an Address Book style, within the field label? :slight_smile:

Actually, I prefer the current Info panel.

I’ve got my window sizes configured to display view windows to the left and the Info panel to the right. I normally have the Info panel open all the time, so that when I select a document the Info data is instantly visible. I would hate a collapsible panel, since I often would need to expand it to find what I need. :slight_smile:

This works well for me, whether on my 15-inch MacBook Pro or my PowerMac with 24-inch screen. DT Pr remembers the window sizes I’ve established and pops tham up each time I launch a database.

Download a picture of this setup at The image is “Modified Vertical Split.jpg”.


I think the panel could have collapsible sections while remembering it’s state. If it’s entirely expanded, keep it that way for all new info panels.

I hope there’s some interface cleanup with v2. :slight_smile: Companies like Apple and OmniGroup have really shown us how nice interfaces can be. :slight_smile:


I agree with your point about wanting to see all the info but I was suggesting an Info panel that could have multiple sections expanded or collapsed. So, for instance, you could expand them all and just leave it that way, whereas I could collapse all but the Comments section. Every user could make his own call and vary it as tasks require. Likewise it would remember what was expanded and collapsed so if you closed and reopened the Info panel it would be just as it was before, ready to go. Just as the Finder’s is right now. I was NOT suggesting a tabbed Inspector type of panel ala Pages & Keynote – that would be frustrating as it can only show one section at a time.

We always welcome suggestions for interface improvements. There have been hundreds of user-requested modifications.

But DT Pro is a Cocoa application that is tied into the OS X environment in ways that try to use standard Apple conventions as much as possible.

For example, DT Pro allows writing and editing of rich text documents with the available tools of TextEdit. Open a new rich text document and there’s a toolbar with those familiar elements, which perform precisely as in TextEdit. Which is to say that character formatting, styles, lists and tables work in the Apple way with no surprises, and in as simple and intuitive way as possible.

DT Pro manages collections of documents. In my main database, about 20,000 of them. There are many document types in my database. DT Pro is capable of capturing text from a number of them. Some of them are rendered like the original file, such as text, RTF, RTFD, PDF, image and QuickTime media. Others are only partially rendered, such as Word files. Still others cannot be directly rendered, such as Powerpoint or Pages files. I’ve got all those example file types in my database, and many more.

DT Pro allows one to organize documents in a traditional hierarchical way, as groups, which are analogous to folders in the Finder, with documents analoguous to files in the finder. Please note, however, that those are analogies, not strick correspondences. A group is not just like a Finder folder, nor is a document just like a Finder file. There are reasons for the differences that exist.

Example: Multiple documents in the same group can have the same name. That’s not allowed in the Finder, but is allowed in DT Pro. Those documents may or may not have the same content. If they do have the same content they may be marked in blue colored names as duplicates, or in red colored names as replicants. The concept of replicating a document is foreign to the Finder and is not the same as an alias of a file. For example, a file may be replicated multiple times, but there is really only one instance of it, with very low overhead for managing the additional instances. Modifying any results in modifying all. Any one of them can be deleted without affecting the others (although if there were only two, the remaining instance will no longer be a replicant). Groups may also be duplicated or replicated. These and other characteristics of groups and documents exist for good reasons and give the user advantages in managing and manipulating the information contained in a database.

How can one “look at” the contents of a database? DT Pro provides 6 ways of viewing the groups and documents in a database, as listed in the View menu. Like the Finder, one can double-click on a group to open it in its own view window (I have my preferences set to Alway open groups in a new window), and any of the 6 view types is available for that window. Unlike the Finder, DT Pro allows one to set a preference so that the database remembers the status of all existing windows (except Search) and will reopen them the next time the datbase is launched.

Of those 6 views, the first three correspond to some simple Finder concepts (the Columns view rather approximates the Finder) and the last three are more like working environments. Note that one can add additional columns to views using Views > Columns, as well as add sorts that are not represented in the available columns by using View > Sort.

My usual working environment is the Vertical Split View, but I may also have open a couple of Three-Panes windows (my top-level view, my Incoming group view and my Bookmarks group) and several document windows.

Scenario: I’m going to create a new project group at the top level of my datbase. Then I’m going to create a new subgroup to hold my new article document segments. Then I’m going to create another subgroup into which I’ll place duplicates of existing documents in my database that I’ll use for reference purposes, pelrhaps excerpting some quotes and creating a citaion to be used for footnotes and’or endnotes when I polish the final article in another application such as Pages. Then I’ll create still another group into which I may add now reference material garnered from web searches, etc.

Question: How Mac-like, simple and intuitive are these various actions?

Step 1: Create a new group at the top level of my database and name it. How to:

[1] From the menu choose Windows and select the top-level view, or use Expose to do that. If the top-level view isn’t available, choose File > New Window (Option-Command-N) and that will create a new top-level view. Note: What about that keyboard shortcut? It’s an adaptation of the shortcut to create a new Finder window, Command-N. Note that we’ve reserved Command-N to create a new document in DT Pro with the contents of the clipboard.

[2] From the menu choose Data > New > Group or use the shortcut Shift-Command-N, which is the same shortcut used in the Finder to create a new folder.

[3] The name field of the new group remains highlighted so that we can type in a name, let’s say “New Article Project”. Or we could open the Info panel to enter the name, by pressing Shift-Command-I, which approximates the similar Finder keyboard shortcut.

Temporary digression about consistency in keyboard shortcuts and other aspects of a user interface: We’ve received several user criticisms asking why we didn’t use Apple’s keyboard shortcut Command-I to open a document’s Info panel. The answer is that there is a limited number of convenient keyboard shortcuts and Apple recognizes this fact by using the same shortcut to do different things in different contexts. So Apple uses Command-I to open a file’s Info panel in the Finder. But in TextEdit Apple uses Command-I to format selected text in italics. Is Apple being inconsistent? I don’t think so. In order to make possible keyboard shortcuts, they assume that the user will recognize that the same shortcut may function differently in different contexrs. And that must be a necessary reality in all Mac applications.

Does OmniOutliner use keyboard shortcuts that are unique to that application, performing different operations than the same shortcut would in the Finder, or in Pages, or in iTunes, or in iMovie? Of course. Does that mean that the OmniOutliner (or iTunes, or DEVONthink Pro) UI is deficient, in the sense that the user must unlearn/relearn the meaning of keyboard shortcuts and/or menu commands in different contexts?

A further digression about dangerous preconceptions: I think the reality is that if a user approaches a new, rich application with the preconception that he has “grokked” the Mac UI and expects the same result for the same shortcut in all good Mac applications (including Apple’s), that preconception will get in the way of learning and using the new application, whether it’s DEVONthink Pro, OmniOutliner, or iTunes.

One of the preconceptions that I think is particularly damaging to a new user appreciating what can be done using DEVONthink Pro is the preconception that DEVONthink Pro is a Finder replacement. It is, to some extent, but that’s not what it’s about. It’s a file management system, but its real purpose is helping the user to manage and “mine” information from a collection of documents. Frankly, I hate it when a new user starts out by dumping the entire contents of a hard drive onto DT Pro, or focusses on mirroring the DT Pro organization and the Finder organization.

Instead, I approach DT Pro from the perspective that it can help me use my information, consisting of many documents bridging a variety of file types, by creating databases holding my topical interest materials. I’ve got enoirmous collections of files. Rather than dumping them all into a single database, which even on my fast machines would become relatively slow (and wouldn’t fit on my MacBook Pro’s 100 GB drive), I create different databases for different purposes. The result is that my databases are very quick and responsive. And on my computers there’s little if any correspondence between the files in the Finder and the files within my database. Most of my database contents don’t exist outside the database. (I back up my databases externally, of course.)

By creating databases that already hold contextually/conceptualy related material, I’m enhancing the abilities of DT Pro’s artificial intelligence features to help me analyze the information content. I make very heavy use of those AI features. My main database currently holds about 20,000 documents and total 24,000,000 words. As I also use it to temporarily hold new material more approptiate to one of my other databases, such as financial records, Apple Newton mail lists, scientific computing topics, etc. I will periodically export those materials to the appropriate databases and delete those exported contents from my main database. It takes only seconds to switch databases on the fly.

Temporary end of digressions and back to the How to:

[4] So now I’ve got the new group created and named. Now double-click it to open it in it’s own window, which I’ll set to Vertical Split view and will leave open for the duration of the project.

[5] Create a subgroup to hold the contents of the article (The Article), another subgroup to hold miscellaneous notes and comments during the project, another to hold duplicates of some of my existing documents that seem useful for reference purposes, and a final subgroup to hold new material added as possibly useful reference materials.

[6] Open The Article as a separate Vertical Split window. Create a new rich text document named “TOC”. Add a title and date to the new document. I start a writing project by using the List feature to lay out a topical outline of the sections of the article. The list feature allows me to indent or outdent the items in the list as needed. When I’m satisfied that I’ve got a grasp of a usable outline (section and sub-section headings) I’ll create a new document for each section and link (Control-click and choose the contextual menu option Link To) the relevant list element in the TOC to it. Now I’ve got the outline for the project and the associated documents to be “filled out” to complete the article. I will usually add a State check to the Info panel for each section document, so that I can quickly check which ones remain unfinished.

[7] I’ll then add to the holding group a new column next to Name that shows the word count of each section. Procedure: View > Columns >Word Count. Especially if I’ve got a limit for the total wordage of the article, this helps keep the relative “space” of each of the sections in balance. Makes a neat and informative writing environment!

[8] Next, I’ll start drafting a section. I’ve already got some ideas about the topic and I will lay those out. Next comes the use of the database to let me look for some ideas that might improve my grasp of the topic, and also serve as reference materials (footnotes or endnotes). I could do that by some searches, by keyword queries, prominent author names, etc. In that mode, I’ll examine some of the search results and duplicate the more promising ones over to my “Existing References” subgroup. Why duplicate? Because I want to feel free to mark up, delete portions or whatever without damaging my existing database items.

[9] Now I’ll start using some DT Pro tricks to look for new ideas and/or supporting references. I can simply press the See Also button at the bottom of the document pane and DT Pro will suggest a list of contextually similar items. I may not like some of them. But perhaps DT Pro will suggest one or more documents that give me new ideas I hadn’t thought about, perhaps even a better perspective on the topic. That happens often enough that I really love See Also. Or I may do a similar exploration by selecting a paragraph or two in my draft and control-clicking on the selection. DT Pro will then look for other documents possibly related to the content of the selected text. Again, often useful. Or I might want to look up other documents containing a selected phrase. With the phrase selected, press Command-/ to do a lookup in the Search window. Or look at other documents that contain a particular term. Select the term and press the Option key. A panel will slide out and display a list of documents that contain that term.

[10] Perhaps when considering my existing reference materials about the topic I find that something isn’t well covered, or I may want to look for more recent material. In that case I can do some searches in DEVONagent to look for more material. I may add that to my DT Pro database as downloaded HTML pages, and select some to put into my “New References” subgroup, or I may use Save As in DEVONagent to save the Digest summaries relevant to my topic.

Another temporary digression: I find DT Pro used in this way a very effective writing/development environment. I’ve got a reasonably simple and effective set of writing tools and, importantly, quick access to my research materials, with “research assistant” helpfulness by the AI features of DT Pro. The writing environment itself is basically Apple’s TextEdit UI environment, and the UI environment for access to my database contents by searches and AI assistance are strictly DEVONthink Pro, but are quickly learned and consistent.

I know of no other writing.research environment that so simply yet powerfully helps me extract information and new ideas from my reference materials in the course of writing. I call DT Pro my Memex machine and, it fact, it’s a step above Vannevar Bush’s original conception of the Memex – much faster, with a much better UI and much more helpful. (For those who don’t know what a Memex machine is, Google it or, better yet, do a DEVONagent search.) :slight_smile:

That’s why I spend almost all my working hours in this environment in DT Pro or in the complementary environment of DEVONagent.

[12] OK. I’ve finished my writing project. Each section of the article is finished and has probably gone through several drafts. Now I’ll tie them together into a single document by selecting View > Sort > Unsorted. If the “pieces” of the article are not in the proper order, I simply drag them into order. Then I select the segments of the article and select Data > Merge. This creates a new document with all the elements assembled.

[13] Now for the publication step. Unless you have designed the project as a Web site like the Tutorial database, DT Pro is not a good environment for final polishing and publication. Nowadays I usually use Pages 2.x for final polishing, because it lets me put everything together easily, add header and footer, footnotes and/or endnotes and set up bookmarks internally (e.g. from the TOC for a long article, and to endnote references), and because it does a good job of exporting the project as a Word document and a magnificent job of exporting as PDF with working internal hyperlinks. The best approach is to copy to the clipboard from DT Pro and paste into Pages 2.x, as images come over very well that way and require little if any tweaking. Done!

More comments on UI: I know of no UI that, from a rigid point of view, can be considered totally consistent. And that’s probably good. Certainly Apple has experimented with a number of UI approaches and continues to do so. There are some Mac operations that are almost universally consistent, such as the Command-S Save for writing or image manipulation environments, Command-A to Select All, and others.

DEVONthink Pro is a pretty rich environment. In addition to the writing/reseach environment I described above, it has features for adding material from the Web or from the Finder, and there are some differences between the capture options. DEVONthink Pro is one of the most “automatable” of Mac operations, allowing automated operations via scripts and Automator workflows – still another environment.

DEVONthink Pro often provides several ways to accomplish a task. One can use the menu bar or a keyboard shortcut to import a file from the Finder, or drag & drop the file to the Dock icon, or drag & drop the file onto the floating Groups icon, or use a script to capture a file. This is common in many Mac applications. It isn’t meant to confuse, but to provide flexibility to accommodate various user preferences.

So we get a variety of comments about the UI, especially about how it looks to a brand new user. Some commend it as being elegantly simple. Some think it’s ugly (maybe the guys who designed NoteTaker?). Some think it’s complicated, with too many choices. Some think it’s unfinished – and I think we would agree with that, as improvements are always being considered. But all in all the UI does let one get at the real strengths of DT Pro, some of which are quite remarkable for a “big” but consumer-level document database that can be used by mere mortals.

What I really want from DT Pro is not just a document database but an information manager, which is quite a big step up from a document manager. I want to slice and dice and compare ideas, and perhaps come up with one of my own, once in a while – perhaps with a little help.

I do feel like I’m in that information manager environment when I dive into DT Pro and create my personal environment for writing an article. I can initiate a search of my references from within my writing space. The See Also button is at the right bottom of a document window or pane. Should it be an icon instead, or placed in the toolbar? Is the slide-out panel of suggestions the best approach? Could there be other navigation aids when I’m following a trail of suggested documents, perhaps triggering searches or See Also actions from them as well? What kind of UI could best help me jump from looking at a document, to getting information, understanding, perhaps inspiration? To me that’s the important question, because that’s why I created my database, and the reason I use it.

Admittedly, some of my other databases are more mundane, such as the one for my financial records. All I need for that database is the capability of very quickly finding anything I’m looking for, with some reasonable organizational structure. How much return did I get on an investment account last quarter? I scan/OCR stuff like that into DT Pro and the search for that information will take about 20 milliseconds to find it – as opposed to looking in a file cabinet. And DT Pro will find it anyway, even it I “filed it” wrongly, i.e., in the wrong group.

Hi, Father Moses:

I’m in agreement with you. I keep the Info panel open, using the full vertical space of the window but narrower than my view and document windows, on the right side of the screen. I find that convenient because I can instantly see or change state, label or lock status for a document. I can launch a Web page or document from the URL or Path field. I can set or kill a script for a group. There’s a fair amount of information available at a glance, from file type and size to location or the number of replicants. I leave a fair amount of space for the Comment field and the panel will scroll if there’s more in the Comment field than is immediately visible.

So I make quite a lot of use of the Info panel, all the time.

But if others want to see only a portion of the Info panel so that it’s collapsed like the one in the Finder. OK as a Preferences setting. I’ll mention that to Eric.

For the time being, place the Info button in the toolbar. Then the Info panel is available at a click, and can be closed or sent to the Dock with another click.

Speaking of UIs, that Finder Info panel bugs me, as it seems I always have to unfold whatever I need. Curse you, Apple! That’s bad UI. The Info panel can’t be expanded horizontally enough to display all the information within the vertical height of my MacBook Pro screen. Why was the horizontal width of the panel arbitrarily constrained so that I can’t expand it another inch, which would let me see all the information I very often need to see about a file, at a glance?

Example: Sometimes I want to set the application to open a file, and also deny access to it by others. That means I have to take time to click on triangles to get to those options, while closing other options I will probably want to see next time I open Info. Irritating.

I do all my drafting/writing in Mori, though I’d prefer to do everything in DTP. It’s not that Mori looks better (it does, but this is not the point). Can I quickly “indent” a note (means: note in notes), can I reorder items via keyboard? I can’t even sort the 3-pane view the way I want to.
DTP is a killer-app for storing information, but not yet for all those little lists and drafts and tasks I have to work on daily.

Hi, Jonas:

You may well be more productive in an environment in which you are comfortable.

In answer to your questions:

Yes, one can indent a note. I’ve got a little document in which I’ve made a variety of paragraph indents saved. I can copy such a "style’ in DT Pro’s rich text mode by selecting a paragraph and pressing Command-1. Then I can paste that indent level by pressing Command-2. The indent remains in memory until DT Pro is quit. As I rarely indent more than one step I’m usually satisfied. A second level of indent means either manually changing the tabs for the paragraph or another trip back to my “paragraph styles” document to copy a different indent.

But for “outline” approximation I use lists. Can I rearrange them by keyboard? No. Takes drag & drop to rearrange. Is it Mori? No.

Can I sort in 3-panes view the way I want to? No. :slight_smile:

That’s why I work most of the time in the Vertical Split view. I like the text pane better in that view, anyway. And of course the Unsorted (manual sort) option is available in Vertical Split.

Why do I write in DT Pro? Because all my reference resources and some useful artificial intelligence features are available to me in that environment. Although I’ve been using it for years, it still feels awesome.

Thanks to a tip from Alexandria, I’ve installed a little utility called Afloat (Google it). Afloat lets me designate one of my DT Pro document windows as floating above all applications. I minimize it to the Dock until it’s needed.

It’s my document that I use as a daily journal and general note-taker (usually in a monthly series), so now it’s not only always available, but as it’s part of my database I can easily find my notes. :slight_smile:

Or select a note about any topic and see if I’ve got related material in my big database. That’s when DT Pro starts getting interesting to me, as it often leads me to write more notes about the topic. Maybe it will turn into an article or a project.

I never do any ‘real’ writing in DT.

I do a lot of work in it (largely info mgt, and now data manipulation), but despite the fullscreen mode it just doesn’t feel usable (unlike Scrivener, which I really like, or Mellel, which I am trying thanks to this forum).

A lot of my ‘real writing’ is either technical with lots of footnotes and in specific formats (usually OpenOffice, given I’m a UNIXer and so is my audience), or code.

There is no way that DT is going to become a full-fledged word processor or programming editor. It’s silly, these guys should stick to what they are good at and not chase user whims as far as applications go (every user has their favorite; it is almost self-defeating to please all).

What I would like to see instead is hooks in DT to allow viewing and working with files using external applictions.
I don’t mean ‘open with external application’, but actually having the application inside of a DT view window – kind of like Adobe Acrobat does on (some) web browsers.

This is quite common in programming IDEs, which makes me believe it is not impossible… though as a programmer who hates GUI work, I have no idea how to go about it :slight_smile:

Anyways, that is my ‘dream use’ of DT: an information manager sitting between the applications and the OS, which allows the use of arbitrary (but well-behaved) applications as ‘slave’ viewing and editing tools but indexes and controls the data obsessively. Hmm, actually, given that anything interesting (.doc, .sxw, source code) is linked or saved as files anyways, it is not hard to feed data to an external app.


PS Back on topic, I like the info window as-is: seeing everything at a glance is important to me. Otherwise I would use finder and spotlight :wink:

Does anybody here remember OpenDoc? :slight_smile:

If you consider DEVON{Note,Think} primarily as “editor” apps it makes sense that command-I be a shortcut for Format > Style > Italic. And if you categorize Finder as “viewer” app, which doesn’t even have a Format menu, command-I is a reasonable shortcut for File > Get Info.

I used similar “editor” vs. “viewer” app logic when someone on the Journler forum recently suggested command-T to be the File > New Tab shortcut instead of Format > Font > Show Fonts.

When a Mac app has a Format menu it always seems that shortcuts for commands under it have higher precedence than they would as shortcuts for commands under other menus. No exception immediately comes to mind; maybe someone else can think of one? Until then I’ll likely keep disagreeing with any criticisms that think its okay for apps to steal shortcuts from the Format menu for other purposes. It’s another matter if some people use methods to personally override any default shortcuts they dislike. I don’t care if they do but won’t recommend; that seems like an extra layer of unlearning/relearning. Other shortcut-related issues, like handling service menu shortcut conflicts, are worth more of my attention than fussing with ones I’ve already learned that work just fine.

That, and Newton soups, are a couple Apple technologies I’m curious about enough to, in hindsight, wish I had some direct experience with. I didn’t take much interest in Apple before OS X and until I bought my first Mac in 2001.

I remember CyberDog, an OpenDoc suite that pulled together web browsing, address book, email and news readers, plus file transfer.

I used it to organize clusters of data, usually for writing projects. Apple supported it for a year, 1996-97, but you could run it on Systems 7 through 9.

Probably what killed it was Microsoft’s investment in Apple. At that point, the Mac OS began to install another browser dog, Internet Explorer. What a woof.

Bill: it’s interesting to note what you “hate”, but I think there is a valid reason that many people come to view DT (partly) as a Finder replacement – namely, they want DT’s AI abilities to help them find their files, which is usually what non-DT people use Spotlight+Finder for.

And while I can understand your need for various databases, let me give you an alternative use of DT, in which DT almost MUST become a Finder replacement.

My life is one of an academic, but my interests and research vary widely. My “formal” training is history and area studies, but I’ve spent many years studying and teaching the liberal arts, such as logic, rhetoric, grammar. I am also interested in world politics and news (who isn’t these days?), “popular” scientific research, social commentary, etc.

Now, someone might make a different database for each of these: one for the “formal” training, one for my files/notes in the liberal arts, one for politics/news, and one for essays and articles of social commentary, etc…

The problem is, I think this completely defeats the whole beauty of DT’s AI abilities. How? Because I want the AI’s “Memex” features to be an extension of my own (weak) memory, to help my mind in the synthesization of information/knowledge. And the thing is, my mind – and my research – is not broken up into neat compartmentlized fields.

Currently I am doing research into medieval educational history and theory, in a particular part of the world, in a particular time. This is my “history” research. However, I might have an article/note/website in my DT database that is about “rhetorical semantics”, that is from a modern theoretical PoV, but is relevant to my historical research. Likweise, a different entry might be a news article culled from BBCnews or the like, discussing applications of some other linguistics theory (that is relevant) in East Asia – this would have been in the “news” database, if they were all separate.

I want my brain to synthesize. I want my DT memex to help me do that. And, to be honest, this requires that I do use my DT database as a finder replacement, to a large extent, because almost every file I personally have in my documents folder is something relevant to my personal cultivation of knowledge, be it academic research, keeping up with big movements in the news, or otherwise. To me, they are all integrated.

So, yes, I do need to see all those .doc files, mellel files, .rtf’s, etc. etc. in my DT database. And, for someone who uses DT as I do, it would be a Godsend if DT did actually become a proper finder replacement, but one with all the AI goodness, all the manipulation of text abilites, all the services and applescript abilities, all the integration with DA, that it has.

That’s the long version. The short: I want my DEVONthink to SEE everything I have on my computer (for now, text or metadata’d non-text files), help me organize it, and remind me of it using it’s AI when need be. And this is best done when everything is seen as one complete, coherent whole “database” (for lack of better term), which cannot be kept divided into smaller databases, as the AI features would be completely rendered null thereby.

This transmission is officially finished. Hope it made sense. :slight_smile:

I love the idea of OpenDoc but it never worked well for me. It crashed often and rebooting it was a long wait on my older machine. This is the main objection I have with a lot of Apple’s latest and greatest technologies – they often require the latest and greatest computers. It seems the minute after I get the new computer up and running, it already has difficulty handling some of Apple’s software. Two recent cases in point:

  • iLife and iWork. Not sold for my 3-yr old iBook. Bummer!

  • high definition video. Not only an Apple problem but also a web-design problem. Go to and you have to wait for an advertising video to download. Bad design. On my machine the wait is more annoying than long but its video card can’t handle playback of Apple’s own videos (most recently the Leopard preview videos) and high definition movie trailers – even after downloading the playback sputters stutters and stops to breathe. So now I avoid the Apple website and no longer bother previewing films in QuickTime. Unfortunately, many websites rely on advertising banners with video. Too bad text-only browsers don’t work well on the modern web. I just spent a couple weeks with OmniWeb – the only browser I’ve found that lets you individually turn off graphics features such as animation, flash, banners, etc. – but found it lacking enough not to spring for the license.

I am wary of most data solutions that should “transform society” such as OpenDoc and Web 2.0 and the many incarnations of xml (how many of them will be useful in ten years?) I keep any data that really matters to me in ASCII files that I back up fanatically.

Please excuse the rant.

Hi, talazem. Well put, and I’m very largely in agreement with you.

My main database reflects my long-term professional interests in science and environmental policy and includes content from a wide variety of disciplines related to those topics from “hard” sciences and engineering to agriculture and forestry, conservation ecology, sociology, political science, archaeology, anthropology, law and economics. For many years I’ve worked with environmental science exchange programs with developing countries.

An old friend, mentor and colleague died last month, Lynton K. Caldwell. Many years ago I coauthored with Keith Caldwell a series of graduate seminar syllabi called Advanced Studies in Science, Technology and Public Policy. Those syllabi, like my DT Pro database, covered topics ranging from philosophy, history, agriculture, energy, communications, public administration, and ethics to many others. Keith is widely credited with being a principal thinker behind the U.S. National Environmental Policy Act, and with “jump-starting” academic programs such as the School of Public and Environmental Affairs at Indiana University.

So my database does cover a wide range of materials, all of which really are related to my interests, and it’s a joy when it helps me recognize and synthesize relationships among seemingly disparate disciplines.

But I don’t put everything into that database, for two reasons. First, at about 20,000 documents my database remains quick and agile on my computers. Given the current state of the art for consumer computers one has to accept limits in resources such as memory and processor speed (especially memory). Second, adding “too much” information about particular matters can lessen the value of DT Pro’s artificial intelligence assistance to the user – and that’s an important consideration.

Here’s a good example. There is a very large literature covering the fields of environmental sampling and analysis and the related issues of evaluation of environmental data. There are thousands of analytical procedures for chemical analysis, for example. I’ve been building a very large database of this kind of reference material, principally from U.S. EPA and other governmental agencies and from the European Union. This database also contains selected case histories of environmental problems and cleanups, including some Superfund sites that I supervised during investigation and remediation. Those case histories are included as illustrating particular problems in choosing investigative, sampling, analytical and evaluative techniques.

If I put that big collection of references into my main database it would result in a much larger and slower-operating database. Of course, that’s a mere computer resource problem that can be corrected in the future by more efficient software (e.g., DT Pro 2.0) along with more powerful computer hardware.

But that’s not the main reason I keep that technical database separate. If I were to merge the databases the detailed technical materials would “swamp” DT Pro’s search ranking and AI features about some topics. Example: my main database contains a number of items about arsenic in drinking water, related policy and regulatory issues, case histories, toxicology references and the like. I can find useful material quickly, and “See Also” really works well in suggesting potentially related material. But if I were to merge the technical collection about sampling/analysis, etc. I would have increased the frequency of the term “arsenic” by orders of magnitude, and this would tilt DT Pro’s search rankings and “See Also” lists heavily toward that technical literature, which really has a limited scope among my usual interests. My usual interests about arsenic in drinking water will be information about areas that have potential problems, issues related to standard-setting such as health effects, available technologies for arsenic removal and costs (especially for small communities or developing nations) and the current status of related regulations and standards. DT Pro does a wonderful job in helping me satisfy those interests now. But it would be less effective if I merged in the technical sampling/analysis database. I’ve got some overview articles and books about sampling, analysis and data evaluation in my main database, but not all those minute details.

I was originally trained as a chemist, so I know which database to use for selecting an appropriate sampling and analytical methodology, with associated statistical and other considerations for data evaluation.

So that’s why I “hate” treating DT Pro simply as a Finder alternative for collecting information.

Mixing a broad-ranging database with a large but topically more limited database can reduce the effectiveness of both collections.

Of course, it’s possible to do some filtering of search results, and future software will be more effective in doing that. But that requires more thought and effort on the user’s part and more work on the computer’s part.

So I keep my financial and tax records in a separate database. When I’m interested in those matters, that database is very useful. But they don’t belong in my main database (except transitorily) nor in my analytical methods database. The same goes for my large collection of news lists and other references about the Apple Newton. I’m interested in the history of science and ideas, which is relevant to my main database, and when I look up something about Newton I don’t want to see thousands of search results for the Apple Newton. And when I’m looking for information about a Bluetooth card for the Newton, I’m not searching for information about Isaac, the laws of motion, calculus or alchemy.

Bob – a most thought-provoking response, as usual.

I still want DT to replace my finder for reasons mentioned above. But I guess I’ll have to wait for the hardware and the software (i.e. DEVONtechnologies’) to catch up to the human mind. :wink: I guess I shouldn’t hold my breath.

If you wouldn’t mind, it might be illuminating if you’d give is a break-down of your own databases; i.e. how many, what they are, and – if it’s not too much of an imposition – WHY you’ve divided them as such.

I personally have two – research, and personal. But I’m being dishonest; the personal includes what it sounds like it would, in addition to a large (perhaps even predominant) collection “practical” computer and software tips (like this type of post) organized by software or hardware groups. Research includes everything else – articles, pdf’s, web-clippings (news and quasi-academic), reading notes, class notes… all organized largely by topic (though formal PDF’s and books by author).

I do want to keep my notes (reading and class) in the same database as related PDF’s and web-clippings for my research/writing purposes (that IS the whole reason I want to work with DT). But – I submit – I do see what you are saying as far as “swamp”.

What do you do, and what do you suggest? (We are also open to anyone else’s experiences).

Bob? :slight_smile:

[never mind; Bob^H^H^HBill explained]

Hi, talazem: In some ways the software exceeds the capability of the human mind, because it can “look at” tens of thousands of documents very quickly indeed. In other ways, it remains well behind human capabilities, as it lack “understanding” of the material in ways that humans understand material, especially with training and experience.

For that matter, my own human mind lacks capabilities that I wish I had. I tend to freeze up when presented a great deal of information all at once. I’m happier when either the amount of information is small enough for me to digest, or I can find a rubric – some synthesizing or relating approach – that lets me pull together details into a simple, comprehensible analysis.

At its best, DT Pro becomes an interactive tool for sifting through large quantities of data to help the human team member get information and, perhaps, understanding.

I do keep my own notes and writings in my main database along with my reference material. That adds value, rather like the Memex. But I lack the time and ability to “swamp” the database; there’s no way I can write that much. :slight_smile:

I’ve already described several of my databases.

My main database started as a single database using the original DEVONthink, more than 4 years ago. When I started testing early alphas and betas of DT Pro, which allows multiple databases, I began to take advantage of that capability.

As previously mentioned, I’ve got a financial database. That covers data from checking, savings and investment accounts over several years. It also contains tax information including notes, spreadsheets, scans of federal and state returns and the PDF manuals for each tax year.

I’ve got a database for Apple Newton material, which is very large as it contains lists to which I’ve subscribed for years and which are voluminous.

I’ve got a large database created to help analyze the impacts of hurricanes Katrina and Rita on the health services infrastructure in Louisiana. This was a volunteer project. I pulled down more than 10,000 pages from the Web using DEVONagent, dumped that material into a new DT Pro database and filtered and organized it to about 5,000 documents. Several months later searches were run again and added to the database. This information dramatically illustrated the severity of the impacts on health services, not just in the areas directly affected by the hurricanes but statewide. Many of the Web pages of important facilities have disappeared entirely (as have those facilities) or have not been updated since before the hurricanes.

I’ve got another big database for a volunteer project to improve policy and procedure documentation for a new health clinic. The content has been added by scanning more than a thousand pages of paper sources plus downloads of Web resources. (I’ve used a beta version of a future enhanced version of DEVONthink to do one-button scan/OCR/save to database, and it works very well.) The database includes model manuals plus federal, state and local regulations that must be met. Several health care professionals are involved in this project and I’m the drudge doing the computer work, including final assembly of the documents in Pages.

I’ve got another database on the Panorama database application, which I’m not using currently but have used years ago, and may have a use for in a future project.

I’ve started a database on scientific computing and will be peeling out some material that currently resides in my main database.

For years, I’ve left mail messages accumulating in Entourage. But I now don’t use Entourage very much, so I’ve created a new DT Pro database that contains almost 20,000 email messages. Now I’ll be able to find things much more easily. I’m using beta software that will be released in an enhanced version of DT Pro in the future. It works! And it handles attachments!

I often create test databases to try out various approaches or test out some tricks. Some of those survive, most don’t.

In general, whenever experience with searches and AI features demonstrates that some of the material in my main database really doesn’t “fit” and may be causing problems, I consider spinning out that material into a separate database.

But one of the types of content that will stay in my main database is a daily journal that I use for miscellaneous notes. I store those in a group that’s exempt from Classification.

Note that these databases have very little commonality, and merging them into a single database would make each of them less effective.

Note also that surprisingly little of the content of these databases exists in the Finder or, for that matter, ever existed as Finder files outside my database packages. The exceptions are Word files, a few very large PDF files, a few very large image files and some QuickTime media. Everything else is contained in the database package file. Obviously, material captured from the Web or by scanning goes directly into the database, including DEVONagent search results. Because DT Pro searches are so much faster and more focussed than Spotlight searches, I only use Spotlight occasionally to check for material that I might want to include in a new or existing database.