StickyBrain 3

StickyBrain 3 has a new database structure. The data is now located in


If one does a show package contents on that there is a folder called Blobs that has the data. There is a folder structure, but the files open with TextEdit.

When I drag the blobs into DevonThink it represents these files as URLs. How do I get DevonThink to index the contents?

I REALLY don’t think you should be pointing DevonThink at an OpenBase database file!

If you want the info to be indexed… drag the original documents into DevonThink (as well as StickyBrain).

But StickyBrain does a very good job of indexing its own documents as well… if you have text/notes in it, use it. Or use DevonThink. Having them point at each other… I predict something bad will happen eventually if you try and do what you’re trying to do.

I’m in full agreement with CatOne. Don’t try to import or index your StickyBrain database into DEVONthink.

StickyBrain 3 is a nice little app and I’ve got it. (Just because I had StickyBrain before I found DEVONthink. I quickly found that DT does much more for me.)

DEVONthink is MUCH more powerful than StickyBrain. I recommend DT as your primary document database. If you want to transfer information from StickyBrain to DT, do the transfers as CatOne recommended.

The only advantage that StickyBrain has for Palm users is that it can transfer files to your Palm. For that use, transfer files from DT to StickyBrain, and then to your Palm.

Where SB is the biggest help to me is HOW it grabs data. I can highlight text, use a contextual menu and CHOOSE the destination folder. Can I do this in DT? It is such a time-saver and really suits the way I work.

I teach and gather info from lots of sources, especially when creating a syllabus, so I am not using DT in a more “orthodox” research kind of environment.

Any help is really appreciated. Feeling a little dense about this.

FWIW – This is most of what is keeping me from purchasing DT.

I’d like to second this. Rather than upgrade to SB3 I chose to purchase Dt which I’m extremely happy with. But even with SB2 you could define keys to capture data to a specific category. Also, SB did not need to be open nor did it launch when those keys were invoked.
Similar functionality in Dt would be welcome.


Thanks …

I’ve found that a lot of what I want to put into DT comes along “accidentally,” while I’m in my regular browser (Camino). A simple, reliable way to capture the URL, page, or a section of highlighted text into a particular DT folder sure would make a difference.

I don’t think I want DT to function like a desktop/application replacement. Maybe this makes my use “non-standard.” Again, my work is not hardcore research.

I don’t think that I do “orthodox” research, either. My interests span science and environmental policy, public administration, international science exchanges and development issues, history and philosophy of science, risk assessment/risk management, and comparative studies of environmental regulatory schemes. I track interesting scientific and technology developments in fields ranging from analytical chemistry to genomics, proteomics, economics and epidemiology, with a bit of computer science and statistics tossed into the mix.

At any given time I may be doing a critique of the literature on sustainability, recommending pollution control and/or pollution prevention approaches for an industry in a developing nation, or looking at investigative/remediative approaches for a groundwater contamination problem.

The number of groups and subgroups in my DT database is approaching 800, because of the diversity of topics I’m following. I track a number of online journals and toss items of interest into my database. I track the EPA and European Union Web sites on environmental issues, science and technology, and also put items of interest into my database. I follow World Bank development projects and find still more items for my database.

Here’s a terrible secret about my DEVONthink database. I cover so many disparate topics, and dump so many items into it every day, that I’ve almost given up trying to keep the content neatly organized into groups/subgroups. I keep a catchall group named “Edit these items” for content that I’ve not placed in my organizational scheme. As of this moment, there are 2,247 documents in “Edit these items.” Yes, I spend a few hours a week whittling away at that group, but it still keeps growing!

What’s wonderful about DEVONthink is that it still gives me quick access to items related to any topic, even if I’ve not “organized” the database structure neatly. DEVONthink has contextual recognition features that aren’t available in StickyBrain. While I’m looking at a document, I can press the “See Also” button and view a list of related items. Of course, I have to exercise some judgment (and know something about the topic) to really use those suggestions wisely. But DEVONthink has, for example, pointed me to a statistical approach for improving confidence in data while I was looking at a paper about the difficulties of evaluating stream water quality databases. That suggested paper was a computer science paper dealing with similar problems of achieving significantly higher confidence levels from limited historical data sets – and seems very promising.

DEVONthink offers still more “knowledge mining” tools, including Keywords, Context, and fuzzy searches. DEVONthink Pro beta offers a “smart folder” script (this can already be emulated in DT PE 1.9.1). A combination of multi-level searches and sorts and Replicate To with the smart folder approach can do fantastic manipulation of your data.

While viewing a document, there’s a “Classify” button. Press that, and DEVONthink will suggest an organizational location for the document. If you are working with a fairly well organized structure and the new additions fit that structure pretty well, “Classify” works pretty well. Suppose a new document fits in two or more groups in your structure. In that case, DEVONthink will replicate the item into multiple groups.

For that matter, if your topics are well organized and you are adding new items consistent with that topical structure, just turn on Edit >Auto Classify and let DEVONthink automatically place additions to your database (of course, this requires that you have already populated your groups with some items for DEVONthink to examine). That’s much easier than the manual effort of telling StickyBrain where to place new items. If DEVONthink can’t figure out where to place something, it will leave it at the root level for you to figure out later – perhaps starting out by looking at DEVONthink’s suggestions when the Classify button is pressed – or perhaps requiring addition of a new group to better fit a topic.

And, of course, there’s the manual approach (not unlike StickyBrain’s). Dump new material into DEVONthink (you can set the group to which imports are sent, in Preferences > Import). Then select an item name, and use the contextual menu option “Move To” to select the desired location. (You also have a similar option to Replicate To, if you wish to place the item in more than one group.)

Truth to tell, I view and use my DEVONthink database more as an information network than as a hierarchical collection of data. My data collection is approaching 14,000,000 words and is encyclopedic in size and scope. This database was started more than two years ago, and the organizational structure wasn’t all that well thougt out or improved in progress. Fortunately, DEVONthink has been “smart” enough to overcome my organizational deficiencies. :slight_smile:

I know that people have preferences for various browsers, for various reasons.

But the most convenient browser for capturing an HTML page (source code), or selected text of imagers, is the WebKit browser built into DT PE 1.9.1. It offers two contextual menu options: Capture Page, or Capture Note (selected text and/or images). The URL for the captured item is automatically pasted into the Info pane of the captured item in DT. The WebKit browser loads pages very rapidly on my broadband connection.

The next most convenient approach is to use the Services options of Safari or OmniWeb 5.x. Services lets one capture selected text/images as plain text or rich text, and the URL is automatically added to the new DT item as above. There’s also an “append” Services option that’s useful if one is capturing a multi-page item. (I’m not certain about this, but I think DEVONtechnologies is planning to do automatic URL captures for the Camino browser, using Services capture.)

I’ve created a “Bookmarks” group in my DT database to hold the URLs of a great many Web sites that I routinely visit. To do this, create such a group. Then copy the URL of a site to the clipboard, go to DT and use Content > New :> With clipboard. The new item will have the name of the URL, which can be changed if you wish. For example, I changed the name of such an item to “Science Magazine WWW.” My Bookmarks group contains subgroups like “Scientific Journals,” “News Sites,” “Mac Sites,” and so on.

There are minor deficiencies in the WebKit browser. There are some Java codes it doesn’t handle, and one cannot do file downloads using it. If, for example, I want to download a file that’s linked to a viewed page, I click on the URL field and the page opens in my default browser, where I can do the file download. But the contextual Page/Note capture options are so convenient that I spend most of my browsing time using WebKit.

DEVONthink Pro beta also offers a “site sucker” feature to download a Web site, including images, for offline archiving and viewing. (I use this feature rarely, if ever.)

You and others also mentioned a desire to be able to place new content into a specific group at the time of capture. This request has been made a number of times. Yes, it could be done with a contextual menu plugin, but the DEVONtechnologies programmers are quite dubious about the safety and value of such plugins, as they frequently cause problems. I did have problems with the CM plugin of an older version of StickyBrain. (And with almost 800 groups and subgroups, it would drive me nuts if I had to manually select a destination for each added item.)

As I noted in a previous post in this thread, DEVONthink offers something that’s even more convenient than StickyBrain’s destination selection: Auto-Classification of new entries. One step below that in automation is the “Classify” button, which triggers DEVONthink to suggest one or more locations for a new item (after capture). DEVONthink also has a Preferences > Import option to designate the destination of new items. Finally, DEVONthink allows manual placement of already captured content either by a contextual menu option, or by dragging and dropping items.

DEVONthink offers still another automatic feature, the Content > Group option. If one has a number of ungrouped files, select them and use the Group option. DEVONthink will examine the files and look for contextual relationships that the application uses to group them into newly created groups. (Your satisfaction with the result may vary, depending on how much you think like your computer. The result is not hierarchical, and you will probably want to rename the groups and arrange them hierarchically.)

DEVONthink also permits batch imports, either by importing entire folders of files, or by receiving batches of HTML pages from a DEVONagent search. I’ve imported hundreds of files at a time. In such cases, the automatic or semi-automatic destination options in DEVONthink are convenient.


I have been using DT and DT Pro for some time now and I feel that I have a pretty good grasp of how to put it to use for me. Having said that, I always learn something new or a different way to think about using DT after reading your posts. I really appreciate your contributions to this forum.

I believe it is unfortunate that AquaMinds does not have a forum as I am sure that I could also learn from you how to use NoteTaker more effectively also! Keep on posting. :smiley:


Bill…great point about hierarchical structure and DT. I’ve got one setup, but honestly I’m not exactly sure what it’s going to do for me long term. I think it’s more for satisfying the anal, organization side of my brain than anything because I can have everything dumped into one huge folder and still find what I need virtually instantly.

So I guess my response to the “let me specify where it’s going to go” request is that it’s hard for me to see the location of documents in DT making much of an impact on their usage. Maybe someone has a radically different usage pattern with this than me that makes the folder structure meaningful, but i’m not seeing this as a very important aspect of this app, and that’s a testament to its other features.

I have a Stickybrain Database with about 11,000 notes. Although Sticky Brain 3 is profoundly improved, the search capability leaves something to be desired and I’d love to be able to get at this data with Devon Think. Sticky Brain now uses the OpenBase engine. Is there any simple way to import this data?

Heh … I must be that person. I teach university freshmen in several classes that are readings based. These classes often have multiple subjects or themes. I suppose you could think of them as courses with an anthology of subjects. As a result, when I find something – for example while surfing in my browser – and want to add it to DT, I am the only one who can know where it best fits. Auto-classify will generally not help me because: a) it does not see the relationship between, say, “infertility” and my writing class, and/or b) there is not enough data in the current writing class for Auto-classify to pick something out and say – in effect – “yes, this matches what is in THAT folder. I think I’ll put it there.” Note: I do not know when Auto-classify would have enough data in the folder. My pattern for collecting information is too short-lived. Finally, the suggestion that I capture data and put it in a “to be filed later folder,” is ‘okay’ but I really prefer doing this all at once: find information, grab it, label it, put it in the right pile. Move on. Ponder the info later. I think of this process the way I did when I worked with notecards. I must be old!!

So I do come away feeling that DT serves a more hard-core research type application AND is adaptable to those who like strict hierarchies or loose ones. I just haven’t been able to make it work for me … and this is a shame because I don’t really care for the alternatives.


frankns…but why does it matter what folder it’s in?

i’ll take my usage pattern to the extreme. I do have folders and subfolders, but let’s say i put EVERYTHING into one folder. I could still find whatever i needed just as quickly because I don’t use the folder structure to retrieve stuff. I use the search functionality to do that.

Yes … I see your point. Here’s how it works.

Any one class I teach will have much unrelated content AND a need to teach both content and writing skills, so I am constantly looking for both in everything I read. When I find something I think: “Now that’s a good example of paragraphing.” or “This is a good reference on infertility treatments.” One is writing, the other content. (I might find a comment in this discussion, for example, and say … “That’s really well done. I want to show this to my Professional Writing students.”)

At this point I want to grab the text, link or whatever and put it into a CONTEXT. The context is critical because it helps me remember why I grabbed that particular item later. Remember that I come across these things “accidentally,” while focused on something else that is more important at the moment.

Later, because I might not have time to come back to the item until days later, I need something to help jog my memory. The context – a folder labeled “Class A” or “Paragraphing examples” – serves this purpose amd makes all the difference. It brings my mind back to where it was when I first saw the piece. This is especially true in planning for classes before a term begins, when I will grab all kinds of things as “ideas” or possibilities," wanting to sort them out later. If I cannot put them in a context right away – meta-data would serve the same purpose – I end up reading the material through again later and puzzling out why I grabbed it in the first place. As a result, I end up spending MORE and not less time.

Goofy as this is, it works extremely well … and is why I haven’t figured out how to make DT work for me. I REALLY want to …

frankns…i think i see your issue. doing a straight search might not work because what you grabbed isn’t necessarily content driven, so what would you search on?

one suggestion that might be of help (provided I’m understanding your issue properly) is something I use in a similar situation. I’ve got a few items in the db where I didn’t feel confident in finding it again searching merely on content, or felt i needed a more overt descriptor to help me locate it. What I did was use the comments field for the item. I selected the item, clicked on ‘show info’, then put one or two keywords in the comment field that i knew i would associate with this item and now i can search for the item by those keywords.

perhaps you could develop a classification/category system to use in this way. You could instantly retrieve any doc that had those keywords without having to remember which folder you stashed it in.

just an idea.

Yes, you’ve understood perfectly. This “scenario” covers most of my usage, so it’s non-trivial for me. And I’m usually reluctant to invest lots of time in adapting to the software’s way of doing things. (Not that the software isn’t “right.”)

BUT – hooray – your solution works fine! I just hadn’t thought of using the comments field for meta-data. Doh …

It IS a little cumbersome. Reminds me of things we were doing years ago. Wouldn’t it be neat to be able to grab and apply meta-data at the same time! Or simply put it in a folder :wink: But I know from looking at other posts that the developers don’t agree.

Maybe grab and classify at the same time?

Thanks for the suggestion!

Maybe I’m not appreciating the full subtleties of what you are doing but why not just add your contextual text to the item itself.

You could add an entry like

CONTEXT: paragraphing example

to the beginning of the item and even change the text, or text colour or use the highlight colour to distinguish it from the original article. Then you could put all your unsorted articles in one folder and use search to find the particular contexts you are interest. You could also of course have multiple contexts per item. Or am I missing the point?

Can you do this with a web page? I can see doing it when I grab selected text, and it’s a great idea.

BTW - anyone know why sometimes the shortcut keys (to the services menu) don’t seem to work? Sorry bad form … two question and wrong thread.

Can you be a bit more specific? How are you bring the webpage into DT? If you’re bring it in as a PDF then unfortunately your only choice is the Comments.

You can leave the Info window open and it will stay on screen while you work with DT. Then when you add a new item to DT all you have to do is click in the comments box and add your comments. As you browse through the database it will be updated to show he details for the current item.

If you are adding the files as any form of RTF or text format then my suggestion of adding the information directly into the file will work.

When you say

what exactly do you mean. What are you adding and how are you adding it to DT?

Hmmm, I wasn’t specific enough, but I see that you have answered my question. If I capture the URL in DT, I will need to use the comment field to store my meta-data. If I grab an excerpt as RTF or TXT, of course I can simply add meta-data to the text.

Thank you.