Devonthink – Papers: 2 Programme für den Mac, um wissenschaftlich zu arbeiten

Gefunden habe ich diese Rezension und Besprechung der Software Devonthing in: Counselling Ressource

Da ich selbst begeistert Devonthink benutze, empfehle ich jedem Wissensarbeiter, diese Besprechung für den praktischen Einsatz der zur Zeit schon etwas überholte Version (V. 1.9) von Devonthing trotz allem zu lesen. Die Version 2.xx wird noch einiges mehr bringen wie hier schon angekündigt ist und im Firmen-Blog des deutschen Geschäftsführers über die Software (zur Zeit als Betaversion in öffentlicher Betaphase erhältlich) nachzulesen ist…

Ich nutze sie für die Verwaltung  meines gesammelten Wissens, das ich vor allem in digitalisierter Form aufbewahre und wiederfinde! Dank intelligenter Suchmöglichkeiten.

[Note: This review has now been superseded by our review of DEVONthink 2.0.] Information management software DEVONthink has a serious problem: it is so uncannily capable that many of its users don’t want anyone else to know about it. Content to let their colleagues or competitors think of it as just some ‘database thing’, they don’t want to let the cat out of the bag: for certain purposes, this software is a Mac user’s secret weapon.

•        What is DEVONthink?
•        Who Needs a Personal Database?
•        How Does DEVONthink Work in Practice?
•        Continue to Part 2 »
What is DEVONthink?
DEVONthink, from DEVONtechnologies, is a pack rat’s dream: as a free form database, it’s a digital analogue of a box into which you can throw pretty much any type of file. DEVONthink does just a bit more than a box, however: built-in artificial intelligence technology automatically analyses the content of what you throw in, classifies it, groups it, interlinks it, summarizes it, and serves it all up back up to you in the blink of an eye.
Want to see a list of the main topics in an article? No problem. How about finding all your other files that are related to a given paragraph of that article? It’s easy. Need to search for all the files that include something roughly like a particular phrase, but which are also focused only on a specific topic? It’s a piece of cake.

DEVONthink handles everything from emails to PDFs, from music files to photos, from movies to browser bookmarks. And the data needn’t come just from files you already have sitting on your hard drive: you can also grab them live from web pages or RSS feeds (browsing: built-in), scan them in from paper documents (optical character recognition: built-in), or create them on-the-fly (text editing environment: built-in).
Who Needs a Personal Database?

If you’re an academic or a researcher — say, working in a mental health field — you already know what it’s like to amass mountains of data…and then have to manage it and make intelligent use of it. But the fact is, you’re not alone: with capacious hard drives, zippy processors, and always-on broadband connections to the internet, today’s computers invite everyone to accumulate files ranging from fairly useless digital detritus through to irreplaceable gems of wisdom and important business documents.
So if you have a mass of information, and you want to organize it and retrieve it easily — or if you want to explore or make use of the interrelationships within it — then the answer to the question ‘who needs a personal database?’ is you do.
How Does DEVONthink Work in Practice?
Long-time readers will know that before reviewing any piece of information technology at — whether hardware or software — we make a point of actually using it, and putting it through its paces in the real live business environment of a small mental health practice and web publishing venture. I’ve personally been using DEVONthink Pro Office (this is one of several versions; see ‘System Requirements and Pricing’) in research and business every week since shortly after I returned from paternity leave in late March 2007. When I came back after 3 months off, I had a fair bit of catching up to do, but I also had a renewed determination to spend less time on organizational tasks. If I could get up to speed with it quickly enough, DEVONthink promised to do some of my jobs for me. Here’s how I got on with DEVONthink Pro Office version 1.3.2, interspersed with a few usage tips that may be of some help if you decide to try out the free demo of the software.
I’ve heard many people comment that DEVONthink has a steep learning curve, but in my experience it’s more appropriate to say it has a steep appreciation curve — i.e., it took awhile before I really began to appreciate what I could do with it. One day I just decided to start with web pages I wanted to save for later use: as I encountered them, I began dropping them into DEVONthink rather than into the various locations across my hard drive where I’d been accumulating bits of information for years. (I wasn’t prepared to jump in headlong just yet: my old archives were staying just where they were!) And at first, this was little different than what I’d always been doing: just grabbing web data and, using the 2-pane or 3-pane DEVONthink view of my collection, stuffing it into hierarchically organized folders that I created along the way.
Usage Tip: If you’d just like to have a quick look at the software’s capabilities, try importing an existing collection of folders full of documents, or grab a sample data set from the developer’s web site.
So at first, nothing much was different, I wasn’t getting any significant benefit, and learning a new software environment demanded that I invest some time rather than saving it. I had little motivation to try out new capabilities, because with such a small database, few of them would prove useful. Hmm. I persevered.

Soon I tried out the ‘auto-classify’ feature and discovered that if I already had a good number of items filed in one ‘group’ (a folder, in terms of the visual interface elements), DEVONthink was pretty good at spotting new items that were similar, and zooming them straight into that same group. No navigating required: just a single keyboard shortcut, and DEVONthink did its best to fit new items into my existing organizational structure. Perhaps this actually could start saving me some time.
Usage Tip: Clicking on ‘Classify’ at the bottom right of the DEVONthink window instead of using auto-classify will slide out a drawer showing DEVONthink’s best guesses at related groups, allowing you to choose easily where to drop your data.
So far, I was just amassing and organizing information, but I wasn’t making much actual use of it. Sure, I could get some speedy search results for items in the database, but this wasn’t terribly different from using the Mac’s built-in Spotlight search facility to search document contents; and it was much slower than using LaunchBar (see Objective Development) to access an item whose file name I happened to know in advance.
Then one day I needed to find the answer to a particular programming problem. (In my case, the web publishing side of my business requires a fair bit of nuts-and-bolts work with mixed-metaphor alphabet soup: PHP, XML, XSLT, AJAX, etc.) I couldn’t remember exactly what the solution looked like, but I was pretty sure I had something relevant. Bingo! A quick fuzzy search and a look at what DEVONthink thought were other documents topically relevant to one I was checking, and there it was: the answer was right there, and getting to it required a few seconds only because that was how long it took me to scan results and click a couple of times.
This ‘See Also’ capability, even though it is reportedly just built atop simple word frequency analysis, has turned out to be uncannily accurate at spotting relationships between different items.
Usage Tip: In addition to DEVONthink’s ‘See Also’ button, which slides out a drawer of related documents, the little right angle quote icon at the top right which looks like a fast-forward button provides a pop-up list of key words from the document you’re viewing. Selecting any of those brings up a list of items relevant to that word, along with DEVONthink’s estimation of the relevance.
OK, now I was getting the hang of it, and DEVONthink had for the first time definitely saved me some hassle in retrieving something I knew I had but which I couldn’t remember or describe in enough detail to locate easily with traditional content searching methods like Spotlight.
It’s now happened several times since then, and the fact that it has happened has begun to save me time in another way. Because I know DEVONthink is so good at finding relationships, I’m starting to devote less time and brain power to reading the items I save. I no longer have to remember as much about the particulars of a piece of information in order to retrieve it again: I do less work, DEVONthink does more.
My primary database is still very modest in size, at under half a million words. As the database grows, however, I find that I do now want to drop in the occasional nuggets I’d previously filed elsewhere, and I certainly find myself taking advantage of the classification and search features more often. And I am distinctly aware that I am getting much more use out of the things that I take the time to save than I used to. (Another database contains around 3.5 million words of psychology texts, but I don’t use that one in the intensive way that I do my main ‘daily’ database, so these types of observations are less relevant there; it does, however, let me know that DEVONthink appears to have no problems at all scaling to handle several million words.)
Usage Tip: DEVONthink needn’t import items in order to index them; if you prefer, you can leave things where they are and just ask DEVONthink to store the information in them, always referring back to the originals for display purposes. (This enables you, for example, to create a database of information physically stored on optical discs.)
The gist of my experience is this: yes, it may take some time and effort before you can really get something out of this software, but if you start realizing the sorts of benefits I’ve experienced, that time and effort will have been well invested.
Part 2 on DEVONthink Pro Office
•        « Return to Part 1
•        DEVONthink Has More Treats in Store
•        DEVONthink Still Has Room for Improvement
•        Conclusions
•        System Requirements and Pricing
DEVONthink Has More Treats in Store
I can’t cover all that such a sophisticated piece of software has to offer in this brief review, but I’d just like to nod in the direction of a few more little morsels in addition to what I’ve covered in the previous section:
AppleScript, Automator and Bookmarklets
Great integration with the Mac’s built-in automation technologies makes it easy to grab text or images from other applications and drop them into the database; I use the Safari bookmarklets for archiving and bookmarking all the time, and DEVONthink’s AppleScripts fairly often.
‘Site Sucker’
DEVONthink can be set to download and archive arbitrary sets of pages — even whole web sites — with just a few clicks.
Bookmarks, Web Browsing and RSS
Although it would be a shame to use it only for organizing bookmarks or reading news, DEVONthink makes a very capable bookmark manager, web browser (using the same underlying rendering engine as Safari) and RSS reader.
Mail Archiving
DEVONthink seamlessly imports email from all major email applications, including Apple’s Mail and Microsoft’s Entourage.
Database Sharing
I can’t quite imagine myself wanting to use it, but if you’re on a LAN and want to share your database with colleagues, DEVONthink provides a front end to enable them to search it via a local web front end. Unfortunately, most of the more advanced capabilities are not available through this interface.
Mentioned only in passing above, DEVONthink’s integrated optical character recognition (OCR) engine enables you to scan paper documents and immediately have them indexed and incorporated into the database: in other words, when you scan the document, you have not only an image, but the actual text of the document to work with. Of course, stand-alone OCR software has been available for a couple of decades, but DEVONthink integrates it well enough to make it actually usable on a routine basis: just scan the document, and there it is, in your database.
Note that some of the above capabilities are not available in the basic (and less expensive) versions of the package.
DEVONthink Still Has Room for Improvement
If you’ve read this review so far, you might have noticed that I’m pretty amazed at what this software can do in terms of saving me organizing time, revealing relationships between sets of information, and helping me get more real benefit from the things I take the time to store away. All of these translate into real, tangible value for me in my business. Having said that, there are still a few rough spots and omissions.
Perhaps the most obvious is the complete lack of transparency for the Mac’s built-in search facility, Spotlight: information which is easy to find once it is within DEVONthink’s database becomes impossible to find using Spotlight. I don’t always know whether I have something tucked away in a folder somewhere, or whether I’ve stored it in DEVONthink, and although I don’t actually use Spotlight very often, it would be a big plus if it could at least let me know when something is stored in a DEVONthink database on those occasions that I do use it. (Responsibility here of course lies with the DEVONthink developers: this is not in any way a weakness of Spotlight.)
This touches on one issue which may be a deal breaker for some users: the use of a proprietary database scheme at all. While DEVONthink does a decent job of exporting (not least by simply dragging and dropping to the desktop), many users will be apprehensive about trusting any large chunk of information to a proprietary monolithic database. I count myself among those users who are apprehensive, although that hasn’t stopped me from using either DEVONthink or a few other software packages which also rely on such databases (Microsoft Entourage comes to mind). Opening up the database to make it accessible in other ways apart from DEVONthink itself would go a long way toward easing these concerns and would no doubt win the software some new converts.
Less obvious are the little interface touches (or their absence) that impact on the experience while working within DEVONthink. For example, while the auto-classify feature is reasonably competent, it doesn’t actually show where it’s decided to put things: they’re just whisked away to their new home without any indication of where they’ve gone. Perhaps that’s fine if auto-classify guesses correctly, but what about those occasions where I’d have preferred it to go to a different place? (And I usually have a clear idea where that place is: I almost never use auto-classify unless I already know where something should go, and I just want the speed of DEVONthink to do it for me rather than taking the time to do it myself.) Just as irksome is the floating palette-style window which pops up to inform you when DEVONthink has been unable to auto-classify something: it can’t be dismissed with a keyboard shortcut and requires a mouse trip over to click it away.
DEVONthink is actually chock full of non-obvious interface elements which unnecessarily obscure functionality for new users. For example, earlier I mentioned the pop-up list of terms used within a document, where selecting a term will generate a list of related documents. It’s a great capability. So why is it hidden behind a tiny little icon that looks like a ‘fast forward’ symbol? Give me a nice big mouse target to hit, give me a contextual pop-up menu, or — better yet — give me a quick keyboard shortcut to the same list, so I don’t have to waste time mousing around.
This problem isn’t limited to the obscure terms pop-up: the whole DEVONthink interface is starting to look a little long in the tooth. If it were merely cosmetic, it would be less of an issue, but some of these things impact actual usage pretty significantly.
In terms of functionality improvements, I’d really like to see thumbnail images for web archives; although DEVONthink already provides thumbnails for PDFs and image files, you get only a generic icon for web archives. Scanning visually when I already know what something looks like is, for me, a very powerful and fast way of finding something — that’s one reason why, for example, scanning through photos in iPhoto can be so handy. But unless you’re looking for a PDF or an image, visually scanning is essentially useless in DEVONthink.
And when it comes to metadata, DEVONthink oddly enough offers very little; it would be handy, for example, to be able to enter comments or other information about a given item without having to generate a completely separate item. If I save a web page, I’d like to be able to store my own notes or annotations about that page in a way that is automatically linked directly to that page — not by writing a whole new note document and putting a link in it to that page, but rather by permanently attaching the annotations to the saved page in a way that allows them to be accessed directly and immediately whenever I’m viewing the page.

I was going to call this section ‘Conclusions for Private Practitioners’, since I think of the software mainly as a time saver in my business and research work — but in fact, I think my main conclusion applies to a much broader set of people than mental health professionals (or academics or journalists, for that matter). My main conclusion is just this: if you find yourself spending significant amounts of time managing or using a large body of information, and you need a tool to reduce the time spent managing it while also enhancing the experience of actually using that information, then I believe DEVONthink is worth a serious look.
You can download a fully-functional, time-limited demo at the developer’s site, along with sample data files and a whole host of tutorial material.

Kommentar verfassen

Deine E-Mail-Adresse wird nicht veröffentlicht. Erforderliche Felder sind mit * markiert