Here are 3 wishes for DTPro’s improvement (they correspond to my own way of using DTPro, and different users may have different needs):
1-Floating (semi transparent) windows, with the possibility to hide/show the bottom strip (with see also and classify buttons) for users who need to save screen space.
2-Possibility to open more than a single DB at a time. This would enable the user to improve organization of his datas (smaller databases, etc…)
3-Spotlight compatibility.
I know that 2 will come with DTPro 2, but what about 1 and 3?
Any GUI design is a reflection of how the programmer believes the user will interact with the system. Getting the design right in an application of any complexity requires reading minds, seeing the future, or trial-and-error. The problem is in “the way users tend to work”; for an email client or a web browser, this is simple.
But how do users manage their information? The answer: they all do it differently. It seems to me that those having problems using DevonThink do not have a clear idea of how it should work, aside from what I would call annoyances or purely technical hurdles [like better OS integration]. This is a pretty good sign; a few iterations following the principle of least surprise should make the UI quite friendly for new users.
To the developers, I would suggest more views. How do the users expect to perceive their data? Should the ‘database as a wiki’ be made a view? Maybe a more filesystem-like interface? Would representing the current group as a document [like the exported outline or rtf] be useful? Brainstorming and experimentation should come up with some interesting views.
Also, the sort methods that Bill mentioned should be a button group like the views: that would make them more obvious
On the whole, I like the interface. When I first tried it out, it worked like a database with a filesystem interface – quite a nice way to manage data! Of course, both databases and filesystems assume that the user is somewhat technical, and since Apple attracts non-technical users, I can see why this approach might be hard for some to swallow.
While I’m posting, I’ll digress a bit and mention something I personally would find useful, but which I haven’t dug around enough in the scripting end to determine if it exists or is possible: plugin handlers for custom document types. As an example, I have binary data files which I have written perl and python code to query, print, plot, etc. It would be great to be able to import this data into DT, and have a plugin I write be responsible for the data … perhaps to import it as a sheet or rtf, or better yet to be called by DT in order to render the data as rtf [in order to display it in-place in DT]. Anyways, this is deserving of another post, let me put some more thought into it and poke around a bit.
OK. But I am a software/hardware technology developer. All I do is develop new software and hardware solutions for my clients. I would hardly call this a non-technical profession.
I don’t believe that the problem is that the users are not somehow qualified to use it. The interface does NOT behave like any file system which I would find acceptable. While its behavior may be predictable after some period of training oneself on what to expect, I think that alone may indicate a problem.
I choose the Mac not because I am not technical, but because I appreciate the logic and beauty which are inherent to many pieces of Mac software. If anything, I feel like DevonThink seems like an extremely good piece of Windows software, which was ported to the Mac and doesn’t exactly capture the Mac’s elegance or expected behaviors yet.
You’ve got me there. I’m a UNIXer, and most of the Mac UI’s “expected behaviors” frustrate me because I cannot configure them to act the way I feel a reasonable system should. Aqua alone reduces my ability to multitask by half, compared to something like Enlightenment… in other words, I wouldn’t notice if DT was not Mac-like.
A hierarchy of branch (group) and leaf (item) nodes that can be arbitrarily reordered without effecting integrity? Creating hard links (replicate) to items on the same filesystem (inside DT), and symbolic links (index) to items on another filesystem (outside DT)? Meta-information such as create/modify times? Maybe you’re more demanding of your filesystems than I am
I’m not trying to imply that users need some special training or to drink the Devon Kool-aid in order to use DT, but rather that it probably helps if the user sees the problem in the same way as the developers appear to. For me, information management has always been largely an issue of organizing tens of thousands of documents and notes on-disk. DT solves that problem well; I was up and running with it in an hour, and by the end of the week it was indispensable. Other users may be trying to solve different problems, such as email or contact management, and their mileage may vary.
That said, there are plenty of gotchas with DT, and a fair share of ‘learning what DT is doing’. The difference between indexes and links, or duplicate and replicate, are good examples. Many of these are scheduled to be fixed in the eventual 2.0 release, which doesn’t help now but at least means they are being addressed.
The 1.1 release which was posted today on VT has some great improvements. Personally, I’m still not ready to buy-off on it and move my data into DT. But 2.0 is sounding like it may resolve a lot of my current concerns. Curio* is pretty much meeting my needs right now, and has better hooks into the OS… unfortunately, it isn’t Applescriptable, so there is no nirvana. Yet.
Yes, I know that Curio is built on a completely different metaphor, but in the end, they actually perform similar functions.
DT Pro 1.1 (just released) simplifies and reduces user options for bringing new content into a database. “Index” results in providing links to files that remain external to the database, and captures into the database the text of all readable document types. “Import” (using File > Import > Files & Folders) copies files into the body of the database (all text-type files) or into the database Files folder for other readable file types, such as PDFs (and, with a Preferences choice, will also import “unknown” file types such as Excel or KeyNote into the Files folder). These import modes will also be used when DT Pro version 2.0 is released, except that in DT Pro 2.0 all Imported file types will be stored in the database Files folder.
Synchronization (one-way, from an edited and saved file to its display in the database) becomes simple and reliable in DT Pro 1.1. Files captured by Index can be launched under their parent application, edited and saved. The reulting changes will be displayed in the database next time that document is selected. Files copied into the database Files folder by Import can similarly be launched under their parent application, edited and saved, and the edit changes will show up in the database.
Both methods of capturing items into the database have pros and cons. Index-imported databases use less RAM and have advantages for those with limited RAM and perhaps a slower CPU. But if the externally-linked files are deleted, or their Paths are broken, their information is lost to the database. Import-imported databases are larger and require more memory (though the memory requirements will be reduced in DT Pro 2.0). On the other hand, Import-imported databases are more portable (they are self-contained), making it easier to transport a database to another computer, or to distribute a database via DVD. Currently, Index-imported documents link to files that are indexed by Sporlight. Currently, Spotlight cannot index an Import-imported database, but that will happen in DT Pro 2.0.
In either case, DT Pro can export to the Finder all of its document/file contents, so that one is never “locked into” DT Pro. For that matter it’s a relatively easy matter to convert a Index-imported database to a Import-imported database, and vice versa. Note also that when DT Pro exports contents to the Finder, the hierarchy mirrors the organizational hierarchy of the database.
Document replication often puzzles newcomers to DT Pro. When a document is replicated, there is still only one document, although it can have multiple instances, whether in the same group or in different groups. Because there is really only one document still, if one edits any one of the replicants, “all” are changed. Suppose a document “fits” into more than one topical group in my database. I can choose to replicate that document into multiple groups. Or suppose I wish to perform multi-level searches and sorts on the results of a search? I can create a new group, select the search results, replicate them into my target groups, and start playing with the logic of my multi-level search/sort possibilities. When I’m finished I may, if I wish, delete my working group (with its contained replicants) without having moved or damaged any of my existing documents. Or I can create a “smart group” that, based on my criteria, pulls together as replicants documents that meet my criteria. The document title of a replicant is colored red.
Duplicates are much easier to understand. A document that is duplicated can be edited without changing the “original” document. In fact, since DT Pro is looking at the contents of documents, if I make significant changes to a duplicate document it is no longer similar enough to be still a duplicate, and the blue marking indicating duplication will go away, replaced by a normal black document title (even if more than one document has the same title). Likewise, two documents with identical content but different titles can be identified by DT Pro as duplicates. How about a plagiarism detector, for example? I can look for similarities between documents and vary the parameters that DT Pro uses to mark duplicates.
But one still cannot “grok” DT Pro without understand the artificial intelligence features that make it unique. DT Pro is “aware” of all the text content in each individual document and in the database as a whole. DT Pro can “see” contextual relationships or patterns of word usage. So, if one is viewing a document that contains the term “dog” and presses the See Also button, DT Pro may suggest as related another document that contains “canine” but not “dog”. And that’s why DT Pro isn’t just a simple “filing system”. It is very different from other programs such as Curio, for example.
I did. Because that is how it feels. Windows apps tend to have interfaces which also expect you to learn how to avoid their pitfalls. This is what DT feels like to me. You have to know how to avoid it from doing something funky, or showing you a view of your data that you never intended.
Yes, it can draw conclusions better between some concepts. However, when what you are saving is stuff like code snippets and server settings, and technical data (rs-232 command APIs, etc) these conclusions don’t seem very relevant. Why would I need a “see also” for something like IP addresses or passwords? The value of this feature is dubious for many users, which pretty much levels the playing field with the competition. I pretty much know what I am looking for and the search terms to use, and rarely need the program to take a guess for me. I’ve tried it a few times out of testing the program, but never seen any particular value in the returned results. For the most part, the only thing good about it is that it doesn’t take up much space in the interface, and you don’t need to see its windows if you don’t use it. I know what it does, I’m sure it’s cool for people storing the right kind of data, but for those of us who don’t, it is pretty much a non-feature. If this represents the feature of differentiation, then I guess I haven’t really missed anything.
Now, factor in that the other tools support address book, mail, spotlight, iCal and Safari in an elegant, mac-like way… and ALSO perform the same in terms of indexed documents as compared to embedded, and also have the concepts of “replicas” and “duplicates”, but because they do it in an extremely intuitive way, they don’t seem to need to explain the differences, or their use… you just use them. Plus they have advanced creation tools built in, incredible clipboard support, presentation modes, dot mac syncing.
Sorry. I’ll be on my way now… I’m glad DT serves you so well. With me, and my recent data and workflow dilemma, DT did not win the prize. The developers now know why, and can do with that information what they will.
Sorry to flog a dead horse here but this argument gives your position towards DEVONthink away completely. This feature is the reason why most of our users have bought our program and makes it stand out from the competition. In your case since you don’t need this functionality and basically are looking for a glorified hierarchical file-system with some bangs and whistles added, the Finder and Spotlight (with its comments) are all you need! You just saved yourself some money! Treat yourself to a nice beer (I recommend Belgian ales, my favourite although Christian will probably disagree).
Heheh, yeah it is not the best for technical info. The loss of syntax highlighting alone is hard to deal with, since an external editor has to be launched for each source file.
When I got started using this, I was doing a huge project porting legacy FORTRAN code to C [from there to C++ probably… baby steps when working with production code]. I imported some domain reference material [e.g. DSP references], sample input and output data, all of the FORTRAN source code, all of the auto-generated C code, and a ton of notes, and went to town. Got more done in a week than in the previous six weeks of the project. Made the files groups, then made the functions groups, then broke out blocks of code into their own text files. Used the search features to track variable usage. Stored emails discussing portions of the code with domain experts, linked them in comments in my rtf ‘working copy’ of the code. Then used the Merge tool to combine the reordered code blocks into the new output c files and ran them through indent. Sure there are Java tools that do this type of stuff, but not any decent C or FORTRAN ones.
After that, imported my programming references, manuals, code from various sites and cookbooks, and started using it as a booshelf. From there, managing projects and keeping track of interesting sites, tools, and so forth. Now there is very little I do without it.
The Ai stuff? Semi-useful, since as you point out context isn’t terribly important in technical manuals. I mostly use that for cutting down my research time in whatever the project I am working on actually does.
DT’s strengths are not in simple file management by attributes, but in managing actual information, much of which is contained within individual files.
UI tweaks aside, It is a very powerful tool for information/data management, as opposed to managing files by attributes.
As a user who sincerely loves DT pro, I’m very much disappointed in seeing this thread turning out in this way. I am not a perfect supporter of Tryll. But I am disappointed in seeing how people from Devon technologies responded to him.
First of all, I know that DT pro is distinct from others because of its powerful AI function. I saw that Bill and other people proudly declared that it’s the essence of DT. I agree. However, have you guys ever seriously investigated the percentage of users who actually use “see also” function or those who decided to commit DT because of the function? I am a Japanese and a member (not a representative) of a discussion board for Japanese DT users. We are excited at the software NOT because of its AI function but because of its powerful data management system. For us, Japanese users, “see also” function does not work very well. Still, we cannot replace DT with other software because of DT’s fast and powerful search function and variety of data formats that can be indexed (+commnts, scripts, smart folder, etc.). DT cannot be replaced by Finder. Furthermore, it seems to be obvious from the forum that non-negligible proportion of DT users appreciate non-AI parts of the software.
So, annard, can you say to Japanese DT lovers who paid for the software “I don’t understand why you guys use DT if you can’t use AI-related functions?” Basically, what you said to Tryll is the same as this. I love DT. But, I’m very frustrated to see you guys who are proud of AI functions tend to ignore or even ridicule serious DT users who don’t use AI functions.
As I have lots of English documents in DT and I really appreciate “see also” and other AI related functions. It is a MUST for me. However, I need to access to information in multiple ways. “See also” is just one of the ways to access to information. In order to fully exploit the AI feature of DT, huge amount of information must be imported to database on day-to-day basis. It’s needless to say this requires non-AI part of the software and efficient import of information is crucial for user experience. We cannot evaluate AI functions separately. It’s well organized with other non-AI parts and, as a whole, impresses me. I don’t understand why you guys often discard the non-AI related feature requests and try to finish a thread by saying “it’s not essential to DT.”
I love DT. If user interface gets nicer (e.g., honestly speaking, I don’t like the icon of ammonite, sorry), I love it more. If I can reduce the number of steps in data handling inside the DT, I love it more and more. If DT can have full outlin function whic can replace OO3 with DT, I don’t hesitate to pay double price. It is up to you and the constraints from OS X to what extent future DT reflects features requested from users. I just want to emphasize that it is non-AI parts of DT which allows us to fully exploit the nice AI functions and finally result .
Finally, I want to say that most of DT users on a Japanese discussion board are excited about v1.1 and announced features of v2.0.
Thanks for chiming in, bamb. I wasn’t really looking for supporters, so it’s OK.
But it does seem fair to say that more than just those looking for AI functions are interested in data management solutions. In my case, I’d trade AI for a smarter interface and better use of the OS functions. I figured that the people from DevonTechnologies would be at least interested to hear that opinion, which they can act on, or not. But at least they’d have it, rather than my just saying “well, this isn’t for me” and moving on.
But it sounds like they pretty much know who they want to sell products to, and I (and some others?) don’t fit the bill. That’s fair too. It isn’t quite the reaction I expected, but hey, I learned something from the experience.
I followed this thread with amusement because it seemed to me like a customer who shouts loud in a shop “I am not going to buy this hammer, it doesn’t work for me. Now that I am fed up with your useless product I go to the other shop and buy the saw which is much better for cutting wood.” Can’t stop me writing that
Now to the serious part:
Bamb,
could you tell me the adress of the Japanese user’s discussion board?
I use Japanese and had a lot of discussions with Christian about how to make the software more usable with East Asian languages. There were some improvements, but if we are many, we might push some pressure on them for using concordances and word indices etc.?
Thanks. Good comments. I bought DT back in 2002 precisely because it let me, for the first time, search across the thousands of text, Word, PDF and HTML files already on my computer. I fell in love with the AI features later on.
Glad to hear that Japanese users are excited about v1.1. I think the developers are more proud of the programs released last Thursday than of any previous releases (possibly excepting that very first one). A great many user comments and experiences were responded to in the new releases.
bamb:
I apologize to everybody who may feel I have stepped out of line in my advice to tryll. This was not my intention.
tryll:
No personal offense meant! I want you to find a good solution for your needs, that was the intend of my message. Maybe I took the “porting from Windows” a bit too personal since I have worked very hard not to have to program on that system.
We shouldn’t forget that opinions voiced on a public forum will probably be received in a different way than we originally intended. Therefore let’s think thrice before pressing that “Submit” button in the future.
One thing only will I say. If you are doing science, real science, and you have to keep up with and have rapid access to the materials that are continuously being generated in your field, DEVONthink is the best way I’ve found to accomplish that goal. I don’t even read everything I throw into DT. If I run across anything anywhere that I think might be needed in my current or future work in environmental science I toss it into DT using one or the other of the numerous methods the developers have given us to put it into the DT database milieu. It’s easier than tossing clothes into a laundry basket! Then, anytime I need to have information even adjacent to that area of scientific study it will be available either directly in my search or in “see also.” This tool is indispensable to me at this point. If you are using it only to store IP addresses and the like you are missing out on its true utility. I’d be willing to bet that if you were to think about it you could readily identify other ways in which DT’s capabilities would enhance your business or profession.
It had occurred to me that I could probably capture more stuff than I currently do, due to the way the DT works. I currently have a lot more than IP addresses, but almost nothing that benefits from “see also”, which really does seem to be DT’s “stand apart” feature.
It’s worth considering whether I want to catalog more, or just allow google/bookmarks/etc. to be my catalog for other people’s content as I currently do.
Ultimately it’s all a matter of what works best within your workflow and what feels right when you are doing it. Just experiment around and enjoy yourself. Perhaps I’m a nerd but part of what I do for fun (who needs video games!?, LOL) is experiment around with trying out new programs and fiddling with features I’ve used very little in programs I already own (DT, DA, NoteTaker, etc.) to see if anything works better than what I’m currently doing. Good luck with your endeavors in these explorations!
