Categories in grids and other suggestions

[crcowan]

Hello, I'm a new registered user and have a few suggestions and a question. Suggestions first:

1) You could provide a very powerful addition by taking advantage of what I will term "categories" in the related items grids. By categories I refer to the hierarchy of links and assignments in the data explorer. By assigning a link you are in effect categorizing an info item. You would need to be able to show multiple items in a cell of the grid (an info item can be assigned via multiple links). You should have the ability to choose what is shown in the grid cell based on a "category" as a header. For example, if an info item is linked to a child of the category which is a header you might show the name of the particular child, or a "yes/no" indication to show it was assigned to any particular child of that category. The model I am thinking has best been implemented in an old Lotus product called Agenda. See this page for more information and the ability to download and experiment with Agenda: http://home.neo.rr.com/pim/alinks.htm . You have many of the beginnings of what you need (and it is one of the reasons I registered) and adding some of the other capabilties of Agenda could rock the marketplace. Everything from hierarchical inheritance to autoassign are things you could do with your model. Some of these things, of course, would show up in the attributes pane as well.

2) The grids should be able to show all of the children of a particular parent (not just the direct descendants).

3) I'd like to be able to have a rich text page that I can place information into for stored files. It is rather stark with just the "click here" line(s). I can imagine pasting an image of the title slide of a powerpoint deck, or the first page of a word document . You could probably just use your rich editor and add a button to the top to launch the document rather than the click here line. (of course, automatically extracting the first page (or so) of a document as it is being stored (with the ability to replace or delete it) and placing it in the rich field would be even better)

4) XML (preferable) or other import/export for items including rich text, attributes, etc as a connected whole.

5) I second requests for user defined "forms" like the contact form (with each field reflected as attributes).

6) Automatic creation of attributes. You could extract some level of information from many types of files when they are imported (eg: document author from Word, exposure settings from digital pictures). I suggest a plug-in architecture for which you publish the API. Let others create plugins the return to you a blob (or blobs) for the document and associated attributes.

7) Please don't turn this into a standard calendar/contact type of PIM, major in information management.

Question: (or perhaps really a request for more documentation) What does the attribute "Original Parent" mean?

Thanks for a remarkable first release. I appreciate the opportunity to make suggestions, and I wish you the best!

[kevina]

I've responded to each question inline below:

1) Agenda Style Categorizing (in Related Items Pane):

Response)
a) We did review Agenda during the development stage of Ultra Recall, which provided some inspiration for the design in Ultra Recall. A few similar suggestions have already been made on the forum, and we are interested in implementing something along these lines. To be honest, we are still grappling with exactly what is being requested and how to design it while maintaining the purposefully uncluttered, intuitive and flexible design we already have. In addition the feature must not compromise the high-performance standard we insist on maintaining.
b) From the different requests we have received, it seems to be a type of data pivot or crosstab of the Info Items where the column headers are Info Item names, and the rows are Info Items that meet a criteria and indicate whether they are parented "related" to the Info Items in the headers. Is that a general synopsis of what is being requested?
c) If you have specific examples or diagrams of what you would like to see provided, please post them to the form or email them to support@kinook.com.

2) The grids should be able to show all of the children of a particular parent (not just the direct descendants).

Response) This is already on our list as well, but we haven't arrived at the optimal implementation yet (one thought was a system Attribute that could be assigned and given a value that would flatten the child list (show all children). Does that sound reasonable or do you envision a different implementation?
Note: currently you can create a quick search with no criteria (maybe named "All Children"), limited to search the bold explorer item, in the Search Pane. To use it, select the item to show all children for (in the Data Explorer Pane), then the 'All Children' Search item to view all descendants of any item.

3) Rich text entry for non-displayed documents in Item Details Pane:

Response) This is available now with the Item Notes Pane. By default this Pane is docked with the Item Attributes Pane, which normally hides it and typically makes it fairly small. You can dock it in a different location to make it larger and/or always visible... One issue with your suggestion is that if the document type (extension) is added to the list to be displayed in Ultra Recall, then the rich text that was previously assigned will no longer be visible/accessible...

4) XML other import/export...

Response) XML import/export is already on the todo list, near the top so it hopefully will be available in an upcoming release of Ultra Recall. Do you have a schema in mind for this import/export (such as OPML)?

5) User-defined forms:

Response) As you noted, this request is already on our todo list, and will take some work to implement, but we agree that will be a handy feature.

6) Automatic creation of attributes...

Response)
a) This feature is available now. It is documented Under Document Summary Attributes in the helpfile (online @ http://www.kinook.com/UR/Manual/?documentsummaryattributes.htm) In a nutshell, if you have the appropriate option turned on (Tools | Options | Import (more) | Map Explorer summary properties to existing Ultra Recall attributes), and you create Attributes (Tools | Attributes) that have the same name as the Summary properties you wish to get Item Attributes for, they will be added automatically.
b) I don't quite understand how your suggested API would work. Can you elaborate?

7) Please don't turn this into a standard calendar/contact type of PIM, major in information management.

Response) We tend to agree with your assessment, however we have had numerous requests for more calendar/contact management as well. Our goal is to excel at information management, but if we can add features that facilitate the latter without compromising the former, we will probably add them as well (within reason).

What does the attribute "Original Parent" mean?

Response) The "Original Parent" system attribute is used to indicate the original location of Info Items deleted to the Recycle Bin (this attribute should only be assigned to Info Items parented directly to the Recycle Bin). Perhaps this should information should be added to the Help File.


Thanks for a remarkable first release. I appreciate the opportunity to make suggestions, and I wish you the best!

We are confident that Ultra Reall is an innovative, useful product with a lot of potential. Thank you for your positive feedback and encouragement!

[crcowan]

[QUOTE]Originally posted by kevina

Great response, thank you. I will respond to your responses (with extracted quotations) :-)

1) Agenda Style Categorizing (in Related Items Pane):

Response)
b and c) From the different requests we have received, it seems to be a type of data pivot or crosstab of the Info Items where the column headers are Info Item names, and the rows are Info Items that meet a criteria and indicate whether they are parented "related" to the Info Items in the headers. Is that a general synopsis of what is being requested?

Yes, that is the general idea. Assume that the gird allowed a column to be chosen from any of the entries in the data explorer - call it the column title. In addition to choosing the data explorer entry, the user would also have the ability to select a "display property" for the column. At the intersection of that column and a particular row the cell would be filled with something that depends both on the item that the row represents and the way you want the cell displayed ... the "display property". For example, if the display property were "name" then the cell would contain the item title for each child (or grand child, etc) of the column title that was linked to the item the row represents. Example: an item called "priority" could have children of "high", "medium", and "low". The "priority" column would show "medium" if the item that row represented was linked to the "medium" item. You would want to allow the selection of "high", "medium", or "low" in the cell and that would cause a link to be created. You might allow an item to be linked to both "medium" and "low" in which case both would be shown in the cell or you could implement "mutually exclusive" capabilities that could be assigned to a category and thus allow an item to be linked to only one of the children.

Other display properties might be "Ancestor" to show the next level up in the hierarchy, "Yes/No" could be to just show if any child were linked to, etc. There are other options, but I hope you get the idea.


2) The grids should be able to show all of the children of a particular parent (not just the direct descendants).

Response) This is already on our list as well, but we haven't arrived at the optimal implementation yet (one thought was a system Attribute that could be assigned and given a value that would flatten the child list (show all children). Does that sound reasonable or do you envision a different implementation?
That might work, or you could create a new grid. For that grid the user provides a list of items from the data explorer (drag and drop probably). For the item in the list you say whether you want just the items linked to that item, or the items linked to that item's children and grand children as well.
Note: currently you can create a quick search with no criteria (maybe named "All Children"), limited to search the bold explorer item, in the Search Pane. To use it, select the item to show all children for (in the Data Explorer Pane), then the 'All Children' Search item to view all descendants of any item.
Good idea!

3) Rich text entry for non-displayed documents in Item Details Pane:

Response) This is available now with the Item Notes Pane. By default this Pane is docked with the Item Attributes Pane, which normally hides it and typically makes it fairly small. You can dock it in a different location to make it larger and/or always visible... One issue with your suggestion is that if the document type (extension) is added to the list to be displayed in Ultra Recall, then the rich text that was previously assigned will no longer be visible/accessible...
Yes, I know about the notes. But the rich text pane is so much more "in your face" and is the place I invision working. To resolve the issue I suggest you think of the rich text pane as having two sections, the top and bottom. Place your link (and the viewer if the extension is in the list) in the top section. Allow rich text entry in the bottom section.
This way, when navigating using the data explorer, what pops up in the item details is consistent across different item types

4) XML other import/export...

Response) XML import/export is already on the todo list, near the top so it hopefully will be available in an upcoming release of Ultra Recall. Do you have a schema in mind for this import/export (such as OPML)?
Yes, I suspect OPML may well be the best solution - with the hope the standard will settle down

6) Automatic creation of attributes...

Response)
a) This feature is available now. It is documented Under Document Summary Attributes in the helpfile (online @ http://www.kinook.com/UR/Manual/?documentsummaryattributes.htm) In a nutshell, if you have the appropriate option turned on (Tools | Options | Import (more) | Map Explorer summary properties to existing Ultra Recall attributes), and you create Attributes (Tools | Attributes) that have the same name as the Summary properties you wish to get Item Attributes for, they will be added automatically.
Yes, I saw that. I suggest, though that you automatically create the attributes so people don't have to know the attribute names to create

b) I don't quite understand how your suggested API would work. Can you elaborate?
Define a standard C or C++ or whatever subroutine that would be passed the file name. The routine would be expected to return an array of attributes and their values. Perhaps it could also return a BLOB that you could save in the database the way you do with the supported applications. It might also return rich text to fill the rich text form. It would probably have to be encapsulated as a DLL. Then provide a mechanism to associate those DLLs with a particular file type and an application. You would then call the routine in the dll when a user dropped a file of that type and then you store the data in your database. There could be a similar routine in the DLL you call when the user wants to open the stored document. You pass it the BLOB and let it invoke the associated application. Publish the specs and how to write such a DLL. Then other people could provide the DLL to support their specific applications and files.

Thanks again.

[prf5]

I can see that there's confusion here about what various people (at least, crcowan and me) are requesting and how our requests are inspired by Lotus Agenda. I'm not requesting the feature that crcowan has identified. I'm not interested in creating a grid of items and category assignments. I'm especially concerned when the moderator's says: "From the different requests we have received, it seems to be a type of data pivot or crosstab of the Info Items where the column headers are Info Item names, and the rows are Info Items that meet a criteria and indicate whether they are parented "related" to the Info Items in the headers. Is that a general synopsis of what is being requested?" For me, the answer is a definite "no." I think that pivot tables are cool, but a pivot table-like capability or a crosstab display is not what I'm interested in, and by the way, that's not how Agenda worked.

Let me devote one paragraph to explaining Agenda, then return to my original request. I think that crcowan has it about right. Here's an example: A hypothetical entry goes, "Lunch with Dave Smith at Scoozi's on Tuesday." This entry might display in Agenda as follows: The item appears in the left-most column (not necessarily, but let's asume it does), then there's an activity column in which the word "lunch" appears, then a person column in which "Dave Smith" appears, then a place column in which "Scoozi's" appears, then a date column in which Tuesday's actual date appears. The assigments can be made either explicitly or implicitly, by means of text-matching. Each of the columnar entries is a category. The categories are related to one another in a hierarchical fashion, governed by the principle of "inheritance." For instance, "Scoozi's" is under "restaurants," which in turn is under "places." By virtue of inheritance, "Scoozi's" is also member of "places."

I'm not interested either in automatic assignment to categories or a matrix-like display -- though others may be asking for that. What interests me is the ability to create views, understood as parts of the database arrayed in the explorer window that meet specific, user-selected, categorical criteria. Example: By now, my explorer window has become populated with old notes and materials that I don't want to throw away but don't want to see every time I open the program. There are various ways to identify these old entries. Using any of them, I'd like to filter them out. Another example: I receive TOC's of new journal issues. I want to create a view that shows only "Social Science and Medicine" and not the other journals. I'm aware that I now am able to filter the database. However, the search results pane does not have the structure that I am looking for, because along with the "Social Science and Medicine" items I want a tree-like view that includes the item's category assignments (for instance, to project A, paper Z, and person X). Similarly, In the "filter out old entries" example, I want to see the entire explorer tree without the entries that have been filtered out.

Paul