Application Architecture?

[Quantum7]

Is there a white paper or documentation (publicly or private) regarding the application architecture of UltraRecall?

In retrospect this is also one of the main things missing from the roadmap; an architectural vision on where the application is heading / what the goals are in i.e. 5 years.

In the roadmaps current form it seems alot of feature's are simply there because users want them, instead of them being embedded in a company vision on the application (making feature bloat prevention easier).

This post was triggered by the following remarks from ashwken:

Anything that can be done to improve the Output-Reporting-Presentation layer would be welcome.

Maybe it's just me, but Information Management sounds an awful lot like Content Management which leads to Database Publishing.

I'm curious to see what they come up with.

[kinook]

Not really. If you're familiar with software development, we do things more in the vein of Agile development than the Waterfall method.

The vision for UR is make its users more productive, and the future direction of UR is highly influenced by user feedback, since they usually know best what makes them more productive. We do believe it's important to take care when adding features to integrate them well while preventing UR from becoming bloated and slow. This is not easy, and that combined with the fact that UR has so many potential uses, means there is still plenty to do before UR is complete (if ever), but we're doing our best to continue improving UR based on user input.

[quant]

Originally posted by Quantum7
In the roadmaps current form it seems alot of feature's are simply there because users want them, instead of them being embedded in a company vision on the application (making feature bloat prevention easier).

I wouldn't see feature bloat as a problem. Take for example Word or Excel, they have literally thousands of functions, if you use only 100 of them, fine ... the problem is that the programmer's time is limited so they have to prioritize "properly" ... ie. in line with their vision of program's future.

Originally posted by kinook
The vision for UR is make its users more productive, and the future direction of UR is highly influenced by user feedback, since they usually know best what makes them more productive.

that's great. Who can know better than the users who use the software all day long every day? Obviously, it's good if the user base is wide enough to cover people from various background and possibly already having experience with different programs and bringing their "know-how" to UR. I'm always glad to see people joining UR forum sharing their experience.

Slightly off topic, nevertheless I state what I consider to be important feature set to be implemented in the future:

1. UR already provides very good file management. But with growing databases (speed is still excellent), UR search features become bottleneck. This needs to be addressed, either by improving the internal search or by allowing other search programs to index UR databases.

2. Time management. Not much to be said here. Proper calendar pane with all the 'obvious' features is needed.

3. I envision that UR could be a very competitive "Mind Mapping Tool", with only a little effort. When I trialled various programmes like TheBrain, Mind Manager, ... I didn't find anything special in there. UR already contains almost everything, logical linking, internal linking, hoisting, the files/web can be attached, item text/notes, attributes. The only missing thing is the "presentation layer". This could be done as a few simple layouts in Data explorer pane.


So to wrap it up, UR motto could be:

Get organized, be productive, be creative:

- organize all your information
- organize your time
- once you organized yourself, that's not all there's to it. There's a next step -> be creative. View your information effectively, and use it to create new information/connections from your ideas.

[Quantum7]

Originally posted by kinook
The vision for UR is make its users more productive, and the future direction of UR is highly influenced by user feedback, since they usually know best what makes them more productive. We do believe it's important to take care when adding features to integrate them well while preventing UR from becoming bloated and slow. This is not easy, and that combined with the fact that UR has so many potential uses, means there is still plenty to do before UR is complete (if ever), but we're doing our best to continue improving UR based on user input.

One of the main reasons I'm attracted to Ultra Recall is that it uses template items for managing info items. I'm pretty sure that extremely flexible approach to information management is not something which would ever score very high as a 'wanted' feature for a tool (more easily understandable features, like calendar pane's will always score higher).

What I'm saying I guess, is that I'd not like to see implementing features that users want, be more important than architectural vision / design. In essence I don't think user input alone is enough for keeping a tool flexible and powerful long term.

Personally in example I'd rather have a a few months of very few user features implemented and get something generic (often usable for situations we never even thought of yet) instead. The roadmap right now doesn't seems to contain a lot of those kinds of features.

Roadmap example: I hope that "Title attribute based on other attribute values" is implemented in a much more generic way.

If you meant something like the above with "integrate them well", then my apologies for this long post .

Notes:
- The users on the forum only represent the current user crop; it doesn't say much about what features people might miss who aren't using it yet.
- Quant: Interesting articles regarding feature bloat:
http://www.codinghorror.com/blog/archives/000980.html
http://www.codinghorror.com/blog/archives/000973.html

[quant]

Quantum7,

I totally agree with you.
I have a similar view on this. For example, when you mentioned "title attribute based on other attributes", I suggested long time ago for this to be very generic (for beginning only based on single attribute). Now, Item Title is only a string, but in fact it could be any supported attribute and then the related items would be sorted not alphabetically but sorted by the natural set order rules of the given attribute. Example is the date ordering, the date can be in various formats, like 2007-11-20 or 11-20-2007 ... if they are ordered by "date", not string, they would always be ordered exactly as you expect, not alphabetically ...

[cnewtonne]

If anything, I would like to see UR adopt a formal beta program. This includes things like (at a minimum)...

- official release of beta versions prior to production ones.
- a formal stress/integration/and or regression testing script.

I see users here are as dedicated as the developers themselves. I do not thing we will have an resource issues organizing them.


Just an urging thought.

[ashwken]

When I first looked at UR it struck me as an Outlook clone, which I already had access to, and even with UR's basis on sql it was difficult to justify the price.

I spent a good amount of time and effort working with TreePad Business Edition before hitting too many walls - I was also put off by the factured development and lack of a proper user forum.

I took a fresh look at UR and was nearing the end of my trial period, ready to bite the bullet on pricing, when the program became available (at substantial discount) at Bits du Jour.

The introduction of Forms in v.3 reaffirmed my belief that this was a program undergoing active development, and that UR might realize the potential that I see in the program.

I understand that Kinook must balance the development of UR as an sql front-end against the preceived needs of their users and the marketplace.

My needs may well fall in the minority and I may be expecting too much from a product in this price range - 'course, now that I'm "in" the upgrade cost has been acceptable.

Coming from a database perspective:

_ the availability of Attribute Values as variables would be welcome (Calculated Fields), but I have no idea as to the ease or difficulty of implementing this in an elegant fashion.

_ Reporting or Database Publishing or whatever you want to call it needs attention. It is my hope that the soultion is more than a banded report writer.

Coming from a business perspective:

_ Multi-user enhancements would be attractive.

I don't know if Kinook can accomplish any of this under their current pricing/upgrade structure, or if some of this funcitionality needs to be offered as an additional cost add-on.

I will say that as things currently stand UR satisfies a large percentage of my information management needs, and I am encouraged by the continued development and responsiveness of the developers.

[eno]

Personaly I do not like companies with programs that start out as offering a solution only to increase the price (out of proportion) as it is developed.

If it changes into another product so be it, but I have found that when a company starts offering all sorts of flavours of a program the customer is often left with cutdown functions on the ones that previously had the functions unless they "upgrade" to the "Super Pro Nuclear Power Version" or similar.

At the moment Kinook has got it just about right, with the standard and Pro version and with not too many paid upgrades but with a program that has evolved into a powerful good and stable offering it is now.

I swapped from another program to Ultra recall because the company listens to its customers and develops the program when needed, Kinook's forum with the personal input of the company is also a strong Yes for me.