in

Emerald Hand

Emerald Hand, Inc. community home page.

This Blog

  • Home
  • Contact
  • About
  • Directory of Computers/Tech Blogs

Syndication

News

I'm back to blogging

Worm in liquid maze

Design and development of information management tools.

Why Sider?

In my earlier post I wrote about Sider, an information management platform. Read it if you want to learn more about what it is. Today I want to explain why I think it's important and why I'm spending time on it. Actually, I'm not working alone. I'm doing most of the stuff, writing code, documentation, etc. My partner, Igor, helps by providing feedback, helping me analyze problems and supplying new ideas.

Igor thought of the Sider and I liked it, so I decided to dedicate my time to it. We both store a lot of information on the computer, notes, e-mails, contacts, project plans, etc. By the way, information is a collection of data chunks with relationships between those chunks. A text document or a table is data, but a collection of text documents organized in such a way that connection between each document is visible, is information.

Unfortunately, current state of managing data on computers is rather weak compared to what is possible. Most programs can handle only a few documents types, and almost all of them are not very extendable. For example, Microsoft Word is focused at working with text, Excel mostly works with tabular data, OneNote stores notes in a tree-like structure, Outlook stores tasks and e-mail in lists a tree (but each task node cannot be a tree node).

As a result I need to split information I want to store on the computer between different programs. My to-do list is in one program - ToDoList, but for recurring tasks that I need reminder on I use Outlook. To store my notes I use Wikipad. To write this blog post I use Word, but to post it - w.blogger. To work with tables (and I do every day, to track time I spend on different projects) I use Excel. For mind maps I might use FreeMind. To work with my information I need to have 5-6 programs to be running and constantly switch between them.

In most cases these programs work separately and I can't reference Wikipad note in a ToDoList task (I can reference wiki file, but usually that's not enough). This need to split up between programs happens because each program was created to operate a single domain and solve only a few problems. In addition many of these programs are inflexible and it's hard to extend them if I want to add a missing feature.

Even when I can create custom extensions for the application, I need to learn it's language and API (Wikipad uses python, Microsoft Office - VBA or .NET, ToDoList is in C++). Moreover, extensions supported are limited in scope. I might be able to add a new view or report, but adding support for a new document type is usually impossible. At best I can try to adopt existing data types, but this is complicated, error-prone and time-consuming.

Sider started off as a more theoretical project, but evolved into an application to address all these problems. Sider stores information in a most-suitable, more intelligent, innter-connected way (a graph is stored as a graph, a tree as a tree, a table as a table, etc.) Working with any information on the computer should be easier and more natural, and that's what Sider is intended to do. Sider makes it easier to reuse information by referencinf it, importing and exporting, and so on.

Sider is flexible and easy to extend. There's no need to try and force existing data type to be suitable for your needs. You can either extend existing type or create new one from scratch, or combine both methods.

Another goal for Sider is to present data in the best way for the user. What is best will change depending on what the user wants to do so it is important to support a lot of different views.

Sider is about helping people to use computers for what they were invented for, to work with data. Of course, it will not solve all problems. When released it will be quite limited compared to most of the existing programs. It probably will never be as good as Word at working with text, or as good as Excel at working with tables. However, it will offer a new and unique approach at managing information and helping people to solve different problems.

In the next post I will talk more about Sider architecture and technologies used. Hopefully it will be an interesting post.

Share this post: Email it! | bookmark it! | digg it! | reddit!
Readability Stats: Word Count: 758; Sentence Count: 45; Grade Level: 8.5, more info...
Published Jul 26 2006, 12:22 PM by Ornus
Filed under:

Comments

 

Worm in liquid maze said:


 I wrote about Sider before, what it is and why I'm working on it. I also maintain a wiki with Sider...
August 7, 2006 8:59 AM
 

PiiXiieeS said:

It's an interesting subject.
What is your opinion about these tools/technologies liken your idea?
 - Ms Sharepoint Services
 - Wiki with plugins allowing you to combine different objects with special features (todo lists, project management, contacts,...)

August 10, 2006 8:03 PM
 

Ornus said:

hey, I'm not really that well familiar with sharepoint services, but I think it and sider have different goals. sharepoint is mostly collaboration platform, to share documents between many users. of course, that means that there should be different types of documents. webparts cover that partially, but I'm not sure sharepoint webparts can be built on top one another to extend supported documents.

sider is specifically designed to allow just that. it's not about collaboration (although it's possible of course) but about storing data in a form that makes it easiest to work with through reuse existing features, and providing separate, independent views for each document. for a example, a document with notes can be shown just as a list of notes, a tree with notes organized, a mind map, etc. I'm not sure this is easy to do in sharepoint.

wiki is really nothing more than graph with wikitext nodes. it's very flexible, but really, there's only single data format, wikitext. it can be used to store tables for example, and there could be many views (openwiking is a good example of this. each skin can actually show data differently). wiki pages can reference each other in any way, but they could be hard to manage, that's the price. in sider each type provides interface to work with it. a tree is always a tree no matter if each node is an HTML note, a task, a wikitext, a list, a graph, whatever. same for all other types. I guess it's possible to build objects on top of wiki that would provide similar features, but I haven't seen anything even close. oh, and also, I really dislike wikitext. it's good for text, easy to edit, but when people try to adopt it for other things, for which it was not originally designed, it quickly becomes cumbersome.
August 11, 2006 11:21 AM
 

PiiXiieeS said:

Ornus,
It really sounds fine.
Anyway I give you some examples from openwiking that seems at least similar to what I understood.
OWNG uses the page object as basic object to store text, links and many other objects that you can embebed by means of macros.
Samples of this objects are TableOfContents, SynonymLinks, groups and authorizations are basic but integrated in special wiki pages, attachments,...
and we can create many others as a Ms Visio embebed, todo lists, ...
All the data can be integrated in the wiki page itself (groups, authorizations, synonyms...) or in special dedicated tables (attachments, interwikis...)

In the easiest way to handle the data can be the difference...
August 11, 2006 6:37 PM
 

Ornus said:

you are right, the easiest way to handle data can be different, however I try to make sure it's simple to add a feature if something is missing, or to fix it if it's cumbersome. everything is done through xml, xslt, js. well, I guess that's really similar to openwiking.
macros seem to be similar to sider types, but are much simpler. I don't think there is a way for them to reference each other, include one inside tho other. there's no OO inheritance, where sider types actually are more flexible than classical OO inheritance.

in openwiking user actually modifies macro. I guess there's a way to create an editor that would provide controls to edit macro code directly, but there's no code sharing between views. another view that would edit macro will need to copy/paste or invent it's own way to edit it. openwiking basically supports two views, editor, in which you have to edit text manually, or themed html view to show the page. there's no view to show wiki map for example, it's possible to do it. but not easy.
August 11, 2006 7:32 PM
 

Recent Links Tagged With "wikipad" - JabberTags said:

Pingback from  Recent Links Tagged With "wikipad" - JabberTags

September 22, 2008 5:34 PM

Leave a Comment

(required)  
(optional)
(required)  
Add

About Ornus

Lead Sider and Xelog developer. I'm interested in information and how we can better manage it using computers. I'm also into design and understanding how to creating cool, useful, simple things.
Copyright © 2006-2007 EmeraldHand, Inc. All rights reserved.
Powered by Community Server (Non-Commercial Edition), by Telligent Systems