Good evening, I am thrilled to announce the launch of my website http://www.officeofgreatideas.com . I do hope that some of you will find time to scope it out and submit a few ideas (I welcome computing-related ideas and knowledge, but please also consider submitting political ideas, and the empirical info that qualifies them!). The site is a simple front end for a database - the goal is to store and organize information in ways that will allow people to make structured arguments with a minimum of redundancy. To this end, I have created eight different types of data, each with a couple unique properties, so that users can use the most appropriate tools when they attempt to convince someone of something. Basic instructions - first create an account. The 'create' link at left allows you to submit data (text, hyperlinks, what have you). After submission, confirmations show up with a link that says 'draw' - click this to add the entry to your drawing board. When you click 'save', all the items on your drawing board are associated with each other. Example - You are reading a book and discover something interesting. Paraphrase it and submit it as an argument. Then create the author and the book, and put all three on the drawing board, and save. Now, whenever someone views that argument, they will see that it is related to that book and author, and vice versa. I plan to update the documentation that is available on the site later tonight. For now, there is already somewhat detailed info up there (if a bit dry). There is, however, an as-yet undocumented feature which I will explain here for those who wish to read on. In addition to a drawing board, each user has a clipboard. This is for wrapping 'old' information into your 'new' ideas without explicitly associating your new ideas with the existing 'old' stuff. Items on the clipboard are not associated en masse like items on the drawing board are; rather they are associated only with the items on the drawing board that are of the same type as them. So, arguments on the clipboard are associated with arguments on the drawing board, and not with books on the drawing board. Specifically: a book on the drawing board will gain a footnote for each book on the clipboard (suggesting that the clipboard books were cited in the drawing board ones) a category on the drawing board will become a subcategory of each category on the clipboard a list on the clipboard will be added to a list on the drawing board a message on the clipboard will be referenced by a message on the drawing board an argument on the drawing board will 'cite' each argument on the clipboard these relationships are dealt with in this way because they are not symmetric. (author-author and hyperlink-hyperlink relationships, conversely, are symmetric so they cannot be put on the clipboard) I hope this is not confusing; I used the words clipboard and drawing board to remind the user that the clipboard is for old things that need not be changed, and the drawing board for new ideas that rely on those old ones. Hopefully this will not be lost on my users. As this is an open source project, my source may be downloaded from http://www.officeofgreatideas.com/app/ I could use a few suggestions [one pressing issue is a bug in the create controller that fails to store files that I attempt to upload]. Please feel free to publicize this site to other communities. My need for a fast take-off is more pressing at this point than my need for a semi-private test period. -Mike
on 2005-11-18 01:01
on 2005-11-18 01:07
Interesting site/idea and all but ruby related?... Whats the net comming to when someone with a yale email address commits spammery?
on 2005-11-18 01:13
On 11/17/05, Lyndon Samson <email@example.com> wrote: > Interesting site/idea and all but ruby related?... > > Whats the net comming to when someone with a yale email address > commits spammery? Mike forgot to mention that this is a Ruby on Rails application which has the full source available. He demoed it at the first meeting of new_haven.rb this month, and it was quite interesting, if a bit in need of continued development. We might even be hacking on it a bit more in coming new_haven.rb mentions. Shame on mike for not pointing out the interesting RoR code that lies beneath! :) http://www.officeofgreatideas.com/app/ This is the source. Hopefully he'll get it onto a SVN repos soon.
on 2005-11-18 01:16
On 11/17/05, Gregory Brown <firstname.lastname@example.org> wrote: > We might even be hacking on it a bit more in coming new_haven.rb mentions. not sure if I meant meetings or functions, but I didn't mean mentions. By the way... there is more information about the core of this application here: http://newhavenrubyists.org:2500/mainpage/show/KnowledgeWeb
on 2005-11-18 02:43
In article <email@example.com>, firstname.lastname@example.org says... > The site is a simple front end for a database - the goal is to store > and organize information in ways that will allow people to make > structured arguments with a minimum of redundancy. Interesting! Have you thought about ways to structure counterarguments, subarguments, debatable points, conclusions, etc.? Maybe a way for readers to "rate" arguments as fallaci.. fall... as containing a fallacy? I remember doing some net-research on this a few months ago, and discovered that there was a whole field of structured debate research, but that the software involved was heavy, crufty, and too complex for people to actually use to debate topics.
on 2005-11-18 20:53
I have spent a fair amount of time pondering my structural approach for contradiction, iff, translation, etc. My current plan is to add a table that stores the two args and an int for the type of relationship between them. But perhaps a good look some of this structured debate research would be a smart plan. Did you notice any algorithms or data structures that stood out as insightful? I think simplicity is probably quite important in a system designed to scale very large and 'judge' things with a subtle touch. -Mike
on 2005-11-19 00:58
On Sat, 2005-11-19 at 04:51 +0900, Mike Schwab wrote: > I have spent a fair amount of time pondering my structural approach for > contradiction, iff, translation, etc. My current plan is to add a > table that stores the two args and an int for the type of relationship > between them. But perhaps a good look some of this structured debate > research would be a smart plan. Did you notice any algorithms or data > structures that stood out as insightful? I think simplicity is > probably quite important in a system designed to scale very large and > 'judge' things with a subtle touch. Just so you're not disappointed later on, be aware that propositional logic has been shown to have serious limitations. Now, just because it's seriously deficient doesn't mean it isn't still useful (the perfect being the enemy of the good...), but there are definitely limitations to what you can do. You'll see similar issues raised concerning the Semantic Web in general (by the way, you really should be looking at the RDF data model for what you're doing). -mental