New Server

[Eliezer Israel]

Had to juggle over to a new server.  

Turns out the AMI I was using wasn't included in the free tier.

Everything should be back online and functioning.

Focusing on enabling general user contributions...

[Theo Armour]

Lev

 

>> Focusing on enabling general user contributions...

 

Each contribution should be a post.

 

People would not usually be happy with strangers mucking around with their personal contributions.

 

When others want to help – they can do so via the comments.

 

What links the posts together is a well-crafted series of categories and tags.

 

Occasionally some people will want to collaborate. They can do that with a Google Doc and then submit the final edit as a new post.

 

Theo

[Yodan Rofe]

Initially, I agree, but gradually as the language is built up, there might be a way to create more definitive edited patterns that are more agreed upon.
Yodan

[Michael Mehaffy]

The Git model (on which Ward's new wiki system is based) is that anything can be coped and re-worked, but the original is retained intact, as are all attributions.  This takes the Wiki model one step farther -- instead of anyone being allowed to nibble at the edges of a project, they can clone the whole thing and re-work it, then pull it (or a portion) back into the master at a later date if that's agreed.  

In Alexandrian terms, that allows people to manage and share wholeness.  In our terms, that is a great fit for "project pattern languages" - you take over a whole set of patterns, then re-work to suit the project.  Some or all can then be taken up by others. 

This gets over a colossal problem that Chris noted with PLs - that they are too generic, too abstract.  You need just this grounded, "R&D" development system - combined with an effective collaborative, open source model.  I think that's the key...

Ward, your thoughts?

Cheers, m

--
Michael Mehaffy

Structura Naturalis Inc.

742 SW Vista Ave., #42

Portland, OR 97205

(503) 250-4449
www.tectics.com

[Ward Cunningham]

Yes, exactly. The patterns are the GATE, our goal is the WAY.

Here are two short presentations I've made on the subject in the last month.

First is one tuned to the general audience of TED where I explain the way wiki redefines linking (twice):

http://www.youtube.com/watch?v=BdwLczSgvcc

This second video, from my keynote at O'Reilly's new javascript conference, describes the evolution of patterns and includes snapshots from Alexander (via this project), EcoTrust, Michael's Mockups and my work on sustainable materials. This addresses federation with emphasis on a new kind of API for data:

http://www.youtube.com/watch?v=mX0M19AqKQw

Finally, CloudSmith has customized their StackHammer product to make a one-page quick start for making federated wiki sites that you can own and control. This will make it push-button simple for anyone to have their own server. We have only to explain how to use this new capability for good. 

Best regards. -- Ward

[Yodan Rofe]

Hi Ward,

Thanks for sharing this. I think I am beginning to finally understand what you're talking about. What it means in effect, is that each of us working on pattern language will have his own version of the language - although probably most of it will be shared. There's something nice about that, almost like a natural language, where each group and even individuals have their own idioms.

I'm intrigued about the metrics and data stuff - I'm not that sure what it means in terms of the pattern language for built environments. I understand that's it's important to keep us all honest, and not write crazy patterns that don't make sense or are damaging. I'm not sure we have the theory to be able to calculate these metrics for all the scales - maybe for certain aspects like energy yes, but less so for human well-being.

Yodan

--

Dr. Yodan Rofe` - Senior Lecturer
Desert Architecture and Urban Planning - J. Blaustein Institutes for Desert Research
Ben-Gurion University of the Negev, Sde Boqer Campus, Israel 84990
Tel. 972-8-6596884  Fax. 972-8-6596881
http://www.bgu.ac.il/CDAUP/yodan-rofe.html

[Michael Mehaffy]

Hi Yodan and Ward,

I think this is the beauty-part regarding metrics:  they get integrated into whole systems.  

The problem with a system like LEED, say, is that the metrics are allowed to run roughshod over the design.  And they're empirically questionable, in part because they're treated so separately.  (Metric[generating factors of course interact with each other in real life, and with the design!)

The same thing is true for economic drivers -- return on investment, and those metrics...   

So coupling these metrics to the design elements, in a way that both can be assembled, transformed and modeled, would represent a great breakthrough.  So we should see if we can pull it off!

I think this is a key goal of Sustasis work, and of my current GHG modeling work.  (Another important but poorly modeled metric - helpful for these purposes because it is, at least, a global metric.)

Cheers, m

[Eliezer Israel]

Sweet to see CA's patterns in your slide-deck, Ward, and to know that our work here helped to put that in there.

I'd like to shift this conversation a notch.  I think we can all sense the potential of a bunch of people working on patterns in this way.  The question is - what are the steps from where we are now to getting a bunch of people working on patterns in this way.  

Last I checked the fed-wiki work was moving along well, but not yet ready for a non-technical user to jump in.  Ward - I'm interested to hear your thoughts on that question.  Also interested in hearing your thoughts on what the right system architecture would be, given this working model.  Should we be thinking of a pattern server that each individual can have a wiki hosted in, and technically apt folk welcome to host their own? 

[Eliezer Israel]

So this is basically there already, and ready to roll.  Working right now on the UI to make it obvious how. 

Questions:

When I post a pattern, can others contribute pieces that show up 'around' the pattern (without the author's explicit approval).

images

places that have the pattern

research

models

etc. 

[Theo Armour]

Lev

 

I have created my first post. The image I uploaded was not saved. Permission error I think.

 

 

>> When I post a pattern, can others contribute pieces that show up 'around' the pattern (without the author's explicit approval).

 

Yes, all they need to do is select the appropriate tags (or category) when they create a post.

[Michael Mehaffy]

Lev - I think it would be terrific to get the technology ready to go open, and build repositor(ies) soon.  From my work on the platform, it seems awfully close.  A few pieces I'm not certain about -- one not a big deal, the other a bigger deal...  (Maybe Ward can comment?)

1) How can we use this platform now/soon, and make it play well with other existing platforms?  Seems that a simple link within each pattern would do that.  That is, these are the condensed patterns, and they link to more discussion and other resources (and even other whole patterns in another format, e.g. the book).  Click on them and it opens a new window, say.  (Ward, is that a capability now?)

2) The other more challenging one is in handling the data.  This is the more ambitious and possibly revolutionary contribution to pattern technology as a whole, and to wiki -- as Ward says, doing for numbers what the earlier wiki did for words.  

And the key breakthrough would be around implementation, as I see it -- always the bugbear of any construction design system, including patterns.  How much does the thing cost to build, how much is the return, are there other costs... what are the regulatory hurdles, how does it conform to specified metrics (e.g. energy code,   GHG emissions, etc)?  

In our experimental model, we modeled how this could work.  But there's a lot more to do!  This is my current focus of research, in fact.

There is a third question, with is simply procedural -- how do we host it, what happens if a million people log on to upload, etc... and again, how much does it cost!?  BTW this is not a proprietary model (though there are plenty of proprietary options for consulting on it, etc). 

Cheers, m

[Eliezer Israel]

Good catch on the bug. You were right.  Upload permissions had to be reset on the new server. 

It's fixed.

Interesting idea on the tagging.  A bit of a different angle than the approach I was taking.  Let me noodle it for a bit.

[Ward Cunningham]

On Jun 7, 2012, at 12:47 PM, Eliezer Israel wrote:

Last I checked the fed-wiki work was moving along well, but not yet ready for a non-technical user to jump in.  Ward - I'm interested to hear your thoughts on that question.  

As an open-source project there are many developers scratching their own itches. 

There is private monies going into ongoing development. That wouldn't happen if they didn't see a future.

Also interested in hearing your thoughts on what the right system architecture would be, given this working model.  Should we be thinking of a pattern server that each individual can have a wiki hosted in, and technically apt folk welcome to host their own? 

There are platform services that will offer Federated Wiki as a menu choice.