Category talk:AWG Design Document

• Any danger of actually discussing what you're doing here Dzon? The sections are no longer editable after what you've done. --Morgaine Dinova 13:09, 21 October 2007 (PDT)
• Please talk to us Dzon, don't just unilaterally turn the wiki into some random HTML junk. When we press Edit, we're meant to be able to edit, not get dumped into some obscure piece of HTML magic. This is what the Talk page is for, to warn of major changes and get them agreed. Or should I revert? --Morgaine Dinova 13:39, 21 October 2007 (PDT)

Wiki sprawl

It appears as though many of these pages are sprawled out versions of existing content. For example, we now have a AWG Agent Domain and an Agent Domain page. Why a whole new set of pages here? -- Rob Linden 18:24, 23 October 2007 (PDT)

I did make a note about it here: AWG_Design_Document_Template#Goals. Also see, Zero's office hour transcripts: User:Zero_Linden/Office_Hours/2007_Oct_18 (~ 7:48 and further). You'll notice his concern with the AWG articles. I started to use Brainstorming as a first candidate to organize discussions by interest to an aspect of the model. I also wanted to make sure we could separate the previous model with the AWG model, so that terminology does not get confused as easily. Perhaps, I didn't make that part with the actions clear. With the example, Agent Domain included all topics about the Agent Domain, which include the Agent Stores, Agent Hosts, Agent Services, and much more. Each one if specified in detail on a single page can create a really long page, as we started to see when an older version of Brainstorming (v35936). It appeared to sprawl out at length over many topics and subjects even though it was a single page. I noticed a lot of the same interest touched on in various areas on the entire page and also on other pages. There was no way to just pull up a wiki page that relates to a more concise interest and find a flow of discussion about it. Also, I constantly saw the concern that people wanted to jump into some form of implementation. The design documents, with the basic template followed, uses the normal wiki abilities to navigate to different interest of the model and subscribe to that part of discussion. Further, I didn't want to obstruct other documents from being more general or plain presentational; the AWG Design Documents carry more of an intention to disambiguate views, boldy. Dzonatas Sol 19:11, 23 October 2007 (PDT)

First, a document template is deeply not the place to propose a completely new process, especially when there is a page, linked directly off the AWG main page, which discusses process, and is marked tentative, and invites discussion on the topic. Posting such a process, in extreme detail, with none of the material marked as "proposed" or "tentative" or "suggested", but rather having the appearance of a consensus document is frankly hostile to forming consensus.

Second, having two sets of pages, which in detail, essence and purpose attempt to describe the same basic concepts, is deeply disruptive to collaboration. Having a template document which suggests a flow through the documents which is alternative to the main entry to the wiki, and leads to pages almost entirely duplicative of the main pages, is simply hostile to forming a single perspective.

If you don't agree with a page in the wiki, on a topic, the normal approach is to post a comment on the talk/discuss page for that topic. Please follow this practice. Forking discussion, and forking it in such that it forms a parallel tree is counterproductive.

Lastly, making it harder to edit pages only makes it harder to hold a discussion. I expect to be able to press the edit tab on a wiki page and edit the page. I expect to see a talk/discuss page that's accessible off the page being edited. I do not expect to have to look into templates to find documents which contain meaning, rather than navigation. Please talk to the group, ideally on the talk pages of the wiki, before making major, complex changes to the wiki structure. -- Zha Ewry 18:08, 24 October 2007 (PDT)

I have to say I agree with Zha here. Dzonatas, can you make sure you build some consensus before creating whole new structures? Where can we move this material so that it's clear it's largely your opinion rather than group consensus? -- Rob Linden 18:23, 24 October 2007 (PDT)

I hate to say, but I feel Zha's comments are a bit hyperbolic about the process in regards to the design documents. Mainly because it does not seem there is much room to propose ideas equally among all individuals. If some informally (like face-to-face) make a consensus and stand on it as the main weight to all other further actions, that would be severely be against ideals of Working Groups (IETF wise). It is obvious that we need to lean more towards a genuine consensus. There appears to be more face-to-face consensus than a Working Group based consensus. If we look at the given VAGs for example, there is barely any members in each to really qualify as more than a face-to-face consensus. As may be noticed here in the logs and changes throughout the wiki (and AWG transcripts if they get posted), some of the members of those VAGs approach others as they are the only way to make consensus. I agree that better consensus needs to be built all around, but I also impress that on other lateral documents as well.

I looked over the IETF and a few other articles to review some thoughts. I noted a couple of things that are being missed by this Working Group, and one is that the mail-list is mandatory as to make more genuine consensus. I propose a secondary mail-list that is more directed at the AWG discussions and decisions to strictly build consensus and to make other announcements. I'd like to use SLDev, but there are members of the AWG that are very biased against SLDev. I feel the real answer to where some of this material belongs is the mail-list, and either those who are biased against SLDev have to sacrifice and use that mail-list, or we will need to create a new mail-list only for Working Group business (not to be limited by who can join, except maybe they need a premium account?)

With these documents more immediately, I intend to mother them for awhile to make sure there is a good example structure of them. They proven not to be completely useless. =) The context most likely is based on my opinion, but the content less than largely my opinion. I've based the content off of a rough consensus of views from several sources, and they should be looked at as examples (even as some "running code") till there is actual consensus (like on the mail-list).

Once there is more genuine consensus, and at that time it is still seen as just my opinion then consider them utterly useless. Until then, I believe the notions made above are more premature. That certainly does not mean I ignore requests. If you can't trust that, then just say it that way. I'm going with your encouragement, Rob, (at the open source meeting) about to make ideas succeed, so does this mean "User:Dzonatas Sol/*" immediately? Or, would it hurt to propose the "Proposal:" namespace? Would that make it clear enough? Dzonatas Sol 23:54, 24 October 2007 (PDT)