CZ Talk:Proposals/Subgroups in addition to Workgroups?

For historical background, or more detailed discussion, there is a forum thread on this topic titled sub-workgroups.

Are subgroup Editors' names and Authors's names automatically found and listed?

Chris, I have not yet digested all of the proposal, but I want to congratulate you on having finally made the proposal ... as you know, I have been "pushing" for this for almost a year. So please accept my thanks.

I do have one question. As I recall, and I may be wrong ... I think that I manually entered my name in Chemical Engineering subgroup's Category:Chemical Engineering Editors and I manually added the 4 authors' names into Category:Chemical Engineering Authors. Will that now be done automatically? Will the names be picked up somehow from users' pages?

Right now, users' pages only list the main workgroups for which each user wants to be an author or editor. Is it going to be up to each user to seek out what subgroups for which he/she wants to be listed as an author or editor at the bottom of his/her user page so that info can be automatically picked up?

Milton Beychok 07:29, 17 January 2009 (UTC)

The section that deals with this issue is How_to_invite_your_colleagues. I think you might misremember adding the names manually to the category page (or I misunderstand you). I believe what you did was add the subgroup category to their user pages. In short, your last paragraph is exactly what I had envisaged. Maybe you could rephrase the section in the proposal to make more sense? Chris Day 16:17, 17 January 2009 (UTC)

Other thing is I myself would rather be and editor at Botany subgroup only than to the whole Biology scope, thus is this option being contemplated or an editor must be listed all the way down? Dalton Holland Baptista 14:53, 28 February 2009 (UTC)

Dalton, at this point, there is no proposal to have workgroup-specific editors. I don't think it wise to spread ourselves so thin. I think that you should apply for editorship in Biology. We could always use another active participant like yourself. Russell D. Jones 19:44, 28 February 2009 (UTC)

From Main Proposal Page

I think its a good idea, but might be unnecessary considering how inactive we are. After all, I'm one to talk, haven't been around here properly since last May. Its definately something that could be used in the future, even as a cousin of the Wikipedia project pages. Denis Cavanagh 15:28, 23 January 2009 (UTC)

Denis, I agree with your assessment. I think the one good reason for doing this now is that it is good to have the mechanism in place for when its ready to be used efficiently. Once it is in place we will be able to fine tune it so it is really ready for prime time. Also, we do have a few users who are already making good use of it. The clear example is chemical engineering. In that case i think it could also be used as a recruitment tool. I might well set up a genetics version too, if nothing else to organise some of my thoughts with respect to topics for class. Chris Day 16:19, 23 January 2009 (UTC)

I agree wholeheartedly with Chris that it would be good to have this mechanism in place now for the reason he gives above and for all the reasons that have been presented in the lengthy Forums discussion of this subject. Milton Beychok 17:25, 23 January 2009 (UTC)

My apologies for barging in! Unaware a discussion had taken place on the forums. Chris offers a very persuading rationale and I do support this. Denis Cavanagh 19:34, 23 January 2009 (UTC)

No apologies needed. In fact, I should link to the relevant discussion in the forum, that makes a lot of sense. Chris Day 19:39, 23 January 2009 (UTC)

Regarding the redundancy question: it seems to me that redundancy would be controlled by the editors. They would have the responsibility to reject redundant subgroups. I'd also like to see some threshold of need met, say two (three?) editors from each group approving the subgroup before creation. Russell D. Jones 02:21, 5 February 2009 (UTC)

Russell, I agree with you that Workgroup editors should control what subgroups need to be created. However, as matters stand now, it is difficult to find two or three active editors in some workgroups. For example, the Engineering Workgroup has only two active editors at best. One of the reasons for creating subgroups is that it might encourage more editors to participate. But for the time being, I suggest that one or two active (and I emphasize the word active) editors be all that is needed to endorse the creation of a subgroup. Milton Beychok 03:29, 5 February 2009 (UTC)

In the section above #Which subgroups are needed? we write that "Editors decide which subgroups are relevant". Given the discussion above why don't we make this more specific, as suggested by Milt above. I think his suggested requirement for active editors is a valid one. Just to clarify though, Milt suggests that two editors should be required to start a subgroup whereas in this proposal it would be two editors are required to endorse a subgroup. I know this is a subtle difference but i think there is quite a significant distinction. When i wrote this i was thinking that anyone could start a subgroup and then, if it was seen to have a useful role, the better ones would "win" endorsement from editors. My reasoning here is sometimes it is not clear what we need and thus a test period might be desirable to see how it works out. Just a thought. Chris Day 17:55, 23 February 2009 (UTC)

Reworking Overall Workgroup Policy Pages

I'd like to see most of this page (once this policy is adopted), minus the discussion, moved to CZ:Subgroups. Should there also be a subgroups page started like CZ:Workgroups or could that be handled by a category page? CZ:Workgroups right now is just a list of workgroups. I think CZ:Notes on launching workgroups should be moved to CZ:Workgroups because that is where the policy on workgroups should be. Then CZ:Subgroups could be branched from there (e.g., "See also CZ:Subgroups"). Russell D. Jones 14:20, 23 February 2009 (UTC)

Name: Subgroup or Subworkgroup?

Should the name be subgroup or subworkgroup? In the end I favoured subgroup since it was shorter but maybe subworkgroup is more descriptive? Chris Day 17:55, 23 February 2009 (UTC)

Subworkgroup is more descriptive and accurate as it is clear that it is an under-level of a workgroup. But subgroup is shorter. CZ does not have "Groups," as far as I know, other than "Workgroups," so there wouldn't be any confusion. Maybe someday "Newsgroups," "authorgroups," etc., which becomes an argument to reserve "subgroup". Once choice is made, though, nearly impossible to change without bot. I'm okay with "subgroup" too. I'll concur with Milt on this and leave it to you. Russell D. Jones 15:23, 24 February 2009 (UTC)

OK, I'll go with subgroup. One of my reasons for switching from subworkgroup originally was that there is an implication (and misconception in the discussion on the forum) that a subworkgroup is a subset of a specific workgroup. Given the emphasis on multidisciplinary interaction we want to encourage between workgroups, subgroup might be a less confusing name. Chris Day 04:24, 25 February 2009 (UTC)

Next Step

I think this prop is mostly baked. Chris, you've listed "Implementation" as the next step. Does this mean that no CZ decision-making body need sanction this proposal? Russell D. Jones 14:20, 23 February 2009 (UTC)

My mistake. I guess I meant everything is in place for implementation. I assume this will need to go for a vote. Chris Day 14:31, 23 February 2009 (UTC)

Should we send it to the ed council? Russell D. Jones 17:41, 23 February 2009 (UTC)

I think it is almost ready. Let's just finalize the name and the role of editors and rewrite the proposal to reflect that consensus (see my two comments above). Then it is ready to got to the EC. Chris Day 17:55, 23 February 2009 (UTC)

I have no preferences. Subworkgroups is a bit more cumbersome but it is more descriptive. I leave it to you, Chris. As for the role of editors, whether they start or endorse (after a test period) the subgroup (or subworkgroup), I still think is should be two "active" editors ... and I hope we don't get involved with trying to define "active" editors. Milton Beychok 18:27, 23 February 2009 (UTC)

Formal deletion/approval of a Subgroup?

(undent) Name: up to Chris. Role of editors: To (1) sign on to approve subgroup (add names to template); (2) one editor (at least) from each covering workgroup; (3) one of the approving editors will place subgroup in some sort of subgroup list (like the article approval system, somebody has to go and change article status; we should have same or similar procedure for subgroup approval). (4) We should also have a policy for subgroup elimination (e.g., how is a subgroup destroyed; should be harder than creating a group). But this can be later amendment to policy. Russell D. Jones 15:37, 24 February 2009 (UTC)

Certainly there could be an approval subpage to document the affiliation endorsements from editors (CZ:Chemical Engineering Subgroup/Approval, CZ:Chemical Engineering Subgroup/Affiliation or similar). As to an approval-like process, this could be done using a metadata page similar to clusters but would such a formal process be required? Or could it be managed on the approval/affiliation subpage or even the subgroup talk page. Possibly the proceedure should involve an announcement on the workgroup mailing list with x days for objections to be lodged with regard to an affilitation? If there are two editors in agreement then the subgroup will become affilitated with the workgroup. As to delisting a workgroups affiliation from the subgroup, would this not just be a reverse of the same process?

As to deleting a subgroup that does not flourish, i.e. no chance of affiliation with any workgroup, I'm not sure what the process would be. Maybe the best solution is cold storage if no workgroups show an interest after 3 months? Recruitment from cold storage is always possible IF editors from a particular workgroup wish to endorse a formal affiliation with the defunct subgroup. Chris Day 04:57, 25 February 2009 (UTC)

Sub-subgroups?

(undent)   What about Sub-Subgroups? e.g. {{Subgroup|History of Biology|History of Science|History|Biology}} or {{Subgroup|History of Chemical Engineering|History|Chemical Engineering}}. Plus, I just saw another problem. No "Science" workgroup. So how do we create a "history of science" subgroup? Russell D. Jones 15:37, 24 February 2009 (UTC)

In my view History of Science does not need to be a subgroup of science. I would envisage it being {{Subgroup|History of Science|History|Biology|Chemistry|Physics}} Now comes the problem, we only have four slots for the affiliated workgroups, clearly more sciences will want to be affiliated with such as subgroup. So maybe we need more? Maybe there should be no limit?

With regard to sub-subgroups, I had never considered this need. Thinking about it a bit I'm not sure there is a need. Using your example, why not {{Subgroup|History of Chemical Engineering|History|Engineering}}? I'll think about this some more. Chris Day 04:36, 25 February 2009 (UTC)

I think that incorporating sub-subgroups into this proposal would be biting off much more than we can chew and would probably create a good bit of opposition. Let's get the subgroups accepted first and leave sub-subgroups to be considered at some future date after we have digested the subgroups. Milton Beychok 05:29, 25 February 2009 (UTC)

I think these are wise words. Also, in time we may well realise it is not necessary. Chris Day 05:45, 25 February 2009 (UTC)

Right, I was just exploring possibilities. It can always be re-addressed at a later date. Russell D. Jones 13:12, 25 February 2009 (UTC)

Part of the problem here is that all these ideas imply strict hierarchy. Data base software stopped being hierarchical a long time ago, first becoming relational (think table-oriented), then more exotic relationships allowing multiple inheritance, semantic relationships, etc. One of the problems is that this may be trying to force hierarchy, certainly a very understandable concept to people, farther than it really can go in data structuring. Even in something as seemingly structured as library catalogs under Dewey Decimal or Library of Congress, each cataloged object still can be in multiple subject hierarchies (i.e., on the catalog record), an author index, etc. This isn't a solution, but might help understand the problem. Howard C. Berkowitz 04:36, 3 March 2009 (UTC)

Certainly we can over think this. In my view the best think it to let it develop on its own. We might be surprised at the relationships that develop. Chris Day 04:54, 3 March 2009 (UTC)

I agree; let's get some experience with workgroups before we think about changes to this aspect of the policy. Russell D. Jones 14:12, 3 March 2009 (UTC)

Chemical Engineering Example

To log editor approval, I assumed that there would be someplace where the approving editors would log their approval. In the Chemical Engineering example, the subgroup points to the Chemical Engineering article. It would seem then that the approving editors should (would) also work on the subgroup main article and get that through approval. Does this mean then that in order for the subgroup to be approved, the subgroup main article has to be approved? Russell D. Jones 15:40, 24 February 2009 (UTC)

Good idea (lead article should be approved), this might well be the carrot that leads to a strong cohesive group and subsequent workgroup affiliation as opposed to a less sound idea. Chris Day 04:51, 25 February 2009 (UTC)

Ah-ha. So is this then the proposed approval process? Russell D. Jones 13:14, 25 February 2009 (UTC)

Policy Proposal

Chris & Milt, I've taken the proposal and re-written it as proposed policy. The way that I have crafted the resolution means that the "action" that the ed council will take upon approval is to move that proposal page to the CZ namespace as approved policy. The "policy" is different from the "proposal" in that they are directed towards different audiences: the proposal towards the ed council, the policy towards all citizens. Thus, there are items in the proposal (i.e., to persuade the ed council to adopt) that are not needed in the policy. I realize that this may create some redundancy (!), but I hope that we can edit the policy and get the mechanics and language in an approvable form. Thanks. Russell D. Jones 14:42, 25 February 2009 (UTC)

Subgroup Approval

We had discussed the approval process as having at least one editor from each workgroup approve the subgroup as a means to get it going. We have also discussed that the subgroup should begin writing the subgroup main article and moving that towards approval too. Given the approval process, this means that it takes at least five editors to formally approve a subgroup (two [assuming a minimum cross-sectioned subgroup] editors to create the subgroup; three more to approve the main article [assuming the two creators worked/authored the main article]. Is this too much?
  In practice, I expect that subgroups would be created and main articles started but not approved. (I note also that CE was approved with only two editors.) This is acceptable, I think. Thoughts? Russell D. Jones 13:05, 27 February 2009 (UTC)

I agree we do not want too many editors to get a subgroup approved. The subgroups main article could actually be approved by one editor IF that editor had no involvement in writing it. If one other editor agrees that the subgroup is a good idea then it could be affiliated to the workgroup too. So theoretically only two are required. I think the maximum of editors required would be three in the scenario where editors are also active in writing the Subgroups main article. It sounds as if you are thinking that the editors that approve affiliation need to be distinct from those that approve the subgroups main article? I had not thought of it that way. Chris Day 15:10, 27 February 2009 (UTC)

I really don't see any benefits to "approval" except in the prestige of being an "approved" subgroup. Anybody can create subgroups, affiliate articles to subgroups, write articles under the subgroup category. There really isn't any benefit to being "approved." I think the reason we came down this road was that we were trying to figure out a way that editors could exercise editorial control over redundant subgroups. Russell D. Jones 15:04, 28 February 2009 (UTC)

I have never really thought about any subgroup being approved. I think we agree on this point. I do not really equate affiliation with approval. On the other hand if no workgroup wants to be affiliated with a particular subgroup their activity is probably doomed. But lack of affiliation does not actually prevent their existence or their activity. Chris Day 04:52, 3 March 2009 (UTC)

Subgroup or Workgroup

Chris, here's another practical matter. Suppose I create an article in the Chemical Engineering subgroup and am about to fill out the metadata. Would Chemical Engineering go in as Cat1, Cat2, Cat3 or would it go in Sub1, Sub2, Sub3? What's the difference? If there is no difference, is it then possible to affiliate an article to six workgroups? Russell D. Jones 13:25, 27 February 2009 (UTC)

A subgroup could be one of three, Sub1-3. Only workgroups names can be used for the Cat1-3 fields. At present there is a maximum of three workgroups and three subgroups per cluster. It would not be catastrophic if they got mixed up but the subpages template would not function correctly in placing categories. Chris Day 14:59, 27 February 2009 (UTC)

I'm not following. Does the subpages template reference the Sub1-3 fields but not the Cat1-3 fields? Russell D. Jones 14:32, 28 February 2009 (UTC)

The hyperlinks in the subgroup header are based on the sub fields not the cat fields. Subgroups never use the cat fields. For example, the 'recent changes' link will track edits to any page in a cluster that has the relevant sub field. The subgroups 'All Articles' category will only show articles that have a relevant sub field. Links go to the Chemical Engineering examples. Chris Day 04:16, 3 March 2009 (UTC)

Affiliation

I just added a new section outlining my interpretation of our discussions to date with regard to affiliation (See at /Subgroups and at /Subgroups in addition to Workgroups?). See what you think. Chris Day 15:40, 27 February 2009 (UTC)

I think it's okay. I wonder about the last paragraph though. I'm not sure exactly what "cold storage" is. Does it need to be defined here? Would the policy work just as well without a discussion of what happens to dead subgroups? Russell D. Jones 14:37, 28 February 2009 (UTC)

Another problem I just realized about "de-affiliating" a subgroup is that there will be a bunch of articles tagged as belonging to the subgroup that will point the article to a dead subgroup. Does de-affiliation mean that all of the affiliated articles' metadata sub1-3 fields be cleared? Sounds like a job for a bot.

Do we really need to address de-affiliation right now in this policy? I know we've discussed it some, and it's good that we have devoted some thought to it. But I'm thinking now that perhaps some experience with subgroups would give us a better handle on the issues associated with de-affiliation. Or is there a more compelling reason that I'm missing why we should have, at least, an initial de-affiliation policy? Russell D. Jones 14:59, 28 February 2009 (UTC)

One more thing, then I'll stop. I see now that we are contemplating three situations where editors should step in to control subgroups. First is where authors have created redundant subgroups, second is to approve subgroups, and third is where the subgroup has died. In the first case, the editors should work towards some sort of consensus or compromise whereby the redundancy can be eliminated and articles affiliated to the agreed-upon subgroup. The second case we've pretty much ironed out. In the third case, the editor would have the subgroup "archived in cold storage." In practical terms (I guess I'm repeating myself now), what does "archived in cold storage" mean? I would prefer that if the content is good that no article or subgroup be deleted; and going through a bunch of articles just to change a field sounds tedious. Russell D. Jones 15:15, 28 February 2009 (UTC)

I've since found CZ:Cold Storage. Russell D. Jones 19:04, 2 March 2009 (UTC)

Good point with regard to depopulating the sub fields, if a subgroup loses all its affiliated workgroups that would be a lot of work (although I presume a bot could be programmed to perform such a task if needed). My rationale writing that was that we only have three subgroup fields that can be used for each cluster, so the sub1-3 fields might become prime real-estate. Thus, if a subgroup is no longer affiliated with any workgroups, then another subgroup should have a green light to substitute a dead subgroup in any clusters metadata. I was imagining that the cold storage option would be that green light. In practice we could just have a rule that a non affiliated subgroup does not have priority with regard to be used in a clusters metadata. Chris Day 03:49, 3 March 2009 (UTC)

Who can start a subgroup?

This is an interesting difference in perspective I found while editing the proposal. I removed the following text from the section on starting a subgroup:

"Once at least one editor from each affiliated workgroup (or two from one workgroup) have agreed to sponsor the subgroup, one of those affiliated editors should complete the subgroup creation process. To do this, the sponsoring editor should..."

My original idea was to allow authors to start any Subgroup they wished to encourage creativity but to have oversight from editors with regard to affiliation. Affiliation is the key since any Subgroup without affiliated workgroups is not "official". If we go with the text above though, we are asking the authors to wait for editors before they can start a subgroup.

While I can see there might be a lot of unofficial subgroups out there if we give authors this much room, I still think it is worth allowing more freedom. We could always have a category that is added to a subgroup with no affiliations and purge them every so often (place into cold storage). As long as the authors understand that without affiliation this is a risk I think we will not have a problem. Thoughts? Chris Day 16:19, 27 February 2009 (UTC)

Yes, yes! I like the idea of having authors initiate subgroups. As a bottom-up approach this would encourage more involvement. So:

1. Any citizen can create a subgroup by creating the pages under subgroup creation; thus authors can affiliate the subgroup with workgroups.
2. Approval of subgroup affiliation is done by affiliated workgroup editors on the Subgroup/Approval subpage.
3. Subgroup approval is not a requisite for affiliating articles with the subgroup (i.e., entering the subgroup in the article's metadata sub1-3 fields).
4. Further legitimation of the subgroup occurs when the Subgroup Main Article is approved by affiliated editors.

Russell D. Jones 14:52, 28 February 2009 (UTC)

In my scenario I was not imagining that the authors would add the workgroups to the template and await for workgroup approval. I was thinking they would advertise their subgroup to the workgroups they wish to be affiliated with at the workgroup homepages. If the workgroups like the idea then editors would then added their workgroup to the subgroup. The act of the editors adding the workgroup would be the "approval". How about the following steps?

1. Any citizen can create a subgroup by creating the pages under subgroup creation
2. Legitimation of the subgroup can occur when the Subgroup Main Article is approved.
3. After this main article is approved subgroup members may canvas at workgroup homepages to request affiliation with said workgroups.
4. Affiliation of a subgroup to a workgroup is documented by workgroup editors on the CZ:Subgroup/Approval subpage and finalized by adding their workgroup to the subgroup template.
5. Affiliation to a workgroup is not a requisite for adding their subgroup to the sub1-3 fields on any clusters metadata page.

I have removed comments to the effect that subgroups will be approved since I think this might be misleading. They can still function without approval. Chris Day 04:07, 3 March 2009 (UTC)

Chris, I see #3 "After main article approval ..." too high a bar. Subgroup participants should start the main article and work towards approval. They should also start and work on other articles in the subgroup field. I do not see the reason why approval of the main article should preclude the participants in a subgroup from publicizing their group or requesting affiliation from workgroup editors. What am I missing here? Russell D. Jones 14:32, 3 March 2009 (UTC)

Nothing, this is a legitimate point and I can go either way. Chris Day 18:27, 3 March 2009 (UTC)

Summary of agreed points

1. Any citizen can create a subgroup by creating the pages under subgroup creation. Agreed.
2. Affiliation to a workgroup is not a requisite for adding their subgroup to the sub1-3 fields on any cluster's metadata page. Agreed.
3. Legitimation of the subgroup can occur when the Subgroup Main Article is approved. Agreed. (whatever "legitimation" means; I know it was my term.)
4. Affiliation of a subgroup to a workgroup is documented by workgroup editors on the CZ:Subgroup/Approval subpage and finalized by adding their workgroup to the subgroup template. Agreed.

Making sure now the proposed policy reflects this agreement. Russell D. Jones 14:32, 3 March 2009 (UTC)

Still thinking

Russell I do plan to reply to your comments above. Still thinking about the various situations you brought up. Chris Day 15:54, 2 March 2009 (UTC)

Subgroups as managing workgroups

Consider this idea: if there is a sizeable enough subgroup, it should be a managing workgroup. So there would be no need for "subgroup" fields in the metadata.

The idea of piling on groups and groups of people all of which claim responsibility for an article makes me very nervous. --Larry Sanger 20:03, 2 March 2009 (UTC)

Would that not complicate the hierarchy of the workgroups? Just for the record, i do not envisage subgroups as having any responsibility for the articles they choose to tag. If that is implied in the proposal i think we need to change the wording or make it more explicit. In my view subgroups have a collaborative role not an editorial role. Subgroup will not have editors, although they will have members who are editors in workgroups. Chris Day 20:11, 2 March 2009 (UTC)

Nor do I believe this would be a case of the tail wagging the dog. There are no subgroup editors; subgroup articles would not be approved without workgroup editors. Subgroups are mostly navigational; a place where like-minded collaboration can occur without "distraction" from all of the other possible topics within a given workgroup. I also think we should have experience here as well. Will subgroups grow to workgroup size? Don't know yet. Let's find out. Russell D. Jones 14:21, 3 March 2009 (UTC)

Pursuant to this discussion, I have merged the sections "Which subgroups are needed?" with the section "Affiliation with workgroups." The "Affiliation with workgroups" section now explains in one place what the role of workgroup editors are regarding subgroups. I have also removed the following text below as Chris's idea about priority seems to mitigate the need for a policy about deactivating subgroups or archiving them. More legitimate subgroups have priority over less legitimate subgroups. Russell D. Jones 15:05, 3 March 2009 (UTC)

Text removed:

"Any Subgroup that cannot find editors willing to endorse an affiliation with a workgroup, or any Subgroup that falls out of favor (a subgroup that formerly had affiliations with workgroups but currently has none) will be archived in cold storage. This can occur after a six month period with no affiliated workgroups. This process can be reversed, however, if active editors wish to revive the subgroup at a later date."

I see another angle to Larry's comment above, which is, "if subgroups become large and active enough shouldn't they become managing workgroups in their own right." Sure, then the subgroup should go through the workgroup proposal process. I think the activity of the subgroup should weigh heavily in favor of the ed-council promoting a subgroup to workgroup status. But does that really need to be made explicit in this policy at this time? Russell D. Jones 16:26, 3 March 2009 (UTC)