Re: Atlas Core - needs examples

[Ron Jeffries]

Hi Mark,

As you know, most discussion of the Agile Atlas is intended to be on the Agile Atlas group, because the Atlas is not a trainers and coaches exclusive thing. If you choose to rejoin, you'll be welcome. 

I'm copying this reply to AgileAtlas and plan to engage in further discussion there, where everyone who is interested can see and no one who is uninterested is forced to see. I hope everyone will join us in the intended forum: it makes it much easier when we have a single repository of ideas. Of course, people can discuss anything in TCC that they want to ... but the "official" discussion forum is the Agile Atlas one -- because that is what people here and on the site forum requested, and we built it for them.

Summary

• More examples would be good. Someone needs to write them.
• Changing the Core is a bit of a big deal as the test is based on it, and by design it is short. In addition, to change the Core itself, we must constitute a "Core Update Group" who will have that responsibility. We would not like to do that lightly.
• We do not have actual information on the extent to which people are or are not confused by the Core. The Core is not all of Scrum, and Scrum is not easy. It would be good to have better information to go on.
• There are other locations on the site where examples might go, GASPs and Commentary.
• There are tools built in, including categories, tags, and links, that could be useful in linking examples into the site.

Details follow:

On Apr 15, 2013, at 11:51 AM, Mark Levison <ma...@mlevison.com> wrote:

The latest discussion around Vertical Slices and Shippable illustrates the issue that many people have reading the atlas as they try to understand Scrum. The Atlas is full of abstract ideas. Even the experts sometimes get confused with abstract ideas. When we have concrete examples points become more clear. For beginners this an especially big issue (see: http://www.infoq.com/articles/science-of-learning for details).

The core document may indeed be abstract. We believe it is no more nor less abstract than the Scrum Guide. Its intention thus far has been to map very closely to the Scrum Guide, without infringing any copyrights thereof, and to remain stable enough that the Scrum Alliance can base its tests on the Core.

As it stands the core document is at best a document for experts. If we want to make it accessible to people new to our game we need to provide some examples to help illustrate the points.

How can we add these?

Sections of the Atlas

As you know, since you have provided material, the Atlas has sections beyond the Core. GASPs, in which your article with Charles appears, is about Generally Accepted Scrum Practices. It is an excellent place for specific examples, such as yours.

There are also a number of "primer" articles, under Commentaries. The original plan, you may recall, was to have descriptions of Scrum by many authors, the sum of which would give readers a broad understanding from many viewpoints. That resulted in too much material to expect people to read it all. In addition, the executive decision was made that the Scrum Alliance required a "Core" page, which we duly created. 

The "primer" articles remain. Some are abstract, some are more concrete. If I recall, one of them is even a little story about Scrum in action. We have talked about breaking those into smaller topical articles but that would require that the authors agree to such treatment. We do not look forward to that.

Core Updates Are Not a Simple Issue

Neither GASP nor Commentaries address your concern completely, but I think they can partly address it. As for the Core itself, there are some issues.

First, it's intended (and required) to be short enough to be required reading for new CSM candidates, and needs to remain stable enough to be consistent with the test. There is, at present, no process figured out for keeping the test and Core in sync, other than mostly not changing them. That's not very good, but it's where we are.

Second, earlier editions of the core had many examples. Strong voices here objected to the examples as being irrelevant to some readers some of the time. Those objections had merit, and -- we think -- resulted in a better document. But this does mean that adding examples to Core is likely to create yet another flurry of only partly productive activity.

Third, there is no official process for updating the Core. The intended plan is to select a small group of experts who will convene meetings -- some of them physical and requiring a day or more of time -- to hash out Core changes. I am only partly optimistic about that. I think the group can surely do the job, but I expect that the TCC will then bubble up some small number of people who will again seriously object and make a decision difficult if not impossible. If the consensus of people on the Atlas group is that we need a core update, or if the Scrum Alliance requests one, we'll form that group and get them working. Up until now we have felt no need for an update and have been waiting for Ken and Jeff's shoe to drop to see what needs to be done. This subject -- examples -- may well be enough to trigger the update group. It would certainly be OK with Chet and me.

Examples Exist on the Atlas, and More Are Welcome.

More facilities exist on the site than are presently used.

Even today, authors such as yourself and others can readily submit articles consisting of examples, just as your GASP had examples of user stories. We could put those into Commentaries or GASPS, of course. 

In addition, we also have categories and tags in Expression Engine, so it would be straightforward to create categories or tags for examples of various kinds, and to add a search capability to find suitable articles. "Core Examples", perhaps, or something more specific. 

We could also add links directly into the Core article. One way would be the usual highlighted link sort of thing, that would link off to a single example or to all the examples on whatever subject. There is also the ability to float a margin comment that could say something like "For examples of User Stories, click here". 

Material Seems to be the Biggest Issue

The system has the flexibility to do any of these things. What we really need is the material. When we get material, we put it up as quickly as we can, and when it comes in a format that isn't too difficult, it now takes very little time at all. As you'll perhaps have noticed from the Kanban boards, we expect some articles but have none in process, so as soon as we got something, we could put it up.

A Meeting Might Be Useful

Perhaps it would be useful for interested people to get together at the Gathering to try to come up with a coherent view of what should be done. Chet and I do not plan to call such a session but if there's interest we certainly would. The technology is ready, the site is ready. Mostly we just lack material, which people need to write.

Summary

See top.

Ron Jeffries

www.XProgramming.com

Wisdom begins when we learn the difference between "that makes no sense" and "I don't understand". -- Mary Doria Russell

[Mark Levison]

I honestly didn't know I was still on the atlas list. A few brief notes below

On Mon, Apr 15, 2013 at 12:54 PM, Ron Jeffries <ronje...@acm.org> wrote:

Summary

• More examples would be good. Someone needs to write them.

You just wrote one on the trainers list about an hour ago. In many cases two sentences will get the point across.

• Changing the Core is a bit of a big deal as the test is based on it, and by design it is short. In addition, to change the Core itself, we must constitute a "Core Update Group" who will have that responsibility. We would not like to do that lightly.
• We do not have actual information on the extent to which people are or are not confused by the Core. The Core is not all of Scrum, and Scrum is not easy. It would be good to have better information to go on.

Scrum is hard to do but fairly simple to understand. The current format of the core has confused attendees to my classes. Since I assume attendees to my classes represent the norm, its safe to assume that many are confused. In the case of "shippable" even some trainers were confused by what we were trying to say.

• There are other locations on the site where examples might go, GASPs and Commentary.

The problem is that most people read only the core. If you put the examples anywhere else the people who need them the most will miss them.

I get the idea of writing separate primer articles, but I don't think they will get read. You've got the traffic figures for atlas but I suspect that less than 10% of visitors do more than see the home page and core.

As to writing one myself - you made some changes in the submission guidelines in the recent past around acceptable tooling (i.e. Word/Pages bad; Markdown good). I gave up at that point.

Cheers

Mark

[Ron Jeffries]

Hi Mark,

This email, and the last, is going to both lists …

On Apr 15, 2013, at 1:36 PM, Mark Levison <ma...@mlevison.com> wrote:

I honestly didn't know I was still on the atlas list.

I don't know either and thought you were not, owing to your "giving up" as you put it below. From our side, a good idea, or a good article, is worth having. And we prefer to do the Atlas work on the Atlas list, which was requested by people far and wide.

A few brief notes below

On Mon, Apr 15, 2013 at 12:54 PM, Ron Jeffries <ronje...@acm.org> wrote:

Summary

• More examples would be good. Someone needs to write them.

You just wrote one on the trainers list about an hour ago. In many cases two sentences will get the point across.

No doubt. As a rule, I've tried NOT to be a major author on the site. It's supposed to be a community site. I've been rethinking that a bit but would prefer that the community get actively engaged to make it be what they want, rather than stand on the sidelines casting in ideas and the occasional stone.

• Changing the Core is a bit of a big deal as the test is based on it, and by design it is short. In addition, to change the Core itself, we must constitute a "Core Update Group" who will have that responsibility. We would not like to do that lightly.
• We do not have actual information on the extent to which people are or are not confused by the Core. The Core is not all of Scrum, and Scrum is not easy. It would be good to have better information to go on.

Scrum is hard to do but fairly simple to understand. The current format of the core has confused attendees to my classes. Since I assume attendees to my classes represent the norm, its safe to assume that many are confused. In the case of "shippable" even some trainers were confused by what we were trying to say.

That report is interesting, but I don't know how to act on it. I don't know what else might have confused them. I don't know if the Scrum Guide would have confused them more, had they been exposed to it first, or less. I would anticipate that the more people read, the less they'll be confused, but this is not always true.

Test passing numbers suggest to us that the site isn't really very confusing. We certainly don't want it to be confusing and would like to make it better.

NOTE: We were required, last time, to have absolutely nothing in the Core to which even a single CST would object, saying that he or she would not use the document. That was an intense pain to accomplish, and resulted in a kind of averaging down that has produced what we have now.

To add things to the Core is to reopen the door to a similar process, which created much heat and shed very little light. It will also require us to convene and kind of "constitutional committee" to work out the update. My personal view is that updating the core with examples will not be cost effective. I would lean toward links or the like.

The Core does follow a different outline from the Scrum Guide, and maybe that makes it harder, though I'd not have thought so. It's different intentionally: the Scrum Alliance wants its own document and we desire to honor the Guide, and to follow it, but not to steal from it.

• There are other locations on the site where examples might go, GASPs and Commentary.

The problem is that most people read only the core. If you put the examples anywhere else the people who need them the most will miss them.

I get the idea of writing separate primer articles, but I don't think they will get read. You've got the traffic figures for atlas but I suspect that less than 10% of visitors do more than see the home page and core.

That may be true. It wouldn't surprise me. However, people who are going to just do that are doing the absolute minimum and it is no surprise that they will exit that activity still uninformed. 

As to writing one myself - you made some changes in the submission guidelines in the recent past around acceptable tooling (i.e. Word/Pages bad; Markdown good). I gave up at that point.

I noticed that and have discussed that in more detail in a separate email directly to you. I will mention here that we have set no required submission guidelines at all, so unless freedom is a problem, giving up may not be the right response. We did discuss at some length the fact that different input formats have substantially different costs.

We certainly want to do the right thing. I believe that the right thing is not intuitively obvious and is not as simple as one might hope. I'd love to be wrong about that.

Ron Jeffries

www.XProgramming.com

You never know what is enough unless you know what is more than enough. -- William Blake

[Mark Levison]

In Las Vegas I commit to working with anyone who volunteers to create a couple of examples of how this might work.

Specifically what I imagine is either sidebar boxes or callouts that provide examples. This would clearly link the examples to the text but would also allow experts to skip them easily.

Off to run a very small business.

Cheers

Mark