Re: [cti] Feature voting and tracking system

["Barnum, Sean D." <sbarnum@mitre.org>]

The intent was not that this “voting” would be the sole arbiter of prioritized ordering.

If you look at the proposed development process for 2.0 that we discussed, agreed to and published a couple of months ago you will see that the intent is that each member review tracker items and then offer their opinion on prioritization “based on importance” using a TBD mechanism (what we are discussing now). These individual opinions would then be considered in aggregation and then factor in other criteria such as cross-dependencies and impacts in order to reach an actual prioritization.

This aggregate prioritization is characterized as:

"

Review set of issues suggested as in scope for 2.0

1. Identify consensus scoping of issues for consideration
2. Identify and eliminate unnecessary duplication between issues (merge/consolidate where appropriate)
3. Identify cross-dependencies and impacts between in-scope issues
4. Identify use cases relevant to each issue
5. Identify consensus prioritization of in-scope issues based on importance
6. Prioritize in-scope issues based on dependence (which issues should be resolved before others as their resolution impacts the others) and importance

“

This “voting” and prioritization technical capability would be used to serve item “a.” in the list above.

I spoke with Aharon and we are both fine if the SC decides that they are okay with Aharon and I doing the above aggregation and prioritization analysis ourselves and then presenting it as a roadmap. Our concern is that all voices are heard and that there is no appearance of bias or favoritism on our part. That is why we have proposed using a mechanism to capture everyone’s opinions in an explicit way. As discussed on the call, one option for the solution is to not have a technical “voting” capability and have the co-chairs establish the roadmap. It should be up to the SC members whether or not they prefer this option over the others.

I certainly agree with "less talk about how to do things, more actually doing things.” That is why I would like to move past talking about capabilities like this and actually decide on which capability to use.

sean

From: <cti@lists.oasis-open.org> on behalf of John Wunder <jwunder@mitre.org>
Date: Thursday, October 22, 2015 at 11:19 AM
To: "cti@lists.oasis-open.org" <cti@lists.oasis-open.org>
Subject: Re: [cti] Feature voting and tracking system

So I have to admit that I still don’t really understand the whole voting thing. What are we using the votes for, prioritization? Or will issues that get a lot of “downvotes” not get addressed?

I think I said this on the call yesterday, but my preferred approach would be for someone (the co-chairs) to lay out a rough roadmap of the issues that we need to address. They can take into account list preferences, dependencies between issues, etc. In particular, they could identify some fundamental issues to talk through first before hitting the specifics. Then they send that roadmap to the list and if anyone wants to add to it or disagrees with it we talk about it.

I worry that if we just rely on upvotes we’re going to tackle things randomly rather than strategically and we’ll end up spinning our wheels. For example, if you look at our previous conversations on relationships we ended up with short diversions to versioning, IDs, markings, and other topics that we probably should tackle separately, first, so that they don’t keep coming up in other discussions.

I do like the idea of threaded conversations. The mailing list is difficult if you miss even one day of a quick discussion. Though it seems like the mailing list + Github is working *OK* (not great, but OK) and I wouldn’t want to spend months figuring out how to switch to Stack Exchange when we could be actually working on STIX issues during that time.

To sum up: less talk about how to do things, more actually doing things.

John

On Oct 21, 2015, at 9:02 PM, Jordan, Bret <bret.jordan@BLUECOAT.COM> wrote:

We really need a way to track features, concepts, ideas and have the ability to vote up / down the ideas.  A solution like this really needs the following key features:

1. Ability to document a concept, idea, or features, similar to what we do today in the issue tracker in GitHub

2. Ability for people to vote up or down the main item

3. Ability to star an the item as a second form of tracking to possibly indicate preference

4. The ability to comment on the top level element just like you can in GitHub

5. Ability to comment on comments. 

6. Ability to vote up / down a comment 

7. Ability to mark a comment as the current train of thought or current consensus of the thread.  This will help people come up to speed more quickly on issues.

In the past I have suggested something like StackOverflow.  Yes, this is usually used for a QnA type system, and that is what it is geared towards.  However, I feel that a lot of the elements that it has could be very useful for us to use.  

Here is an example of a QnA question that shows most of the elements or requirements that I have illustrated above.  See how multiple can comment on the item and multiple comments can get votes, and how it tracks edits to comments.

http://stackoverflow.com/questions/21029174/whats-this-operator

A solution like this would be a replacement for the GitHub Issue tracker.  Meaning, we would not store issues in GitHub anymore.  Further, I could see a lot of the discussion around topics moving out of email and in to a system like StackOverflow. 

Thanks,

Bret

Bret Jordan CISSP

Director of Security Architecture and Standards | Office of the CTO

Blue Coat Systems

PGP Fingerprint: 63B4 FC53 680A 6B7D 1447  F2C0 74F8 ACAE 7415 0050

"Without cryptography vihv vivc ce xhrnrw, however, the only thing that can not be unscrambled is an egg."