| In addition to Jason;s request I still havenât heard whether the metadata writeup that was written will be merged/added to the spec to help explain all the metadata properties we have in the spec.
We seem to be going in the wrong direction as far as getting close to being complete on the spec. We might need a special session that everyone can make to discuss and resolve the issues. Unfortunately the meeting times havenât worked out for me and given that Marlon wasnât able to make the call today either then we might want to consider a specific call next week to resolve.
Allan On Nov 1, 2022, at 3:08 PM, Jason Keirstead <Jason.Keirstead@ca.ibm.com> wrote:
Hi folks - wondering if someone can link to the detailed proposal being discussed? As Stephanie has recently left IBM I am trying to get back up to speed on CACAO and Allan's argument is concerning.
This Message Is From an Untrusted Sender You have not previously corresponded with this sender. Desiree - This proposal effectively still makes the feature required.
Most, if not all, playbooks will like have commands and if we are stating that a command must also define the meta-data that describes what the command is doing then this effectively makes the entire feature required.
Secondly, this is absolutely the worst possible scenario of all because it is equivalent to requiring every single shell command within a shell script have to have a functional description of what that command is doing *in addition* to the actual command itself.
This proposal results in extremely verbose playbooks with absolutely every single command having to have a descriptor in addition to the actual command itself.
I am strongly opposed to this proposal as it will result in making playbooks bloated, overly cumbersome to define and effectively will result in playbooks not being defined or used at all.
I suggest that all playbook metadata be optional and let the market/authors decide what is practical and reasonable to document. This is exactly the same as what occurs with other scripting languages. People comment on commands and scripts based on what they want to convey. They donât do it on every single command as itâs not required and would slow the development of the scripts down to a crawl.
Allan
Regarding the required/optional issue â as discussed on todayâs call, weâd like to propose the following compromise: - The playbook_functionalities property on Playbook is optional - The playbook_functionalities description includes a normative "SHOULD" (i.e., âThe values for this property SHOULD come from the playbook-function-type-ov open vocabulary.â) - The function property on Command Data is required Dez On the required/ optional issue, I favor optional. I hear Allanâs arguments and I resonate with allowing new users to crawl before running. If it's that valuable the market will reward the optional feature. Remember customers can require optional features. It just means not everyone requires it in all cases. Plus making required would be a breaking. Just my 2 cents. iPhone, iTypo, iApologize
Bret -
a) The writeup that I did on meta data remains open on adding to the spec or not as far I can tell from the last call.
b) The proposed feature on functionality is not yet completely specified. Whether we make it required or optional, it is missing a lot of the meta-data attribution/classificaiton information that the feature allows a user to specify. So at this very moment the proposal is not ready to add to the spec regardless of required/optional.
c) Adding this functionality feature brings in a lot of work upon playbook authors to define such functionality instead of focusing on the mechanics of defining a playbook in the 1st place. From my perspective it is a significant increase in the burden to do for every playbook and Iâm not convinced all use cases will rely on it or want to have to do it. Therefore I was willing to support its inclusion in the spec (which it HAS NOT BEEN APPROVED) provided it is optional. However, if the proponents insist that this feature be required then I will say that the entire feature SHOULD NOT BE ADDED to the spec at all.
Iâm a firm supporter of proving value in the marketplace and making the barrier to entry for using playbooks very low and easy. If an optional feature (any of them) gets used reliably and well in the marketplace then the outcome is the same.
The idea that a spec can force people to do something they donât want to do is a fallacy. And increasing the complexity of CACAO playbooks even more than they already have is a mistake that will only result in the entire project being undermined.
Allan
> On Oct 27, 2022, at 10:23 AM, Bret Jordan <jordan.oasisopen@gmail.com> wrote:
>
> All,
>
> We have very few remaining work items that need to be addressed before we can ship the next version of the CACAO specification. We really need people to speak up and voice their opinions on the following:
>
> 1) MITRE/DHS has proposed a new property to track some of the features of a playbook. Several people seem to like this and support this. However, there is some concern on if this property should be "required" or if it should be "optional". Allan strongly views this should be optional. MITRE/DHS really want it to be required. But I have not heard from the rest of you.
>
> Please respond to this email with your views on this subject.
>
> 2) A while back Allan proposed the attack target types. I think we all generally agree that this is a good idea. However, when I have played around with it, it feels like there are some minor issues with the modeling. I have not yet been able to put my finger on it. To be clear I really like the idea of the Attack stuff that Allan proposed and really want it. But something just feels off with it. Has anyone else seen issues while playing around with it?
>
> Bret
>
---------------------------------------------------------------------
To unsubscribe from this mail list, you must leave the OASIS TC that
generates this mail. Follow this link to all your TCs in OASIS at:
https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
|