Re: [MLS] MLS: the WG name should include "group"

[Nadim Kobeissi <[e-mail]>]

I do not believe the name should be changed:

1. MLS is a protocol that is equally suited for pairwise messaging as it is for group messaging
2. The MLS name is elegant and mirrors TLS.

I would recommend not spending a lot of time on this.

Nadim Kobeissi
Symbolic Software • https://symbolic.software
Sent from office

> On Mar 24, 2018, at 4:31 PM, Suhas Nandakumar <[e-mail]> wrote:
> 
> 'mls' seems to be a fine name as it indicates e2e security for messaging layer of the stack. Group vs 1:1 is a realization of an application context on top of such a abstraction layer is what i feel.
> 
> ./S
> 
> 
> 
> On Sat, Mar 24, 2018 at 8:25 AM, Raphael Robert <[e-mail]> wrote:
> We looked at a number of alternatives to MLS and in the end none of them seemed to be substantially better than “Messaging Layer Security”, but that doesn’t mean the name cannot be changed..
> 
> In my opinion “messaging” implies groups in 2018 and it’s not a special case of messaging anymore. It's the norm for non-E2EE products, and a number of E2EE messengers support it as well already today: WhatsApp, Signal, Wire, etc. In the context of MLS, “groups" are also interchangeable with “multi-device”, which is also becoming the norm.
> 
> The 1:1 pairwise mode is still important technologically, but application-wise it will most likely be perceived as a special case of (group) messaging..
> 
> Also just my 2c.
> 
> Raphael
> 
>> On 24 Mar 2018, at 15:05, Katriel Cohn-Gordon <[e-mail]> wrote:
>> 
>> +1
>> 
>> I have no strong feelings regarding the name, but I do think we are working on groups because you need them for general messaging.
>> 
>> Katriel
>> 
>> 
>> On Sat, 24 Mar 2018, at 1:43 PM, Jon Millican wrote:
>>> Thanks for raising this point. I'm also no good at naming, but agree that there's potential scope to name more clearly. A couple of points in response to the specific issues you raised though:
>>> 
>>> While most of the discussion has focused on groups, my feeling is that this is because group is a more challenging superset of the pairwise case; and we should still care about pairwise messaging. In the pairwise case I think the main differentiator compared with TLS is the requirement for asynchronicity; so if we're clearly carving up the territory it might make sense to highlight this.
>>> 
>>> Just my 2c though. I have no objection at all to changing the name, and certainly agree that group support is a crucial aspect that could be worth highlighting.
>>> 
>>> 
>>> Jon
>>> 
>>> On 24 Mar 2018, at 13:15, Shivan <[e-mail]> wrote:
>>> 
>>>> +1 the name seems to be inspired by TLS but the drafts concern themselves explicitly with group messaging, not messaging as a whole.  
>>>> 
>>>> On Sat, Mar 24, 2018, 11:28 AM Daniel Kahn Gillmor <[e-mail]> wrote:
>>>> Thanks for all the work that has gone into the proposed MLS architecture
>>>> and protocol drafts already.
>>>> 
>>>> As we move toward working group formation, i'm a little bit concerned
>>>> that the ostensible name of the project ("message layer security")
>>>> doesn't adequately describe the most salient features.
>>>> 
>>>> We spent most of the time at the BoF in IETF 101 London talking about
>>>> the design of the group key agreement scheme and the strong
>>>> cryptographic properties we want it to provide -- and i suspect that
>>>> most of the work in any formed WG will center around that problem.
>>>> 
>>>> So it seems odd to me that the word "group" doesn't appear in the name
>>>> "MLS".
>>>> 
>>>> I'm no good at acronyms, but i'll just throw out two alternatives:
>>>> 
>>>>   MSG: Messaging Security for Groups
>>>> 
>>>>   MESSAGE: Message Encryption and Security Standard for Asynchronous Group Environments
>>>> 
>>>> I hope since the WG isn't yet formed, it's not too late to consider a
>>>> change to the name.  We ought to be up front about the group-specific
>>>> context that we're talking about here, as we already have a security
>>>> layer that works for messages passing between two endpoints
>>>> synchronously (TLS) and it'd be nice to clearly carve up the territory.
>>>> 
>>>>         --dkg
>>>> _______________________________________________
>>>> MLS mailing list
>>>> [e-mail]
>>>> https://www.ietf.org/mailman/listinfo/mls
>>>> _______________________________________________
>>>> MLS mailing list
>>>> [e-mail]
>>>> https://urldefense.proofpoint.com/v2/url?u=https-3A__www.ietf..org_mailman_listinfo_mls&d=DwICAg&c=5VD0RTtNlTh3ycd41b3MUw&r=M0CVEJydBVUX_bvEqMa84Q&m=q5KP1AJUu-4ir-39yrQWELe1gULZ7UQbevfDBrgr0ss&s=jlPCTk9fCDAvKt6l2DuFKYLJoRLEL_KXV7lPHhtlA3M&e=
>>> _______________________________________________
>>> MLS mailing list
>>> [e-mail]
>>> https://www.ietf.org/mailman/listinfo/mls
>> 
>> _______________________________________________
>> MLS mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/mls
> 
> 
> _______________________________________________
> MLS mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/mls
> 
> 
> _______________________________________________
> MLS mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/mls