Re: [dhcwg] WGLC: draft-ietf-dhc-duid-uuid-02.txt

[Thomas Narten <[e-mail]>]

Hi Thomas.

Thomas Huth <[e-mail]> writes:

> 1) You're neither referencing RFC 2119 nor using the keywords from that
> RFC. However, I also didn't see a lot of sentences in your document where
> you need it really urgently, so you might get away without it. Anyway,
> please consider using it, e.g. in the following sentences:

> - "DHCP UUIDs should be persistant across system restarts ..."

> - "Implementations of this specification using DUID-UUID must select a UUID
> that is persistent across system restart..."

Personally, I prefer not using 2119 language unless there is a
compelling need. I.e., to make very clear that not doing something is
a violation of the spec.  I don't see a strong need in this
document. But if others think this is necessary, I'll make the change.

> 2) You might want to add a sentence like this in the beginning of your
> document, to make clear where the definitions come from:

> Terminology specific to IPv6 and DHCPv6 are used in the same way as is
> defined in the "Terminology" sections of [RFC3315].

Will do.

> 3) In the "Background" section, you explain the previous DUIDs again:

> "Three DUIDs types have been defined previously:

>    DUID-LLT -  the Link-Layer address of one of the device's network
>            interfaces, concatenated with a timestamp

>    DUID-EN -  an Enterprise Number plus additional information specific
>            to the enterprise

>    DUID-LL -  the Link-Layer address of one of the device's network
>            interfaces"

> In my opinion, this is too much detail here, you can simply refer to
> RFC3315 instead, e.g.

> Three DUIDs have been defined previously: DUID-LLT, DUID-EN and DUID-LL
> (see RFC3315 for details).

I could, but then the reader would need to go look at 3315 to
understand what info they contain. And since I talk about the how the
existing definitions are inadequate, I think it's better to include
the detail.

> 4) In some sentences, you're using a lot of "would"s ... that sounds
> somewhat uncertain to me, and that seems unappropriate for an RFC to me.
> For example:

> In some environments, a UUID-based DUID would be preferable to the other
> existing DUID types.

> Please consider whether you could use a more direct language in those
> sentences, e.g.:

> In some environments, a UUID-based DUID is preferable to the other existing
> DUID types.

I will do this.

> 5) In chapter 4, you've labeled the figure with "DUID-UUID format. ." - I
> think you can remove the second (or even both?) dots here. I also doubt
> that you need the "Figure 1" label here, since you never reference this in
> the rest of the text.

Removed the extra ".". The "figure 1" is generated automatically by
xml2rfc and may (I'm not sure) be required as part of the general RFC
format.

Thanks for the review!

Thomas

> Mit freundlichen Grüßen / Kind regards,
>    Thomas Huth
>                                                                            
>     IBM Deutschland                 Vorsitzender des Aufsichtsrats:        
>     Research & Development GmbH     Martin Jetter                          
>     Schönaicher Str. 220            Geschäftsführung: Dirk Wittkopp        
>     71032 Böblingen                 Sitz der Gesellschaft: Böblingen       
>     Tel.: [phone]          Registergericht: Amtsgericht           
>                                     Stuttgart, HRB 243294                  
>