Re: [hrpc] review drafts with guidelines on human rights considerations

["Giovane C. M. Moura" <[e-mail]>]

Hi,

I was having this conversation in private with Niels and Corinne, and he
asked me to maybe have this discussion on the list. So I'm forward it
here, and will answer his response in a following email.

Best,



Begin message.
From: Giovane Moura
To: Niels ten Oever, Corinne Cath
Date: Fri, 3 Jun 2016.


Dear Niels,

Thanks for reaching out.

I just went through your draft and, wow, congratulations on it, also on
its comprehensiveness and clarity.

Some feedback:

Section 1:

  * I guess it would be nice to have a phrase alerting about the values
that are implicitly embedded into a protocol.

In other words, engineers have their own set of values which they, often
without much thought, bring to their protocols, as it if was just a
given. This may have unintended/unforeseen consequences -- namely human
rights violations.

I've seen this first hand: I had to do a exercise like that for my
dissertation, by going through my school's ethics commission. In the
end, the ethicist (Aimee) and I wrote a case study paper on value
sensitivity design (VSD) and Internet security, and its consequences[1].
I was introducing bias, prejudice, unfairness with my approach, while
proving other values at the same time.

So what I mean is: engineers may unintentionally violate HR and values
while being oblivious to that, and this may be worth mentioning. I do no
know, however, how often this happens on IETF.

Relevance to the draft:

 - My draft can be found at [2], and when I tried to apply the questions
from 5.3 on it, and some of them didn't seem to apply to it.

And that's where I have another question: are those guidelines targeted
to protocols that assume a real human using it? E.g.: e-mail, chat, etc.
I.e.: protocols that are use to transmit data by anyway related to human
communications.

I am asking because our protocol is a control protocol, designed as a
SOS mechanism to be used by one server under attack (DDoS) to signal to
another servers that the attack is happening.

Therefore, it does not carry any metadata/data related to human
communication. If it is not human communication related, does it imply
that it does not violate human rights? I am not so positive on that
either...

For example: BGP is also machine to machine but if *used* poorly, can
and has been used in censorship cases (see Pakistan Youtube case[3]).

Could you please share your thoughts on this  so I can carry on here
with the analysis?

Thanks a lot,

/giovane


[1] http://doc.utwente.nl/87095/

[2]https://www.ietf.org/id/draft-francois-dots-ipv6-signal-option-00.txt

[3]
http://www.cnet.com/news/how-pakistan-knocked-youtube-offline-and-how-to-make-sure-it-never-happens-again/




> I hope this e-mail finds you well. As discussed in Berlin, Corinne and I
> worked hard to addressed all brought up issues in the
> hrpc-research-draft. You mentioned you would be interested in reviewing
> a draft your currently working on in the light of the guidelines for
> human rights considerations.
>
> I would like to kindly ask you to:
>
> 1. Take your draft
>
> 2. Read through the guidelines here:
> https://tools.ietf.org/html/draft-tenoever-hrpc-research-02#section-5.3.2
>
> 3. Make notes about your findings, especially relating the following
points:
> 	- Relevance for your draft
> 	- Changes you will make to your draft
> 	- Perhaps even a considerations paragraph that could be added to your
> draft.
>
> 4. Perhaps write a short evaluation paragraph in which you answer the
> following questions:
> 	- Was this a useful exercise
> 	- What can be improved?
> 		- Examples
> 		- Number of questions
> 		- Granularity of questions
> 		- Did going through the questionnaire feel like 'time well spent'?
> 	
> All your thoughts are very welcome.
>
> Best,
>
> Niels
>
> PS If you preper to talk about this rather than do a write up, am also
> more than happy to schedule a call with you.
>
>