Re: [Tails-ux] [review] Onion Circuits and Tor Status extens…

このメッセージを削除

このメッセージに返信
著者: intrigeri
日付:  
To: Tails user experience & user interface design, alan
題目: Re: [Tails-ux] [review] Onion Circuits and Tor Status extension strings
Hi,

first, thanks a lot for spending time on this!

Spencer wrote (10 Mar 2016 17:33:24 GMT) :
> I don't understand the difference between Onion Circuits and Tor Status [#11161].


I've tried to explain this on
https://tails.boum.org/doc/anonymous_internet/Tor_status/

Tor Status is the software that handles the onion icon in the "system
tray".

Onion Circuits is the software that shows you Tor circuits and
connections. That is, the window whose title is "Onion Circuits".

Clear enough?

> • Unless there are plans for more information in the dropdown, the Onion Circuit
> click would be more effective if it opened the dialog instead of presenting an option
> to do so in a dropdown. The softkeyboard functions like this, so it seems okay.


This was discussed in the "How to replace the green onion [was: What
do we miss to replace Vidalia]" thread in March last year
(https://mailman.boum.org/pipermail/tails-dev/2015-March/).

A few arguments were raised in favour, and against, the menu.
sajolida and I were undecided, so Alan (who was doing the actual work)
picked the menu option.

> • The log fills up nicely but the info disappears. It would be great to have control
> over this, if not only to have enough time to read and understand.


What do you mean with "the log"?

Onion Circuits is a real-time view of the _current_ status, not a log.

> • Path section shows both circuits and log info. Could be either/or, or emphasize
> the difference between with color coding.


I'm sorry, I have no idea what you mean with "log info" :(

> • There are no details when clicking on the path, as the instructions suggest we
> expect. I am unsure if this is the case with some log items but all items would
> benefit from having details of some sort.


I think I've seen reports about that when using Tor bridges.
If you were using a bridge: can you retry without using a bridge?

> • Access to an explicit log would be great.


It's not clear to me what you mean here. Please clarify what use case
this would address, what value this would add, and then we can check
if someone wants to write+maintain the corresponding code, and how
exactly it should work :)

> • It is not clear what the one hop paths are. They eventually disappear.
> • Some are 2 hops. These also eventually disappear.


My initial reaction is: these circuits are parts of how Tor works.
Onion Circuits shows how Tor works. So you get them.

But really, I'm not sure what we should do.

Should we display only 3 hops circuits that are about some proxied TCP
connection, to match the understanding of Tor that many people have?
Or should we keep doing what we do, which is more accurate, and more
useful to some people perhaps, but more confusing to some others?
Is it worth spending coding time to change the current state of things?

> • Maybe country could be the path labels, moving the router name to the details
> section where country is.


Why?

> • Regarding the subtitle of the dialog, the difference between circuits and streams
> isn't clear.


I'm afraid that's Tor's nomenclature :/ ... and indeed that's exactly
what Onion Circuits does ("Display Tor circuits and streams").

Maybe we should hide this precise wording, and instead find something
more vague and perhaps less confusing because it'll carry less
technical information?

OTOH telling the user exactly what we're showing might empower them if
they want to learn more: they'll know exactly what they are seeing.

> • Status labels aren't clear: Extended?


I believe this is essentially the same topic as the 1-hop /
2-hops issue.

> • Maybe circuit visualization should accompany 'Open Onion Circuits' option in
> the dropdown.


This was proposed, and then rejected for security & technical reasons
in the aforementioned thread. Let's come back to it, if you want,
after you've read it.

(I'm sorry I can't point you to a summary of the arguments, since
nobody wrote down any such thing back then.)

> • Why are there these circuits? How are they determined? Can these be manual or is
> there value to automatic potential path generation?


>From a UX PoV I don't know what I can do from these 3 questions of

yours, so perhaps I should reply with a question: why exactly does it
matter in _this_ discussion?

If it's just that you are personally curious, I'm not aware of
not-too-technical, simple answers to these questions, written down
anywhere. So at this point, either you trust Tor (the software) to do
the right thing because it's Tor™ and we rely on it anyway, or you'll
need to learn more about it and dive into its design. IIRC Tom Ritter
has published a PDF that is very helpful to anyone who wants to
understand more.

Cheers,
--
intrigeri