Re: [PATCH v2 2/2] Documentation: sketch the threat model on the front page

[Valentin Gagarin <valentin@gagarin.work>]

On 6/22/26 15:26, Alyssa Ross wrote:
> Valentin Gagarin <valentin@gagarin.work> writes:
> 
>> Also capture the poking-holes security model and hint at the trust
>> boundary.
>>
>> This way of phrasing it intentionally doesn't go into any detail.
>> Proper exposition in dedicated documents would still be fruitful,
>> especially for readers less experienced with the domain.  But this way
>> we already mention all the most important ideas in the first few
>> paragraphs, which should be more likely to capture attention of experts.
>>
>> Signed-off-by: Valentin Gagarin <valentin@gagarin.work>
> 
> Good addition.
> 
>> ---
>>   Documentation/index.html | 7 ++++++-
>>   1 file changed, 6 insertions(+), 1 deletion(-)
>>
>> diff --git a/Documentation/index.html b/Documentation/index.html
>> index 6d4f07b..e1c5ebb 100644
>> --- a/Documentation/index.html
>> +++ b/Documentation/index.html
>> @@ -21,9 +21,14 @@ the principle of security by compartmentalization.  It aims to have a
>>   lower barrier to entry and to be easier to use and maintain than other
>>   such systems.
>>   
>> +<p>
>> +By isolating each application in a virtual machine, Spectrum prevents untrusted
>> +software from accessing what you haven't explicitly allowed or disrupting other
>> +applications.
>> +
>>   <p>
>>   Persistent data will be managed centrally, while
>> -applications remain isolated. This means that the system can still be
>> +preserving boundaries between compartments. This means that the system can still be
> 
> Do you think it's already understandable by this point what a
> compartment is, to a reader that is learning about security by
> compartmentalization for the first time?

My gut feel is it will be alright, because the intuitive notion of 
compartment maps good enough to what's really going on.  Open to 
suggestions for further improvement of course.