Opened 6 years ago
Closed 21 months ago
#1087 closed task (fixed)
Compile list of open reseach questions
| Reported by: | str4d | Owned by: | jaruga |
|---|---|---|---|
| Priority: | minor | Milestone: | 0.9.35 |
| Component: | www/i2p | Version: | 0.9.8.1 |
| Keywords: | docs | Cc: | slumlord |
| Parent Tickets: | #1085 | Sensitive: | no |
Description
Tor has several lists of research ideas and open questions. We should have a similar list.
Tasks to complete:
- Collect the many research questions listed in the existing documentation.
- Read our minimal existing literature.
- What questions do they try to answer?
- What answers to they provide?
- What new questions do they pose?
- Think about what areas of I2P need more research.
- What changes have we made to I2P recently?
- How do those changes affect the performance/anonymity tradeoff?
Subtickets
Attachments (3)
Change History (13)
comment:2 Changed 6 years ago by
| Priority: | major → minor |
|---|
comment:3 Changed 4 years ago by
| Milestone: | → soon |
|---|---|
| Priority: | minor → major |
| Status: | new → open |
comment:4 Changed 22 months ago by
| Owner: | set to jaruga |
|---|---|
| Status: | open → accepted |
This is something I will gladly tackle over the next couple weeks, having a compiled list would certainly be valuable for increasing research participation.
Changed 21 months ago by
| Attachment: | research_v1_2018.04.18 added |
|---|
V1 draft of research page - no questions
comment:5 Changed 21 months ago by
Update: I've almost finished going through the documentation and a lot of the relevant research papers, the list is looking quite nice. I imagine it'll be done in the next week or so.
I've also written a first draft of a general outline for researchers who choose to work on the I2P network and some safety ethics that it's recommended they follow. I've attached it here. If anyone had time to give it review while I finish compiling the accompanying questions it would be greatly appreciated.
comment:6 Changed 21 months ago by
We actually have a list already - it's here http://i2p-projekt.i2p/en/research/questions
WRT research guidelines page which is what jaruga attached, we have one of those also: http://i2p-projekt.i2p/en/research
neither has been updated in a while
comment:7 Changed 21 months ago by
Attached merger of existing research page and additions from draft v1.
Changed 21 months ago by
| Attachment: | open_research_questions_2018.04.30.txt added |
|---|
Proposed additional research questions
comment:8 Changed 21 months ago by
| Cc: | slumlord added |
|---|---|
| Milestone: | soon → 0.9.35 |
| Priority: | major → minor |
All looks fine to me.
These may be pages that we start updating regularly, so no need for perfection.
If you don't get a response from str4d in a couple days I suggest you proceed by working with slumlord to get the changes in. You two can figure out how to do that, either manually or by you generating a patch against the actual website files in mtn or github (get them from https://github.com/i2p/i2p.www/tree/master/i2p2www/pages/site/research ) - but not a github PR because it's just a mirror.
comment:9 Changed 21 months ago by
edcc8f079eb91933ef321cab7a4454bbeac39db4 - #1087 - Update 'Academic Research' page
and
5f0e9227fc62a9f8fda733591afe4b5fa6122551 - #1087 - Update 'Open Research Questions' page
Thank you, jaruga.
comment:10 Changed 21 months ago by
| Resolution: | → fixed |
|---|---|
| Status: | accepted → closed |
Yay 
No problem. I'll mark this as resolved for now.
I'm sure we could find some with searches (I may look at this later) but one came to mind. Unidirectional tunnels are a design choice seemingly not backed by any research papers but could be beneficial.
Grothoff's paper decided unidirectional tunnels were a bad idea with little evidence to support this. A page in our docs exists for it: http://www.i2p2.i2p/unidirectional-tunnels.html
As with many of the research questions we have, I'm sure research can and will benefit both ourselves and the field. It's possible unidirectional tunnels were a positive change and Tor may, some day, want to use them if the research is positive.
Relevant zzz.i2p post: http://zzz.i2p/topics/905
A brief mention of another: profiling.