Notices by 死 (firstname.lastname@example.org), page 2
死 (email@example.com)'s status on Tuesday, 06-Dec-2016 23:26:38 UTC 死 #goals
Kaiser (firstname.lastname@example.org)'s status on Tuesday, 06-Dec-2016 23:23:04 UTC Kaiser Fuck, I forgot to mention. My aunt? Literally a witch. She even goes to these secret meetups somewhere and does tonnes of research on wicca.
- In conversation from activity permalink
死 (email@example.com)'s status on Tuesday, 06-Dec-2016 23:25:19 UTC 死 @shpuld yes. delicious, screaming legos.
死 (firstname.lastname@example.org)'s status on Tuesday, 06-Dec-2016 23:22:59 UTC 死 finally found some shit worth building
死 (email@example.com)'s status on Tuesday, 06-Dec-2016 23:20:05 UTC 死 there is only flow
死 (firstname.lastname@example.org)'s status on Tuesday, 06-Dec-2016 23:19:49 UTC 死 4:20
死 (email@example.com)'s status on Tuesday, 06-Dec-2016 16:37:57 UTC 死 Mission Improbable: Hardening Android for Security And Privacy https://blog.torproject.org/blog/mission-improbable-hardening-android-security-and-privacy
死 (firstname.lastname@example.org)'s status on Tuesday, 06-Dec-2016 16:29:00 UTC 死 Understanding and Hardening Linux Containers https://loadaverage.org/url/3340169
死 (email@example.com)'s status on Tuesday, 06-Dec-2016 16:26:53 UTC 死 Large-scale cluster management at Google with Borg https://research.google.com/pubs/pub43438.html
死 (firstname.lastname@example.org)'s status on Tuesday, 06-Dec-2016 16:25:58 UTC 死 Linux Container Security https://linuxcontainers.org/lxc/security/
- In conversation from activity permalink
Mikael (email@example.com)'s status on Tuesday, 06-Dec-2016 07:55:17 UTC Mikael Join my project of replacing google. My idea is to make a local cache of stuff I have viewed that will be searched before google. An anonymized search to google will help complete the results.
The idea is to selfhost an elastisearch db with html documents, contacts, locations, mail and chatlogs. That db will serve a browser search provider which will search local stuff first and when getting bad matches, complete the search from google.
It will be an android app that hijacks some data and disables some google spy settings.
死 (firstname.lastname@example.org)'s status on Tuesday, 06-Dec-2016 07:58:59 UTC 死 @golbette maybe consider anarchism then. https://www.gutenberg.org/files/23428/23428-h/23428-h.htm
死 (email@example.com)'s status on Tuesday, 06-Dec-2016 07:57:01 UTC 死 אוּרִיאֵל
死 (firstname.lastname@example.org)'s status on Tuesday, 06-Dec-2016 07:54:30 UTC 死 # Ntrack: The Network Tracker protocol
Spec Version: 0.1.1
Ntrack is a simplification and generalization of the bittorrent HTTP tracker protocol.
The aim is to produce a more lean and streamlined protocol that removes bittorrent specific aspects and can be used for any number of peers to associate IP:PORTs to a given key (info_hash in the case of Bittorrent).
The main goals are:
- Simple: retain the simplicity and transparency of textual data, and further simplify the data structures and URL scheme.
- Lean: as bandwidth and computationally efficient as reasonably possible.
- General: Should be useful for any task that requires a bunch of decentralized clients to publish their connection information.
- Backwards compatible: we don’t want to break (too much) existing Bittorrent clients.
Note: this effort almost completely unrelated to the new UDP bittorrent tracker protocol.
- BT: Bittorrent tracker protocol.
- NT: Ntrack tracker protocol.
- Key: An arbitrary UTF-8 string, in BT it is the info_hash.
- Track: A set of IP:PORTs associated with a key.
- Tracker: A server that collects and publishes tracks from and to clients.
- Peer: A client that connects to a tracker to publish its connection information to a given track and retrieve the connection information of other peers in the track.
## Changes from BT
Proposed backwards compatible Changes from the BT tracking protocol:
- ‘info_hash’ is an arbitrary string (ie., a key).
- ‘peer_id’ is deprecated, if included it will be ignored.
- ‘ip’ is deprecated and can be ignored, always use the HTTP client’s ip. [XXX]
- ‘port’ is optional, if missing use the port of origin of the HTTP connection.
- ‘uploaded’, ‘downloaded’ and ‘left’ can be ignored.
- The semantics of ‘event’ are ambiguous and bittorrent specific enough that it can be safely ignored. [XXX]
## New ntrack REST-ful interface
A ntracker serves by convention in the path /ntrack, so only an ip or domain, and no full url is needed to refer to a given ntracker.
An HTTP GET request from IP 22.214.171.124 and port 1234 to the URL:
Is equivalent to this request in the classic BT tracker protocol:
## Open issues
They are marked with [XXX] in the text of this document.
- Should the Key have a length limit?
- Some kind of generic scrape interface.
- Maybe we need something like ‘event’ but with better defined semantics, so peers can easily remove themselves from the list and so on.
If we could find a way to avoid having even the key in the URL this might improve the privacy aspects a bit (many http servers log the URL and client ip/port which are enough to reconstruct the tracker’s data structures). Perhaps using HTTP POST with bencoded data rather than the querystring.
## Related protocols
One could think of ntrack as a centralized, dynamically updated, and extremely simplified DNS system.
Ntrack is compatible with the HTTP 0.2 subset of HTTP.
- Online games
- Chat systems
- Distributed computing
## see also
dtluna (email@example.com)'s status on Tuesday, 06-Dec-2016 01:27:53 UTC dtluna "Mecha-pussy" - @xj9 2016
死 (firstname.lastname@example.org)'s status on Tuesday, 06-Dec-2016 01:41:14 UTC 死 @dtluna i'm good now. i already hit bottom. starting fresh.
死 (email@example.com)'s status on Tuesday, 06-Dec-2016 01:16:49 UTC 死 @dtluna reflecting on my self-destructive tendencies.