aboutsummaryrefslogtreecommitdiffstats
path: root/include/zot.php
Commit message (Collapse)AuthorAgeFilesLines
* this should finish up hubzilla issue #113zotlabs2017-12-311-0/+9
|
* sql compliance issuezotlabs2017-12-101-1/+1
|
* default profile assignzotlabs2017-11-241-0/+2
|
* use loggerzotlabs2017-11-081-2/+1
|
* :bulb: Improving Doxygen documentation.Klaus Weidenbach2017-11-031-193/+292
| | | | | Fix some Doxygen parsing errors. Improve hooks documentation.
* put deferred queue logic every place we create a delivery process (except ↵zotlabs2017-11-021-0/+17
| | | | for protocol 'friend requests' which aren't likely to swamp the delivery system). Remove it from the queue_delivery function which was too late to do anything.
* don't perform zot_refresh on dead sites unless $force is setzotlabs2017-11-011-0/+10
|
* do not send message_list responses to dead sites (this delivery method ↵zotlabs2017-11-011-1/+1
| | | | bypassed the notifier)
* Maria's profile photo issue, please push to master.zotlabs2017-11-011-0/+4
|
* wrong keyzotlabs2017-10-221-4/+2
|
* regression in zot site info after zot6 workzotlabs2017-10-161-4/+6
|
* typozotlabs2017-10-111-1/+2
|
* for federated forums we need to keep track of what protocols are available ↵zotlabs2017-10-101-0/+16
| | | | to red/hubzilla channels to determine which posts need to be mangled for transport to individual networks - based on what networks the author can communicate with. We probably need this for all xchans and hublocs but at some point it becomes increasingly difficult to scrape this information and map out a compatibility matrix. It's also doubtful anybody will use this system because we basically have to forge comments between network sites involving different protocols and this is going to look like crap to anybody that isn't on red/Hubzilla. Eventually they have to fix their protocols for this to work correctly, but nobody seems to believe me that their networks are basically anti-federation, so we need something like this to highlight their walled garden specifications and the resulting federation problems in a form they can't really argue with.
* more zot6 basic stuffzotlabs2017-09-251-9/+5
|
* zot_site_info - always sign the site block with the site private keyzotlabs2017-09-251-4/+4
|
* tagging changes - provides ability to mention a forum by using !forumname as ↵zotlabs2017-09-241-1/+1
| | | | well as the traditional red style (@forumname+). This should probably not be advertised on a wide scale until after a critical mass of sites have updated to a version containing these changes. This adds yet another option type to the ACL module which probably needs refactoring soon since it is turning into option type spaghetti.
* encrypt delivery reports. This will cause a bit of consternation because ↵zotlabs2017-09-211-3/+20
| | | | this is not backward compatible and older sites will see some incorrect delivery reports from newer sites until the next upgrade; as their systems will not know how to read the ecnrypted final reports.
* make wiki page mimetype default to wiki mimetype (this was the original ↵zotlabs2017-09-131-1/+3
| | | | intent but the wrong variable was used)
* more zot6zotlabs2017-09-131-0/+1
|
* Zot6: some things that need to be done ahead of time so we can turbocharge ↵zotlabs2017-09-131-11/+35
| | | | the delivery engine and split off site records from channel records.
* avoid a security patch and resultant compatibility issues; instead restrict ↵zotlabs2017-09-051-2/+1
| | | | the input characters we accept in token verification strings to hex digits. This will all be changing in the coming weeks/months anyway.
* validate the security contextzotlabs2017-09-031-0/+1
|
* lognoise: don't log sync packet contents if there are no clones to receive themzotlabs2017-08-231-2/+3
|
* consolidate disable_discover_tab config if it was not yet set to either 1 or ↵git-marijus2017-08-061-1/+3
| | | | 0 in DB
* more work on signature algorithm choicezotlabs2017-07-261-12/+16
|
* start to implement the code needed for verifying variable signature algorithmszotlabs2017-07-251-15/+31
|
* sync key changes to cloneszotlabs2017-07-251-6/+108
|
* make signing methods variable and signablezotlabs2017-07-251-1/+2
|
* new field names did not get savedzotlabs2017-07-241-2/+2
|
* provide a keychange operation to rebase an identity on a new keypairzotlabs2017-07-241-0/+44
|
* hubzilla issue #832zotlabs2017-07-191-1/+1
|
* update some addon docs and ensure we only generate statistics once a dayzotlabs2017-07-131-1/+1
|
* prevent expiration of conversations you are involved with - allows you to ↵zotlabs2017-07-101-1/+11
| | | | find your own comments months from now
* minor weirdness in zot finger results after deleting a clone from a channel ↵zotlabs2017-06-281-5/+9
| | | | that was on a site which was previously migrated from http to https and still had the old hubloc
* explain what is happening herezotlabs2017-06-201-0/+3
|
* more import_author_zot fixeszotlabs2017-06-201-2/+9
|
* start deprecation of server_rolezotlabs2017-05-291-7/+0
|
* start using the abook_not_here flag (requires changes in addons, not easily ↵zotlabs2017-05-251-0/+5
| | | | portable to hubzilla)
* issue #769zotlabs2017-05-111-2/+2
|
* abook_not_here flag created to indicate singleton connections which are ↵zotlabs2017-05-031-3/+1
| | | | connected to this channel but not on this hub. abook_instance enumerates which hubs the connections is valid, but we ultimately need something more efficiently searchable to decide what operations are supported w/r/t this connection in the context of this hub. This flag is ignored during sync to clones although the code to set it correctly during channel creation, import, and sync has not yet been implemented.
* make sure the new fields don't leak through sync packets and break older ↵zotlabs2017-04-121-0/+2
| | | | versions
* only log zot_refresh content if json decode was successful.zotlabs2017-04-071-2/+2
|
* revisit the import_author_zot algorithm yet again. There was one bug that we ↵zotlabs2017-04-071-2/+34
| | | | weren't returning necessary information in the first SQL query - and performance/loading problem if one tries to refresh a dead site.
* import_author_xchan - since we rarely refresh zot-info for non-connections, ↵zotlabs2017-04-061-1/+1
| | | | force a cache reload once a week to catch things like profile photo updates and location changes
* create site_store_lowlevel() to initialise data structures for the site tablezotlabs2017-04-041-15/+17
|
* import_author_zot() - check for both hubloc and xchan entries. This should ↵zotlabs2017-04-031-5/+14
| | | | catch and repair entries which were subject to transient storage failures.
* crypto updateszotlabs2017-04-021-3/+0
|
* We've passed the deadline for allowing unsigned zot finger tokens. It is now ↵zotlabs2017-02-241-4/+1
| | | | a protocol requirement.
* more work on permcats and consolidating calls that try to discover connect ↵zotlabs2017-02-091-33/+21
| | | | permissions, also create lowlevel store functions for abook and profile - since these currently may have issues with sql strict mode.
* fix typosMario Vavti2017-01-311-1/+1
|