0.9.0pre1:
-* FS: [CG]
- - Migration test error scenarios:
- 1) peers fail to connect (very, very rare) => maybe explained with bandwidth bug discovered with transport test
- 2) 1 out of 3 bits of content fails to migrate (destination peer never sees it);
- suspicion: core packs two migration messages into one, and someone throws away the 2nd half!
- => likely reproduced with new core api reliability test (which still fails)
- 3) client fails to connect to FS service for download (requests never make it to FS)
- FS was running initially, but is shut down around the time the source peer (!)
- is supposed to die; the OTHER FS service (the one that was supposed to die) is still up!
- - TTL/priority calculations
- - hot-path routing, load considerations
- - statistics
- - active reply route caching design & implementation of service; gap extension!
* TRANSPORT:
- HTTP backend [MW]
-* DV: [Nate]
- - write DV API (need to move declarations from dv_api.c to gnunet_dv_service.h!)
- - implement DV service
- - implement DV library (looks done)
- - implement DV transport plugin
- - implement testcases
- - implement performance tests (needs tbench)
-* TOPOLOGY:
- - needs more testing (especially F2F topology) & transport blacklisting
+* CORE:
+ - derived key generation [Nils]
* UTIL:
- only connect() sockets that are ready (select()) [Nils]
[On W32, we need to select after calling socket before doing connect etc.]
- - add support for UNIX domain sockets [CG]
-* GNUNET-GTK:
- - write main loop:
- + use g_main_context_set_poll_func to integrate GTK with GNUnet Scheduler!? (YUCK!)
- + OR: add scheduler API to enable integration with GTK main loop instead of doing our own select
- + use g_main_context_pending, g_main_context_query / g_main_context_check / g_main_context_dispatch
- and NEVER g_main_loop_run (can this be done? might be the clean way to do this! But how
- to integrate this with "gtk_main"? Docu says:
- "It's OK to use the GLib main loop directly instead of gtk_main(), though it involves
- slightly more typing. See GMainLoop in the GLib documentation."
- => so maybe it "just works"?
- - implement glade loader
- - implement main quit handler (file-quit and main window close)
- - implement about menu handler
- - add license to about dialog
- - add authors to about dialog
- - add translators to about dialog
- - add documenters to about dialog
- - add artists to about dialog
- - test logo in about dialog
- - test logo in main dialog (window icon)
- - implement search dialog opening and actual search
- - NS list in search dialog should use colors to offset our own namespaces from the others
- - double-clicking on NS list in search dialog should move 'root' to keyword line
- - double-clicking on search result in master list should open save-as dialog for CHK/LOC URIs
- - right-clicking on search result should open menu that allows:
- + viewing full meta data
- + starting download
- + starting recursive download
- + aborting active download
- - right-clicking on NS list in search dialog should open menu that allows
- + viewing full meta data
- + deletion of namespace info
- - master should add tab to main notebook with label with 'new' icon that allows starting of a search (same as via menu)
- - tabs in notebook should contain:
- + "X" to allow closing of tabs
- + "play" to allow resuming of search OR "pause" to allow pausing of search
- - implement open directory dialog
- - implement new-pseudonym dialog
- - implement publish dialog
- + add (single file)
- + edit (meta data add, meta data del, meta data type population, keyword add (button, activate entry line),
- keyword remove, anonymity level, priority, index/insert, normalize keywords,
- previews, preview widget for file chooser!?)
- + mark 'insert file' button inconsistent for directories
- + up
- + down
- + left
- + right
- + delete
- + open (import directory structure)
- + cancel
- + execute (publish)
- - pick appropriate sort-by columns for various tree view columns
- - anonymity spin buttons should turn red on zero:
- + search
- + publish
- + download
+* BUGS:
+ Jun 27 11:51:54 core-7670 ERROR Assertion failed at gnunet-service-core.c:3616.
+ (transport notified us that we connected to ourselves!!!)
0.9.0pre2:
* CORE:
disconnect in a timely fashion (waits for connection timeout);
need to figure a way to make it near-instant in those cases
(i.e. rapid reduction in timeout on transport-level disconnect)
+ - encrypted P2P transmission (loopback) tops at 2 MB/s on my quad-core;
+ why is it so slow? Fix! (see core_api_reliability test)
+ - implement API to get notifications about bandwidth assignments to individual peers
+ - implement API to get notifications about updates to liveness of individual peers (before disconnect)
+* STATISTICS:
+ - implement notification-based statistics API
+* PEERINFO [CG]
+ - better API for trust management (or move into FS!?)
+* TOPOLOGY:
+ - needs more testing (especially F2F topology) & transport blacklisting
* TBENCH: [MW]
- good to have for transport/DV evaluation!
+* DV: [Nate]
+ - implement performance tests (needs tbench)
* DHT: [Nate]
- implement DHT service
- implement testcases
- implement performance tests
-* ARM: [Safey]
+* MONKEY: [Safey]
- better crash management (attach debugging support, capture and analyze
debug output, detect random vs. deterministic crashes)
- - discover dependencies between services
- => need UNIX domain sockets first!
+* ARM: [CG]
- handle gnunet-arm -k in combination with auto-start magic (what is the right thing here?)
+ - discover dependencies between services
* FS: [CG]
+ - trust & peerinfo integration
+ - load considerations (migration, routing)
+ - bound our priorities based on priorities used by other peers
+ - artificial delays
+ - active reply route caching design & implementation of service; gap extension!
- datastore reservation (publishing)
- location URIs (publish, search, download)
- unindex on index failure
+ DHT integration for search
+ CS-DHT-functions (DHT-put of LOC)
+ P2P-functions (DHT-get)
+ - listing of learned namespaces
- collection API & tests
+ gnunet-pseudonym (collection support)
+ - publishing a directory sets the embedded filename to "empty"
+ (likely an issue with cutting off everything before '/' in the filename,
+ which in the case of a directory would leave nothing)
- implement FS performance tests
+ insert
+ download
+ search
+ unindex
-* GNUNET-GTK:
- - handle view-metadata toggle
- - handle view-preview toggle
- - implement API to get notifications about bandwidth assignments to individual peers
- - implement API to get notifications about updates to liveness of individual peers (before disconnect)
- - handle view-neighbours toggle
+* GNUNET-GTK: [CG]
+ - implement publish dialog details:
+ + open (import directory structure)
+ + normalize keywords (edit subdialog)
+ + set/view previews (edit subdialog)
+ + mark 'insert file' button inconsistent for directories (edit subdialog)
+ - right-clicking on search result should open menu that allows:
+ + starting recursive download
+ + aborting active download
+ + copy URI to clipboard
+ + save-as gnunet-directory
+ - events:
+ + implement clear button in search tabs (to stop downloads)
+ + search resume
+ + search error
+ + download error
+ + download resume
+ + publish resume
+ + publish error
+ - implement new-pseudonym dialog
+ - namespace search
+ - NS list in search dialog should use colors to offset our own namespaces from the others
+ - double-clicking on NS list in search dialog should move 'root' to keyword line
+ - right-clicking on NS list in search dialog should open menu that allows
+ + viewing full meta data
+ + deletion of namespace info
- extend peer dialog with green-yellow-red connectivity status lights
+ - extend peer dialog with country flags and names
- implement statistics tabs
- - handle view-statistics toggle
- do meaningful update to status line (starting up, peer running, #connections, shutdown, ...)
- implement unindex operation (use dialog with all indexed files for selection)
- implement download by URI dialog; figure out where to display those downloads!
- figure out where in the GUI we should show active uploads/unindex operations and allow aborts
-* SETUP:
- - design & implement new setup tool
+ - handle 'lost parent' case for recursive downloads (need to move children!)
+
0.9.0pre3:
* TRACEKIT: [MW]
* DATASTORE:
- GNUNET_DATASTORE_cancel method not tested
* TESTING: [Nate]
- - modify configuration to allow controlling connections for non-local starts
- - testbed creation with topology (needs working F2F topology)
- - testbed with churn
- - implement testcases for distributed testing
- test basic peer re-configure
- - test topology creation
- - test churn generation
- consider changing API for peer-group termination to
call continuation when done
* NAT/UPNP: [MW]
* MYSQL database backends: [CG]
- datacache
- datastore
-* FS:
+* FS: [CG]
- reconstruct IBLOCKS from DBLOCKS if possible (during download; see FIXME in fs_download)
- add support for pushing "already seen" search results to FS service for bloomfilter (can wait)
- use different 'priority' for probe downloads vs. normal downloads
-* ARM:
+* ARM: [CG/Safey]
- better tracking of which config changes actually need to cause process restarts by ARM.
+* SETUP:
+ - design & implement new setup tool
0.9.0:
* new webpage:
- - expand bibliography
- convert documentation pages to books
- update books (especially for developers)
- create good Drupal theme for GNUnet
* TRANSPORT:
- need to periodically probe latency/transport cost changes & possibly switch transport
- should use hash map to look up Neighbours (service AND plugins!)
-* DATASTORE (?):
- - check for duplicates on insertion (currently, same content is frequently
- stored again [seen with KBLOCKS and SBLOCKS]!)
* PEERINFO:
- merge multiple HELLOs of the same peer in the transmission queue
(theoretically reduces overhead; bounds message queue size)
projects / oweals / gnunet.git / blobdiff