Commit graph

25 commits

Author SHA1 Message Date
685a9b84ee Library: replace boost::noncopyable by our own library solution
Benefits
 - get rid of yet another pervasive Boost dependency
 - define additional more fine grained policies (move only, clonable)
2018-03-24 05:35:13 +01:00
937ad64596 DiffMessage: now uniformly plays the role of MutationMessage (closes #1066) 2017-08-13 07:25:32 +02:00
5fbc4b84bf DiffMessage: switch to moving DiffMessage over the bus
basically the opaque-buffer based MutationMessage implementation is obsoleted now
2017-08-12 17:59:02 +02:00
0d5ca55019 GuiNotification: define outline of this service interface 2017-01-20 03:27:32 +01:00
06a61773fa Nexus/CoreService: consider handling of bus connections.
Now I've realised that there are two degrees of connectedness.
It is very much possible to have a "free standing" BusTerm, which
only allows to send uplink messages. In fact, this is how CoreService
is implemented, and probably it should also the way how to connect
the GuiNotification service...
2017-01-20 01:54:49 +01:00
e75152b29d install and activate UI-Bus in the actual GUI (#1043)
Decision was made to use the CoreService as PImpl to organise
all those technical aspects of running the backbone. Thus,
the Nexus (UI-Bus hub) becomes part of CoreService
2016-12-10 04:01:06 +01:00
a54990de7c define the plan for some scaffolding to drive the UI-Session connection (#1042)
...following a similar idea as employed when developing the Player-Engine connection
2016-12-10 01:21:08 +01:00
f995dd51e2 define creation and control structure of TimelineWidget 2016-12-03 05:42:34 +01:00
1725a31df1 Bugfix: insidious dangling pointer caused by move after construction
Damn sideeffect of the suppport for move-only types: since we're
moving our binding now into place /after/ construction, in some cases
the end() iterator (embedded in RangeIter) becomes invalid. Indeed this
was always broken, but didn't hurt, as long as we only used vectors.

Solution: use a dedicated init() hook, which needs to be invoked
*after* the TreeMutator has been constructed and moved into the final
location in the stack buffer.
2016-10-03 23:54:09 +02:00
c8ad698ac4 MutationMessage: limit to treating of gui::model::Tangible
the generic typing to DiffMutatble does not make much sense,
since the desired implementation within gui::ctrl::Nexus
is bound to work on Tangibles only, since that is what
the UI-Bus stores in the routing table
2016-10-02 23:51:45 +02:00
d2e4f826ed UI-Bus/mutation: expand on draft for mutation message 2016-10-01 23:09:08 +02:00
27ba8d5896 UI-Bus/mutation: draft idea for mutation message on UI-Bus 2016-09-30 22:23:55 +02:00
44bb044eee message broadcast implementation unit test PASS
...was indeed dead easy to implement
2016-02-14 02:20:51 +01:00
eaa12499f3 back to UI command invocation: basically implement a placeholder command
based on the previous experiments, this adds a fake operation
and a definition frame to hook this operation as pseudo Proc-Layer command

WIP: the invocation itself is not yet implemented.
     We need to build a custom invocation pattern for that,
     in order to be able to capture the instance-ID of the command
     on invocation

NOTE: also, because of #989, we can not bind a time value for this test
2016-01-22 12:19:25 +01:00
c9ea9de54d cover basic up-link and down-link communication in BusTerm_test 2016-01-03 03:37:52 +01:00
540151b56b provide a mock handler for commands and state marks
in the real system, this will be the task of the CoreService,
while here, in test mode, we allow to install handling closures
from the unit-test-code
2016-01-03 03:23:39 +01:00
2e4d74747e implement logging overrides in the Test-Nexus implementation
the actual functionality is forwareded to the base class,
which is the regular Nexus Hub
2016-01-02 19:16:37 +01:00
603f9e2b7d DOC: fix some Doxygen link syntax
I worked under the erroneous assumption, that Doxygen
will use its internal entity-IDs as the link-IDs when
generating mardown-links. Yes, this seemed logical and
this would be the way I'd implement it....

But seemingly, Doxygen is not so consistent when it
comes to questions of syntax. The same holds true for
markdown, which lacking a coherent definition anyway.

Another problem is that Doxygen's auto-link generation
frequently fails, for reasons not yet clear to me.
Sometimes it seems to be necessary to give it a nudge
by including the \ref command. While I'm not willing
to go into focussed invstigation of Doxygen syntax
right now, at least I've done a search-and-replace
to remove the malformed links I've written the
last days
2015-12-27 03:16:49 +01:00
4d1fcd6dcb implement logging/identification of mock UI elements 2015-12-25 00:41:14 +01:00
972045d8f8 fix a segfault caused by re-entrance
at the point when we're connecting a new node to the UI-Bus,
the new node's BusTerm is not yet initialised, since we need
the uplink connection we're about to create for setting up
this BusTerm.

Consequently, the new nodes's ID is not yet initialised,
so we need to pass this endpoint-ID explicitly to the
registration function.
2015-12-24 22:43:41 +01:00
38e6d635dc implement routing in the UI-Bus hub 2015-12-24 22:27:37 +01:00
37802fd699 complete wiring of nexus and core services
since, by definition, the Nexus is "the" up-link,
all we need is clever overriding of the relevant
handling functions, so the nexus will care for the routing,
while the CoreService cares for command and presentation
state handling
2015-12-19 04:05:21 +01:00
449fa16444 WIP: draft of the UI backbone setup
- front-end handle
- Nexus == routing hub
- CoreService == service access point for command invocation
2015-12-19 03:40:19 +01:00
0f793e0a79 untangle the setup of test nexus 2015-12-18 19:50:57 +01:00
f19ebd63d0 pondering about how to connect the central bus hub
...especially since we need a faked UI backbone for unit testing
2015-12-18 19:50:02 +01:00