extras merge requestshttps://gitlab.inria.fr/openvibe/extras/-/merge_requests2024-03-25T17:09:01+01:00https://gitlab.inria.fr/openvibe/extras/-/merge_requests/295Resolve "Fix share folder for acquisition-server"2024-03-25T17:09:01+01:00GARNIER LaurentResolve "Fix share folder for acquisition-server"Closes #255Closes #255GARNIER LaurentGARNIER Laurenthttps://gitlab.inria.fr/openvibe/extras/-/merge_requests/288Draft: Resolve "Add Generic Saw Tooth driver"2024-03-19T11:39:17+01:00GARNIER LaurentDraft: Resolve "Add Generic Saw Tooth driver"Port Acquisition Server to Qt/QMLGARNIER LaurentGARNIER Laurenthttps://gitlab.inria.fr/openvibe/extras/-/merge_requests/287Resolve "Fix log manager multi lines in acquisition server qt"2024-03-12T11:07:02+01:00GARNIER LaurentResolve "Fix log manager multi lines in acquisition server qt"GARNIER LaurentGARNIER Laurenthttps://gitlab.inria.fr/openvibe/extras/-/merge_requests/277Draft: Resolve "Feature: Add entropy measures"2024-03-08T16:08:26+01:00BOUNEAU AxelDraft: Resolve "Feature: Add entropy measures"Closes #248Closes #248BOUNEAU AxelBOUNEAU Axelhttps://gitlab.inria.fr/openvibe/extras/-/merge_requests/245Draft: Resolve "Update: Keyboard Stimulator always detect key press"2023-10-20T17:28:31+02:00Thomas PRAMPARTDraft: Resolve "Update: Keyboard Stimulator always detect key press"Closes #207
There is no GTK way of doing this, so the implementation is platform specific using lower level libraries.
- [ ] Linux X11
- [ ] Windows
- [ ] MacOSCloses #207
There is no GTK way of doing this, so the implementation is platform specific using lower level libraries.
- [ ] Linux X11
- [ ] Windows
- [ ] MacOSThomas PRAMPARTThomas PRAMPARThttps://gitlab.inria.fr/openvibe/extras/-/merge_requests/242Deep Learning Boxes2023-09-28T10:51:00+02:00PEPI AlexDeep Learning Boxeshttps://gitlab.inria.fr/openvibe/extras/-/merge_requests/237Added Accumulate option to box Matrix3DTo2D2024-01-04T14:40:09+01:00DESBOIS ArthurAdded Accumulate option to box Matrix3DTo2DCloses #217Closes #217DESBOIS ArthurDESBOIS Arthurhttps://gitlab.inria.fr/openvibe/extras/-/merge_requests/228Draft: Resolve "CI: Ubuntu 20 tests broken"2023-02-08T09:51:42+01:00Thomas PRAMPARTDraft: Resolve "CI: Ubuntu 20 tests broken"Closes #203Closes #203https://gitlab.inria.fr/openvibe/extras/-/merge_requests/174Draft: Resolve "Test: Confusion matrix test"2022-06-15T13:40:45+02:00MONSEIGNE ThibautDraft: Resolve "Test: Confusion matrix test"Closes #153Closes #153MONSEIGNE ThibautMONSEIGNE Thibauthttps://gitlab.inria.fr/openvibe/extras/-/merge_requests/131Draft: Resolve "Update: Classifier's architecture"2021-11-17T13:52:13+01:00MONSEIGNE ThibautDraft: Resolve "Update: Classifier's architecture"Closes #106Closes #106MONSEIGNE ThibautMONSEIGNE Thibauthttps://gitlab.inria.fr/openvibe/extras/-/merge_requests/116Draft: gtk3 initial port2021-08-03T11:33:40+02:00PAPADOPOULO TheodoreDraft: gtk3 initial portThis is a preview of the gtk3 changes for extras.
acquisition-server seems to work on my limited testing. I checked that the driver properties interfaces are working for all the drivers installed on my computer (i.e everything default + ...This is a preview of the gtk3 changes for extras.
acquisition-server seems to work on my limited testing. I checked that the driver properties interfaces are working for all the drivers installed on my computer (i.e everything default + EEGO). Fully tested with EEGO. I'm fairly confident that it is working fine, but there is always the possibility that I missed something or that something is wrongly connected between Gui and the acquisition process itself. For plugins, I just tested the SignalViewer which mostly works (an annoying bug remains because the signal window is cleared when the window is entered or or leaved, but this is not due to the callbacks. Correction will come later). All the .ui files have been adapted to gtk3 (removal of {HV}{Box,Paned,Separator} and of GtkTable, occasional conversion of GtkComboBox to GtkComboBoxText). Other big changes, are the use of cairo for all drawings and the use of css styles instead of functions to change the window styles.
A big bug which puzzled me for a while was that the SignalViewer was not working at all initially. It was a hard-to-find bug which boiled down to classes with the same name bug differing content in extras and designer. Issue opened #85 . I DID NOT SOLVE THE FUNDAMENTAL PROBLEM, I JUST MADE THE TWO STRUCTURES COHERENT... I guess this problem will be better solved when Thomas will merge the trees...
As for the designer branch, I have not removed Stock icons and several other deprecated features in gtk3.
Different bugs will be corrected, when I find them. Some of them are at the designer level. It also seems that some minor issues that were present on linux before are solved... These are mostly some information that was no longer correctly displayed (e.g. in the "window manager").
As for the designer case, this has been branched from master and not develop as I had to use my fedora changes. Sprry.https://gitlab.inria.fr/openvibe/extras/-/merge_requests/115Draft: Resolve ""Feature: Box Z score""2022-03-08T11:05:42+01:00VENOT TristanDraft: Resolve ""Feature: Box Z score""Closes #82Closes #82VENOT TristanVENOT Tristanhttps://gitlab.inria.fr/openvibe/extras/-/merge_requests/114Draft: Resolve "Feature: ERD Detection"2023-10-16T17:39:03+02:00Thomas PRAMPARTDraft: Resolve "Feature: ERD Detection"Closes #76Closes #76Thomas PRAMPARTThomas PRAMPARThttps://gitlab.inria.fr/openvibe/extras/-/merge_requests/113Resolve "Box: BrainHero Game Box"2024-01-04T10:29:16+01:00MONSEIGNE ThibautResolve "Box: BrainHero Game Box"Closes #74Closes #74https://gitlab.inria.fr/openvibe/extras/-/merge_requests/110Draft: Resolve: "Feature: Graph Laplacian box"2021-04-23T17:31:47+02:00DESBOIS ArthurDraft: Resolve: "Feature: Graph Laplacian box"Graph Laplacian + Denoising computation from Connectivity Matrices
Closes #71Graph Laplacian + Denoising computation from Connectivity Matrices
Closes #71DESBOIS ArthurDESBOIS Arthurhttps://gitlab.inria.fr/openvibe/extras/-/merge_requests/36Use the parameters instead of hardwiring some values.2019-07-16T21:13:28+02:00PAPADOPOULO TheodoreUse the parameters instead of hardwiring some values.The box ovpCBoxAlgorithmP300MagicCardVisualization (contrarily to the other boxes of the same family) has parameters to define the host and address to use for finding the acquisition server... but these parameters are not used and hard c...The box ovpCBoxAlgorithmP300MagicCardVisualization (contrarily to the other boxes of the same family) has parameters to define the host and address to use for finding the acquisition server... but these parameters are not used and hard coded values are used (localhost and 15361). See issue *6.
Having these as parameters seems a good thing as otherwise this box will not work in cases where the acquisition-server and the designer are not on the same machine... Here I took the simple solution of keeping the parameters and using them.
Other possible solutions:
- Remove the parameters from the box and simply forbid using this box in configurations where acquisition-server and designer are not run on the same machine... But then this needs to be documented.
- A third solution would be to introduce some "global" parameter space where this can be set once for all the boxes of a scenario. But this raises the case of using multiple acquisition servers in a same scenario.... A solution would be to introduce an identifier for each acquisition server in a scenario and allow boxes to refer to this identifier. This is way beyond what I was willing to do here....https://gitlab.inria.fr/openvibe/extras/-/merge_requests/35Use lua current version (not lua5.1) for all fedora builds.2019-07-17T10:46:44+02:00PAPADOPOULO TheodoreUse lua current version (not lua5.1) for all fedora builds.There are strange things hardwired in FindThirdPartyLua.cmake.
Lua 5.1 is forced on windows and "unix if not fedora21". On recent fedora (tested with fedora 28/29), this (legacy?) library does not contain lua_setglobal (it seems it is co...There are strange things hardwired in FindThirdPartyLua.cmake.
Lua 5.1 is forced on windows and "unix if not fedora21". On recent fedora (tested with fedora 28/29), this (legacy?) library does not contain lua_setglobal (it seems it is contained in ubuntu lua5.1). FindThirdPartyLua.cmake has a "hardwired" correction for the specific version of fedora 21 (which is ~4 years old). Corrected by applying the fix to all fedora distributions (it currently selects lua5.3.so which indeed contains the lua_setglobal symbol).https://gitlab.inria.fr/openvibe/extras/-/merge_requests/34Return the application return code in all cases (instead of just background m...2019-07-16T18:40:48+02:00PAPADOPOULO TheodoreReturn the application return code in all cases (instead of just background mode).This small patch correct issue #2 where the return code of the designer binary is not propagated through the shell script wrapper (it was only in background mode). This is probably true of many other scripts as they are based on the same...This small patch correct issue #2 where the return code of the designer binary is not propagated through the shell script wrapper (it was only in background mode). This is probably true of many other scripts as they are based on the same wrapping snippet....https://gitlab.inria.fr/openvibe/extras/-/merge_requests/32Add a parameter in the DisplayCueImage box to not send the triggers to the ac...2019-07-17T09:55:54+02:00PAPADOPOULO TheodoreAdd a parameter in the DisplayCueImage box to not send the triggers to the acquisition server.It is sometimes useful to create a control display that shows the cued images to the experimenter (not the subject) on a different screen. But doing so, duplicated events (one per display) are sent to the acquisition server. This is wron...It is sometimes useful to create a control display that shows the cued images to the experimenter (not the subject) on a different screen. But doing so, duplicated events (one per display) are sent to the acquisition server. This is wrong.
This PR adds a boolean parameter to the DisplayCueImage box. By default this parameter is true (i.e. triggers are sent), but using this parameter it is possible to add multiple DisplayCueImage boxes but have only one that sends the triggers to the acquisition server.
The parameter is set to true by default to keep the old behaviour. However, this setup can lead to scenarios in which multiple triggers are sent, without the experimenter noticing it immediately. Putting the parameter to false by default would have the advantage that the experimenter would immediately see that sthg is wrong (triggers would not appear on the signal display)...
Note also that the testing has been done only at the interface level as lua boxes currently do not work on linux (see issue #3).