graphics/twm4nx/src/ciconmgr.cxx: Integrate use of CButtonArray; implement first cut at event handling.
graphics/nxwidgets/src/cbuttonarray.cxx: Add a method to CButtonArray that will allow us to dynamically resize the array (at the cost of losing all button labels).
graphics/twm4nx/src/ciconmgr.cxx: Add some fragments of CButtonArray logic.
graphics/twm4nx/src/cwindow.cxx: Finishes first cut at window event management.
version 1.0.10 to NuttX NX windows server. No, a port is not the right
word. It is a re-design of TWM from the inside out to work with the NuttX
NX server. The name Twm4Nx reflects this legacy. But Twm4Nx is more a
homage to TWM than a port of TWM.
The original TWM was based on X11 which provides a rich set of features.
TWM provided titlebars, shaped windows, several forms of icon management,
user-defined macro functions, click-to-type and pointer-driven keyboard
focus, graphic contexts, and user-specified key and pointer button bindings,
etc.
Twm4Nx, on the other hand is based on the NuttX NX server which provides
comparatively minimal support. Additional drawing support comes from
the NuttX NxWidgets library (which necessitated the change to C++).
Twm4Nx is greatly stripped down and targeted on small embedded systems
with minimal resources. For example, no assumption is made about the
availability of a file system; no .twmrc file is used. Bitmaps are not
used (other than for fonts).
The TWM license is, I believe compatible with the BSD license used by NuttX.
The origin TWM license required notice of copyrights within each file and
a full copy of the original license which you can find in the COPYING file.
within this directory.
STATUS:
This port was brutal. Much TWM logic was removed because it depending on X11 features (or just because I could not understand how to use it). The logic is partial. A lot more needs to be done to have a complete system (hence, it is marked EXPERIMENTAL). The kinds of things that need to done are:
1. Update some logic that is only fragmentary for how like resizing, and menus.
2. Integrate NxWidgets into the windows: The resize menu needs a CLabel, the menus are CListBox'es, but not completely integrated, the Icon Manager needs to be a button array.
3. Resit Icons. They are windows now, but need to be compound widgets lying on the background.
4. Widget events are only partially integrated. A lot more needs to be done. A partial change to thoe event system that hints at the redesign is in place but it is far from complete.
apps/system/nxplayer: Fix some logical errors from recent comments. They broke the build of the nxplayer as a library vs. a task. apps/graphics/nxglyphs: Fix yet more namespace problems.
graphics/nxwm/src/nxwm_main.cxx: Move from graphics/nxwidgets/UnitTests/nxwm/nxwm_main.cxx.
apps/graphicx/nx*: Fixes numerous build issues mostly associated with include paths and namespaces.
apps/nxglyphs: Decouple (mostly) from nxwidgets and nxwm so that they can be used elsewhere. Creates include/graphics/nxglyphs.h
apps/graphics/nxglyphs: Put all NxWidgets and NxWM glyphs into a common directory where they can eventually be shared. Not very sharable at the moment due to header file entanglements. Need a separate nxglyps.hxx header file.
Separate nxwidets and nxwm into separate directories. Remove old, common NxWidgets directory.
1. A memory corruption issue that occurs from a paste operation that would overflow the fixed buffer size for keyboard processing.
2. A stall in getch() processing when there are cached keycodes in the termcurses emulation (tcurses_vt100.c).
The patch provided in issue 135, which adds support for HTTP Chunked Encoding, covers all paths in the webserver app except CGI callbacks. As a result, if a page being served happens to use CGI, it could generate stream content that does not comply with HTTP Chunked Encoding.
The patch attached amends the webserver app's CGI callbacks to use the HTTP Chunked Encoding sender function instead of using send() directly.
The webserver/httpd app is currently broken when script support is enabled (CONFIG_NETUTILS_HTTPD_SCRIPT_DISABLE). The root cause has been tracked down to the "Content-length" not being available ahead of time in this case (length of -1 passed to send_headers() ). On the other hand, the server closing the socket does not result in FIN being sent to the browser either (FIN not supported by NuttX yet).
Simple solution: Add support for HTTP Chunked Encoding to webserver/httpd (attached patch).
The attached patch is simple. It adds a configuration option to enable chunked encoding. When enabled, the implementation will auto-detect the cases where content length is not available ahead of time, and will automatically engage chunked encoding transfers.
Without this patch, the browser/client hangs forever, as it is expecting more data. With this patch, the browser displays the content.
NxWidgets improvements
* NxWidgets: fix garbage returned when CCycleButton::getValue() is called after removeAllOptions().
It's a bit questionable whether returning 0 here is reasonable or
if it would be better to assert(). But either is better than reading
into random memory and returning a garbage value.
* NXWidgets CScrollingPanel: Don't draw outside the widget area.
The port->move() function doesn't support clipping the
copied area to the client area, so we have to manually
calculate the part that can be moved without going outside
the widget.
* CNxString: Add string + operator and ::format() function
* CNumericEdit: Add option to include unit name after the value
* NxWidgets: CNxWidget: Make useWidgetStyle() public
Makes it easier to update styles when multiple controls are nested.
* NxWidgets: Add CLabelGrid control for displaying text in grid format.
Approved-by: GregoryN <gnutt@nuttx.org>
Reviewers note: Reviewing the changes I also see that these file did not follow the NuttX codings standard. I ran all files through nuttx/tools/indent.sh, manually reviewed all files for coding style issues. I also changed occurrences of non-standard types u8_t, u16_t, and u32_t to the standard uint8_t, uint16_t, and uint32_t.
Squashed commit of the following:
Clean-up some include path problems introduced with the previous set of changes.
app/include/graphics, apps/NxWidgets/UnitTests: Update include paths for file in new location.
apps/NxWidgets/nxwidgets and nxwm: Update include paths for file in new location.
apps/graphics/NxWidgets: Move nxwidgets and nxwm include/ directories to apps/include/graphics
- fix error "sz_main.o: No such file or directory"
- support -p <path> for rz to change the folder for the recevied file
- switch debug output from printf to syslog
- send the next packet for ZME_ACK in ZMS_SENDING to avoid rz on the host side stuck
- make send work reliable even without hardware flow control
testing: add Unity - unit testing library from ThrowTheSwitch.org
* examples/powerled: add arch initialization; examples/smps: fixes in some printf and in Kconfig
* testing: add Unity - unit testing library from ThrowTheSwitch.org
Approved-by: GregoryN <gnutt@nuttx.org>
apps/examples/ft80x: Fix some size calculations. Add option to disable primitive tests... just too boring to have to watch over and over again.
apps/graphics/ft80x: Fix a typo in backlight fade logic; Fix error in formmatted display light debug dump output.
apps/graphics/ft80x: Fix some warnings when debug features are enabled.
apps/graphics/ft80x: Replace polling logic with gentler signal logic when waitinf for the CMD FIFO to become empty.
apps/examples/ft80x: Add first co-processor example.
apps/graphics: Extend library to manage display lists either in display memory (RAM DL) or provided directly to th co-processor FIFO (RAM CMD).
apps/examples/ft80x: Some minor clean-up. Trivial changes for coprocessor support.
apps/graphics/ft80x: Add support for accessing graphics ram. apps/examples/ft80x: Add more primitive graphics demos.
apps/examples/ft80x: Add a couple more demos of primitives. Use new ft80x_dl_create() to simplify.
apps/graphics/ft80x: Add ft80x_dl_create() which simplies writing of very simple display lists.
apps/examples/lvgldemo: Fix up some include paths
apps/examples/lvgldemo: The last final now conforms to the NuttX coding style.
apps/graphics/littlevgl: Two more files are closer to the NuttX coding style.
apps/graphics/littlevgl: Two more files are closer to the NuttX coding style.
apps/examples/lvgldemo: Three more files are closer to the NuttX coding style.
apps/examples/lvgldemo: Two more files are closer to the NuttX coding style.
apps/examples/lvgldemo: Two more files are closer to the NuttX coding style.
Add include/graphics/.gitignore
apps/include/graphics: Remove lvgl.h. graphics/littlevgl/Makefile now copies the file into position.
apps/graphics/littlevgl: Add .gitignore file
apps/graphics/littlevgl: Most of these files follow the coding style now.
apps/graphics/littlevgl: This gets a couple of files closer to the NuttX coding style.
apps/examples/lvgldemo: Add lvgldemo example
apps/graphics/littlevgl: Add littlevgl library
FTP: Adds support for IPv6 and fixes various transfer issues
* netutils/ftpd: Fix support for IPv6 FTP server
* netutils/ftpc:Adds support for IPv6 sockets
* ftpc: Must convert port received by EPSV reply to network order
* ftpc: Data socket address for passive connection should be same as server address
* ftpc: Must skip human readable string in EPSV response before scanning for port
* ftpc: Don't send PASS command if USER command was sufficient. ie no password required
* ftpc: Generically handle permanent negative completion in shard ftpc_cmd logic
* ftpc: Minor addresssing fix
* ftpc: Tweak logic for overriding network debugging output
* FTP: Adds option for setting stack size of various threads in FTPD and FTPC
* netutils/ftpd: Fixes build error setting IPv4 address. sin_addr => sin_addr.s_addr
* netutils/ftpd: Protects against partial write returns by looping in that case
Approved-by: Gregory Nutt <gnutt@nuttx.org>
apps/fsutils/mkfatfs: New user-space fatfs appears to work fine.
apps/fsutils/mkfatfs: Move mkfatfs from the OS to here. Not fully integrated on the intial commit.
apps/netutils/netlib: netlib_ipv[4|6]adaptor() will now use the new routing table interfaces if they are available: If the desination IP address does not correspond to the network served by any device, then the address of the router that should get us to the correct sub-net is is determined. Then the network adaptor that can communicate with the that router has the IP address that will returned.
apps/netutils/netlib: Add new utilities netlib_ipv[4|6]router() that can be used to determine the IP address of a router that would be used some some destination IP address that is not locally accessible.
apps/netutils/netlib: Add helpers for reading the routing table: netlib_read_ipv4route() and netlib_read_ipv6route()
Merged in merlin17/apps/i8sak-sock (pull request #108)
i8sak: Fixes mistake where socket was accidentally bound to endpoint address
i8sak: Fixes bad logic that bound outbound port to the destination address
i8sak: Adds sniffer port option and a few other get/set parameters
Merged in merlin17/apps/i8sak-sock (pull request #106)
i8sak: Endpoint address is updated whenever 802.15.4 addressing information is updated
wireless/ieee802154/i8sak: Converts PF_INETX to PF_INET6 and SOCK_IWPAN to SOCK_DGRAM
wireless/ieee802154/i8sak: Catches case where arguments are included, but the interface was never set
wireless/ieee802154/i8sak: Endpoint IP address is now updated whenever 802.15.4 addressing info changes
wireless/ieee802154/i8sak: Routes all error output to stderr rather than stdout
Merged in merlin17/apps/i8sak-sock (pull request #104)
i8sak: Fixes socket version of sniffer logic. socket logic for sniffer now works
i8sak: Fixes socket version of sniffer logic. socket logic for sniffer now works
Merged in merlin17/apps/i8sak-sock (pull request #103)
wireless/ieee802154/i8sak: Adds socket interface support. You can now use both socket or char driver to control the MAC layer
* wireless/ieee802154: Removes useless wpanlistener folder with build files
* wireless/ieee802154/libmac: Merges some files together and adds some missing sixlowpan functionality
* wireless/ieee802154/i8sak: Starts adding support for both character driver and socket
* Removes unneccessary Kconfig statement
* ieee802154: Adds function prototypes missed on previous commit
* wireless/ieee802154/libmac: Fixes sixlowpan_resetreq
* wireless/ieee802154/libmac: Small fixes and removes enableevents helper
* apps/ieee802154/i8sak: Moves wpanlistener functionality into i8sak directly
* apps/ieee802154/i8sak: Major restructuring to allow both char driver and socket control
Also changes Kconfig options and certain args held previously by i8sak. This is to bring it closer to a real CLI tool rather than a development tool. For instance, no longer does the app change the addressing of the device automatically. This is because you probably don't want i8sak overwriting the extended address after board bring-up. Additionally, the endpoint logic has been simplified. There is only ever one active endpoint now.
* wireless/ieee802154/i8sak: Non-functional clean-up.
* i8sak: Removes chan and coordinfo commands and introduces generic get/set commands
* i8sak: startpan now requires PAN ID argument
* i8sak: Adds ability to set endpoint address using set command
* i8sak: Minor fixes and cleanup
* i8sak: Adds addrmode setting to i8sak. exposed via get/set calls
* i8sak: Clean up poll logic and allow polling address mode to be based off endpoint address setting.
* i8sak: Trivial
* i8sak: Adds get/set hook for endpoint address mode (destination address mode)
* i8sak: Changes default address modes to short.
* i8sak: Fixes build erros associated with enabling 6LoWPAN support
* i8sak: Adds socket version of get/set implementation. Simplifies some logic as well.
* wireless/ieee802154/i8sak: Removes unused files
* wireless/ieee802154/libmac: Fixes sixlowpan_assocresp. resp was not being copied into the ioctl argument.
Approved-by: Gregory Nutt <gnutt@nuttx.org>
Here are some offenders: examples/cc3000/shell.c, examples/nsh/nsh_main.c, examples/nxterm/nxterm_main.c, and examples/pty_test/pty_test.c. It is possible that for some of those, that represents the first call to nsh_consolemain(). But probably not.
To give other initialization options, I added nsh_consolechild() which is identical to nsh_consolemain() other than it omits all of the one-time initialization. Of course, you would not get the result that you wanted if nsh_consolechild() were called without nsh_consolemain().
