Hello, please take a moment to read this. I will appreciate any help.
I have a gtk application that runs on linux and win32.
I writing on a serial display but on linux the com flushes too slow while on
win32
seems ok.
I test it the display with the hyperterminal and didnt see any problem with it.
On Mon, 2006-11-13 at 02:15 -0500, Freddie Unpenstein wrote:
On Mon 10/23, James Scott Jr [EMAIL PROTECTED] wrote:
I am about to write the socket module of a program that follows the
gnu standard structure for a daemon. I am using glib wherever
possible and have a question concerning the
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
James Scott Jr wrote:
On Mon, 2006-11-13 at 02:15 -0500, Freddie Unpenstein wrote:
On Mon 10/23, James Scott Jr [EMAIL PROTECTED] wrote:
I am about to write the socket module of a program that follows the
gnu standard structure for a daemon. I
to the gnu standard structure for
a daemon part of your post. In either case, I would like to know
whence _that_ documentation.
Chris.
Ok, I may have misunderstood the question, as my answer related to
g_io_channel... usage.
If I am back on track, the standard daemon model is based on my
On Mon 10/23, James Scott Jr [EMAIL PROTECTED] wrote:
I am about to write the socket module of a program that follows the
gnu standard structure for a daemon. I am using glib wherever
possible and have a question concerning the use of linux signals.
Basically, I have blocked all signals on
and use a single thread with a sigwait() to process the ones I'm
interested in; like SIGHUP to reload the config file.
Question: Will having all signals blocked interfere with the normal
operation of glib, in such a way that that I need to rethink how to
handle signals? I plan to use g_io_channel
James Scott Jr wrote:
Folks,
Hi James,
in a phrase - I've done it and it works fine.
some places might call for special attention though - for
example - if you were to call g_spawn_sync() or g_child_watch(),
you may want to have leave SIGCHLD available.
Cheers,
-Tristan
On Mon, 2006-10-23 at 13:42 -0400, Tristan Van Berkom wrote:
James Scott Jr wrote:
Folks,
Hi James,
in a phrase - I've done it and it works fine.
some places might call for special attention though - for
example - if you were to call g_spawn_sync() or g_child_watch(),
you may
when it moves on to build gobject, it hangs up while running
glib-genmarshal. If I look at the task manager, ntvdm.exe is using
a lot of CPU - why would glib-genmarshal be using the 16-bit VM?
Heh, no reason. One thing to look out for is that the
glib-genmarshal.exe you built needs to run
On 11/16/05, Tor Lillqvist [EMAIL PROTECTED] wrote:
See the patch below that clears up some of the issues: 1) After
getting FD_CLOSE, make the knowledge of that stick and always set
G_IO_HUP. 2) Plug a handle leak, WSACloseEvent the socket's event when
the channel is closed.
If you can build
There are some problems both in your code, and in giowin32.c.
For efficiency, your code should not read just one byte at a time in
the callback. It should read in bigger chunks, and as much as possible
in a loop, until it gets a WSAEWOULDBLOCK or an EOF. If it did this,
it would probably also
On 11/15/05, Tor Lillqvist [EMAIL PROTECTED] wrote:
There are some problems both in your code, and in giowin32.c.
For efficiency, your code should not read just one byte at a time in
the callback. It should read in bigger chunks, and as much as possible
in a loop, until it gets a
I fixed that (and what do you know... it works!). Thanks for noticing.
Good to hear!
Should I make an attempt at the changes to giowin32.c, or is that
something you've already got sorted out?
See the patch below that clears up some of the issues: 1) After
getting FD_CLOSE, make the knowledge
Tor Lillqvist writes:
See the patch below that clears up some of the issues:
Forgot to mention, 3) Calls WSAEnumNetworkEvents() only if the event
for the socket has fired. This means the spurious
WSAEnumNetworkEvents() calls you noticed being logged when just moving
the mouse don't happen any
g_io_win32_check: WSAEnumNetworkEvents (1576, 0x618) revents={}
condition={IN|ERR|HUP} events={}
What seems to happen is that after watching a socket io channel once,
I'll continue to see these messages until I restart the program - even
if the io channel should have been cleaned up by
On 11/14/05, Tor Lillqvist [EMAIL PROTECTED] wrote:
g_io_win32_check: WSAEnumNetworkEvents (1576, 0x618) revents={}
condition={IN|ERR|HUP} events={}
What seems to happen is that after watching a socket io channel once,
I'll continue to see these messages until I restart the program - even
Greetings,
I'm having a strange problem in trying to get a program (gaim) working
with Glib 2.8.x (2.8.3) under win32.
I'm watching for IO on a socket using g_io_channel_unix_new(fd), and
then watching for the result using g_io_add_watch_full() (basically
using functionality modeled after
I need to send a data structure over an io_channel. In the
reference
manual, I found that the only available instruction to write data
in a
channel is:
GIOStatus g_io_channel_write_chars(GIOChannel *channel,
const gchar *buf,
Vivien Malerba writes:
if it's on the same machine, then no problem!).
...as long as the writer and reader processes are compiled with the
same compiler and compiler options, that is.
--tml
___
gtk-app-devel-list mailing list
On 7/8/05, Vivien Malerba [EMAIL PROTECTED] wrote:
Beware however that depending on the contents of your structure, you
may have some problems if you exchange data this way between machines
which do not have the same endian-ness (if it's on the same machine,
then no problem!).
should not be
On Fri, 8 Jul 2005, Vivien Malerba wrote:
On 7/8/05, [EMAIL PROTECTED] [EMAIL PROTECTED] wrote:
g_io_channel_write_chars(channel, (const gchar *)your_struct, sizeof
(your_struct), written, errors);
Beware however that depending on the contents of your structure, you
may have some problems if
how i can use g_io_channel to make a simple echo-server?
libgnet comes with a simple example which does exactly that.
Which reminds me... Does anyone know the differences between gnet and
gnetwork? Which is the more GNOME sanctioned of the two?
Fredderic
22 matches
Mail list logo
