Re: invalid path nma-gconf-connection
On Wed, 2010-04-14 at 23:30 +0200, Sébastien Fillaudeau wrote: > Le 14/04/2010 22:31, Dan Williams a écrit : > > On Sat, 2010-04-10 at 12:20 +0200, Sébastien Fillaudeau wrote: > > > >> Le 09/04/2010 02:04, Dan Williams a écrit : > >> > >>> On Thu, 2010-04-08 at 20:13 +0200, Sébastien Fillaudeau wrote: > >>> > >>> > Hi, > > For my work, i try to modify the mn-applet to order connection by there > priority. > > The priority for the nma-gconf-connection can be get by the path of this > connection. > > > >>> What behavior are you trying to get here with the priorities? NM > >>> doesn't use priorities for a specific reason, usually because they're a > >>> pretty bad way to interact with users. It's not often that you have > >>> more than one visible wifi network in the same area that you switch > >>> between frequently; that would usually indicate bad network planning :) > >>> There may be other ways we can adjust the behavior. > >>> > >>> But I don't think your patch is the right way to do this. You don't > >>> really want to be modifying nm-connection.c or really libnm-util at all, > >>> and you don't need to. > >>> > >>> Just abuse the 'timestamp' option to be your priority, which NM will > >>> already use to sort the connections. Disable the code in nm-applet that > >>> updates the timestamp periodically, and then add some bits in > >>> nm-connection-editor that change the timestamp based on what the user > >>> wants the priority to be. > >>> > >>> Dan > >>> > >>> > >>> > I have a probleme when i tried to get te connection path. It seems to > have illegal caracter at the beginning of the path. > > Maybe, any modification of char in gchar could cause it? > > The get_path function give a char *. > > If you know something about comparable error can you help me. > > I search a lot about this errors and i don't found any help. > > I attach the patch of my modification, maybe you could help me better if > you see the code i generated. > > > Thank for any help > > Sebastien > ___ > networkmanager-list mailing list > networkmanager-list@gnome.org > http://mail.gnome.org/mailman/listinfo/networkmanager-list > > > >>> > >>> > >>> > >>> > >> I can take my situation to make an example. For my internet connection > >> my FAI is free. > >> They propose to share your internet connection by a specific wifi network. > >>And if you share you can connect to other one wich share. > >> In my home i don't want to connect to this second limited connection. > >> But in my appartement for my study, i connect to it. > >>So if we can set priority in applet it could be useful for case like > >> this, don't you think? > >> > >> For you NetworkManager should'nt explicitly modify the priority? > >> > >> about the code : nm-connection is the representation of system connection? > >>so it is on this class i need to interacte with it ? > >> > > NMConnection is the representation of all connections, both system and > > user. But you shouldn't need to modify it if you're using the timestamp > > for the priority. > > > > > >> timestamp is a good example but it is not a element that user by using > >> applet can modify > >> > > No, but my suggestion was to make it modifiable by the user. NM just > > uses the highest value there to pick the first network to try, so > > timestamp is really a good mapping of what you're trying to do. > > > > Dan > > > > > > > > > > > ok , i was thinking that the priority of a connection was pick up by the > number of the folder it is in. No, the GConf folder is an implementation detail of how the applet stores the connection data. It's not sent or used by NM in any way. Dan ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
RE: Web login
On Wed, 2010-04-14 at 11:28 +, Mattias Bengtsson wrote: > > > Just got it - it is simple as that. Try to open any page you're sure > > > isn't accessible without authentication. It will redirect to login > > > page. If there is clear shot that is indeed login page, go ahead. If > > > not, fall back to current default - do nothing :) > > > > > > bool requiresWebLogin() > > { > > return (http://www.google.com > > == http://www.yahoo.com/ > > == http://www.hotmail.com/ ) > > } > > > > Something like that. > > > > I think Dan was against this sort of functionality. I think he referred to > some netgear routers that autoconnected to some hosts on startup or > something. A sort of unintentional DoS. Search the archives. > Some at least slightly more likely URLs then google, yahoo and hotmail are > "http://www.canonical.com/nmtest"; for ubuntu and "www.redhat.com/nmtest" for > fedora etc etc. I wasn't against it per-se, but we need to really think about this since it certainly does have the potential to make people really mad. The reference was about some DLink routers that DoS-ed the UW Madison NTP servers because they'd hard-coded the NTP server address in the firmware and didn't have any rate-limiting when the server couldn't be reached, so they'd just retry over and over and over. Think about a couple hundred thousand of these wifi APs trying to connect to a server that doesn't exist anymore, which still uses up bandwidth for the initial TCP connection setup. The point was that blindly adding some "am I really connected" query into NM itself is error-prone in a few ways: 1) if a bunch of NM-enabled machines boot up at the same time, you get a DoS effect against the site that you're trying to ping 2) you'll need a few sites to try, and you'll need to handle errors if for example a routing split means you can't contact one of the sites: http://gigaom.com/2008/10/30/cogent-sprint-un-peer-may-cause-web-slowdown/ 3) it's more complicated than you may think; some captive portals (I've seen this with Bluesocket ones) will respond to *any* ping request, even if you ping www.google.com, they'll respond to that. If you go to www.google.com, they'll respond to that too with the portal login page. This is pretty popular. So you can't just ping a site. You'll need to keep a database of login portals and associate "magic"; I know some exist out there already. You do a quick request to your preferred "am-i-on-teh-intarwebs" server, and if it matches anything in the portal database, then you know you're not connected yet. Plus this means you can have something automatically enter your username/password for you too. Otherwise, there are some standards for captive portals (WISPR?) that help you autodetect and log in more easily than doing screenscraping of the HTML. I've long wondered what Windows Vista and later do for this, since they have some sort of functionality to detect whether you're connected to the internet or not. Maybe we could take a similar approach? Windows certainly doesn't make you enter the addresses of various sites you'd like to ping. Next, this is a great use of pre-up scripts. I'd like to roll this into the same infrastructure, since for auto-login to portals, you're not really "up" and online unless you've gotten through the portal first. Which is what pre-up was all about. Dan ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: Web login
On Wed, 2010-04-14 at 09:42 -0400, Martin Owens wrote: > On Tue, 2010-04-13 at 14:57 +0300, Peteris Krisjanis wrote: > > Just got it - it is simple as that. Try to open any page you're sure > > isn't accessible without authentication. It will redirect to login > > page. If there is clear shot that is indeed login page, go ahead. If > > not, fall back to current default - do nothing :) > > Logging in is really a second step to providing a simple boolean about > whether the network is in such a state to begin with. Right now a > fundamental frustration is the lack of notification and api to make > work-flow more intelligent. > > Get that in there and then we can do the more interesting auto-logging > on functionality. We have some designs cached for how notifiers would > work in such a situation, but NM api is holding them back. I assume you mean something like pre-up scripts? That's where I'd put this sort of functionality. Dan ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: GUI thoughts: confusing menu with multiple wifi cards
On Sun, 2010-04-11 at 05:08 +0100, Richard Neill wrote: > Dear All, > > I've used NM a bit with a machine which has multiple network cards, and > I thought I'd attach the following screenshot. I think that the layout > is rather confusing, and perhaps could be improved. > > I think it's confusing because the horizontal rules appear to take > visual precedence over the slightly less-indented headings. Thus it > seems to be divided into 5 sections: Yeah, the multi-card case is somewhat special. But in this case, you almost always dont' want both cards to connect to the same thing, so you need to control the separately. Second, as you've seen, both cards don't necessarily see the same APs. Your iwl3945 is a/b/g, while your p54 is just b/g. Thus we simply shouldn't show any APs under the p54 card that are in the 5GHz band, because the p54 can't connect to them. Same goes for encryption capabilities, some older cards can't do AES/CCMP, so we shouldn't show any APs for those cards that are only WPA2 capable. We've got various ideas for redoing the menu into a more flexible format [1], but not a lot of time to commit to that yet. But it's been clear for a while that we need to focus on the UI more, since the current UI is no longer fulfilling all the needs people have of it. Dan [1] http://blogs.gnome.org/dcbw/2009/05/26/face-transplants-are-the-new-botox/ > > > 1. Wired Network,Disconnected, Wireless Networks (IntelPro), Mobile, > Disconnect > > > > 2. Mobile-Eapsim, Wireless Networks (Intersil), disconnected. > > > > 3. Mobile, Mobile EAPSIM > > > > 4. VPN connections > > > > 5. Connect to Hidden, Connect to New > > > If there is only one WiFi card, it's far less confusing. However the > screenshot attached could, imho, be much improved by a better layout. > > > Best wishes, > > Richard > > > P.S. I seem not to be thinking very clearly - I was going to add this > comment as a post-script to my previous message; it then occurred to me > that it would be clearer in a separate thread, but added the > (irrelevant) attachment to the last message too - sorry for the confusion! > ___ > networkmanager-list mailing list > networkmanager-list@gnome.org > http://mail.gnome.org/mailman/listinfo/networkmanager-list ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: invalid path nma-gconf-connection
Le 14/04/2010 22:31, Dan Williams a écrit : On Sat, 2010-04-10 at 12:20 +0200, Sébastien Fillaudeau wrote: Le 09/04/2010 02:04, Dan Williams a écrit : On Thu, 2010-04-08 at 20:13 +0200, Sébastien Fillaudeau wrote: Hi, For my work, i try to modify the mn-applet to order connection by there priority. The priority for the nma-gconf-connection can be get by the path of this connection. What behavior are you trying to get here with the priorities? NM doesn't use priorities for a specific reason, usually because they're a pretty bad way to interact with users. It's not often that you have more than one visible wifi network in the same area that you switch between frequently; that would usually indicate bad network planning :) There may be other ways we can adjust the behavior. But I don't think your patch is the right way to do this. You don't really want to be modifying nm-connection.c or really libnm-util at all, and you don't need to. Just abuse the 'timestamp' option to be your priority, which NM will already use to sort the connections. Disable the code in nm-applet that updates the timestamp periodically, and then add some bits in nm-connection-editor that change the timestamp based on what the user wants the priority to be. Dan I have a probleme when i tried to get te connection path. It seems to have illegal caracter at the beginning of the path. Maybe, any modification of char in gchar could cause it? The get_path function give a char *. If you know something about comparable error can you help me. I search a lot about this errors and i don't found any help. I attach the patch of my modification, maybe you could help me better if you see the code i generated. Thank for any help Sebastien ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list I can take my situation to make an example. For my internet connection my FAI is free. They propose to share your internet connection by a specific wifi network. And if you share you can connect to other one wich share. In my home i don't want to connect to this second limited connection. But in my appartement for my study, i connect to it. So if we can set priority in applet it could be useful for case like this, don't you think? For you NetworkManager should'nt explicitly modify the priority? about the code : nm-connection is the representation of system connection? so it is on this class i need to interacte with it ? NMConnection is the representation of all connections, both system and user. But you shouldn't need to modify it if you're using the timestamp for the priority. timestamp is a good example but it is not a element that user by using applet can modify No, but my suggestion was to make it modifiable by the user. NM just uses the highest value there to pick the first network to try, so timestamp is really a good mapping of what you're trying to do. Dan ok , i was thinking that the priority of a connection was pick up by the number of the folder it is in. and timestamp was just a attribut to sort the list display for the user. I will try using this approach Sébastien ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: Preconfiguring WEP keys for network manager
On Sun, 2010-04-11 at 04:56 +0100, Richard Neill wrote: > Dear All, > > I've just joined this list, so I hope I'm asking this question in the > right place. What I'm trying to do is to deploy a large number of new > company laptops, all with the ESSID/WEP key preconfigured. We're using > Ubuntu's Ubiquity with automatic install from a USB stick. So I need to > write some shell script whose effect will be that Network Mananger > already knows the WEP key for the default network, with no further user > interaction. See my reply to your other message about system connections. > If I put the key into /etc/network/interfaces, then on boot, the machine > does come up with the network connected, but as soon as it goes out of > range, the connection is lost, and NM is unable to reconnect. Yes, this might happen, mainly so that NM doesn't keep reconnecting over and over and over to an AP that you're not able to connect to. This is a bit tricky to solve, because not every AP shows up in every scan. So if we clear the inhibit flag every time the AP drops out of the scan list, we may end up trying to connect again to an AP for which you hit the "Cancel" button when NM asked you for the passphrase. That doesn't help you; but it's clearly something we need to fix. I think we do that by differentiating between a "failure" of the connection and cancellation of the password request. For some devices (bluetooth, 3G) where we can't scan for something to connect to, we treat failure as "don't try again automatically". For wifi though, we only set the inhibit when the user hits cancel since at that point, they clearly don't want to connect again until they choose it from the menu. If you're interested in helping out with a patch for this, I dont' think it would be that hard to do, and I'll happily guide you through various bits. Thanks, Dan > The wep key simply isn't "precious" data, and I'd like to get NM to use > a configuration file which doesn't need unlocking, and which never > prompts the user. These machines will need to be remotely accessible, so > if the relevant Wifi is in range, the latop must automatically connect > to it, without waiting on the user to click OK, or authorise anything. > > Is there anything I can do here? > > Thanks for your help, > > Richard > > ___ > networkmanager-list mailing list > networkmanager-list@gnome.org > http://mail.gnome.org/mailman/listinfo/networkmanager-list ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: is there any way to get NM to store WEP keys without using a keyring?
On Wed, 2010-04-14 at 15:38 +0100, Richard Neill wrote: > Dear All, > > Am trying to get a set of laptops preconfigured for various wifi > networks and WEP keys. I've set the laptops to autologin, but now > gnome's keyring is prompting for a password every time, which rather > defeats the purpose; it also means that an unattended or remote reboot > will not come back online. How can I get NM to treat WEP keys as > "world-readable" on this computer? You're probably looking for "system connections": http://live.gnome.org/NetworkManager/SystemSettings which are stored not in the user's gconf/keyring, but in your distros normal network configuration files (or via 'keyfiles'). You could also do this with gconf/gnome-keyring, but I don't think that's quite what you want. So create the connection you want in nm-connection-editor, then click the "Available to all users" button and hit Apply. It's now a system connection, and where it's stored depends on what plugins you have enabled in /etc/NetworkManager/nm-system-settings.conf. Dan ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: [PATCH] [nm-applet] Fix compilation using -DGSEAL_ENABLE (bgo#615653)
On Wed, 2010-04-14 at 17:24 +0200, Mirsal Ennaime wrote: > Hello, > > This patch substitutes all direct access to Gtk+ object fields > with accessor functions in order to fulfill the UseGseal goal > as described at http://live.gnome.org/GnomeGoals/UseGseal > > It also requires to bump the required Gtk+ version to 2.18 We probably won't be able to do this until the next major version of NM (0.9) since it's not nice to bump reqs in the middle of a stable release stream. I'll keep the patch around until then though. Thanks! Dan ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: complex openvpn - can nm just launch?
On Fri, 2010-04-09 at 09:43 +0200, Robert Vogelgesang wrote: > Dan, > > On Thu, Apr 08, 2010 at 05:15:54PM -0700, Dan Williams wrote: > > On Tue, 2010-04-06 at 22:01 +, Alessandro Bono wrote: > > > On Tue, 06 Apr 2010 09:25:44 -0700, Dan Williams wrote: > > > > > > > On Tue, 2010-04-06 at 10:05 -0600, Scott Serr wrote: > > > >> I have an openvpn config file that works fine with openvpn. (ubuntu > > > >> lucid beta) As far as I can tell there is no way to create a like > > > >> config in the nm openvpn editor. I can make one somewhat similar and > > > >> export, but it doesn't look enough like mine to work. > > > > > > > > Which options? > > > > > > Hi Dan > > > > > > this is my (short) list of missing options/features > > > > > > - support for external dhcp on the server side, normally I configure > > > openvpn server to push only data that I can't provide via dhcp server. So > > > ip/mask/dns is taken from dhcp and additional route from openvpn > > > This configuration works perfectly for windows machine, on certain > > > customer I have a dedicated openvpn only for me because I can't use > > > "normal" openvpn configuration :-( > > > > Yeah, we need support for this internally. Right now we pretty much > > assume a tunnel approach, not TAP. It's not that hard to fix that I > > guess; but in general the real fix for this would be helped by some of > > the activation changes that I'd like to do to fix the PPPoE issues that > > people currently have. > > I'd like to see this feature, too. Could you please elaborate on what > has to be done to support this? 1) add an "method" item to NetworkManagerVPN.h: /* string: IP4 configuration method */ #define NM_VPN_PLUGIN_IP4_CONFIG_METHOD "method" /* Values for NM_VPN_PLUGIN_IP4_CONFIG_METHOD */ #define NM_VPN_PLUGIN_IP4_CONFIG_METHOD_DHCP "dhcp" #define NM_VPN_PLUGIN_IP4_CONFIG_METHOD_STATIC "static" if the plugin doesn't send 'method' in the config dict, or the item is a zero-length string, 'static' is assumed. 2) In the openvpn plugin, if we're supposed to use DHCP (is tap always used with DHCP, or are there cases where it's not?) then we add the NM_VPN_PLUGIN_IP4_CONFIG_METHOD item to the returned IP4 config struct with the value "dhcp". 3) Then we need to modify nm_vpn_connection_ip4_config_get() and split it into two functions, one for DHCP and one for static. Take the stuff at the bottom of that function (everything below print_vpn_config()) and put that into a separate function that gets called by both the static and dhcp processing bits. So you'll have something like: nm_vpn_connection_ip4_config_get() { const char *method = "static"; /* remove the timeout since the plugin replied */ g_source_remove (priv->ipconfig_timeout); priv->ipconfig_timeout = 0; val = g_hash_table_lookup (config_hash, NM_VPN_PLUGIN_IP4_CONFIG_METHOD); if (val && G_VALUE_HOLDS_STRING (val)) method = g_value_get_string (val); if (!method || !strcmp (method, NM_VPN_PLUGIN_IP4_CONFIG_METHOD_STATIC)) { if (handle_static_ip4_config (connection, config_hash)) return; } else if (method && !strcmp (method, NM_VPN_PLUGIN_IP4_CONFIG_METHOD_DHCP)) { if (handle_dhcp_ip4_config (connection, config_hash)) return; } else nm_log_err (LOGD_VPN, "unknown vpn IP4 method '%s', method); /* same error stuff as at the bottom of the function now */ } For the DHCP4 bits, we'll want to build up the NMIP4Config object as much as possible and cache that in priv->ip4_config while DHCP is completing. We'll need to add a few things to teh NMVPNConnection object's private data, like: NMDHCPManager * dhcp_manager; NMDHCPClient * dhcp4_client; gulong dhcp4_state_sigid; gulong dhcp4_timeout_sigid; (see nm-device.c for DHCP stuff). When the NMVPNConnection is initialized, lets grab a reference to the DHCP manager in nm_vpn_connection_init(): priv->dhcp_manager = nm_dhcp_manager_get (); and then in handle_dhcp_ip4_config() we'll do something like: static gboolean handle_dhcp_ip4_config (NMVPNConnection *vpn, GHashTable *config) { NMVPNConnectionPrivate *priv = NM_VPN_CONNECTION_GET_PRIVATE (vpn); NMSettingConnection *s_con; NMSettingIP4Config *s_ip4; const char *uuid; ip4_config, see nm_vpn_connection_ip4_config_get() for how to do this> s_con = NM_SETTING_CONNECTION (nm_connection_get_setting (priv->connection, NM_TYPE_SETTING_CONNECTION)); s_ip4 = NM_SETTING_IP4_CONFIG (nm_connection_get_setting (priv->connection, NM_TYPE_SETTING_IP4_CONFIG)); priv->dhcp4_client = nm_dhcp_manager_start_ip4 (priv->dhcp_manager, , nm_setting_connection_get_uuid (s_con), s_ip4, ,
Re: AD-Hoc network specify channel
On Tue, 2010-04-13 at 21:36 +0200, Simon Schampijer wrote: > Hi, > > I am trying to create an Ad-Hoc Network with a specific channel. From > the spec [1] I know that I have to set the band as well. However the > created network always has channel 1. The nm-applet does not have an > option to set the channel, and is using channel one, too. > > Using iwconfig to create an Ad-Hoc network I can set a channel fine. Is > it fully implemented or do I miss something else? It should be, though it's likely not possible if you're using nm-connection-editor since the channel/band aren't accessible there. But looking at it now, they *should* be accessible for ad-hoc connections. So we'll need to whip up some code for page-wifi.c to hide/show them when ad-hoc is selected. (this is because wpa_supplicant only has the ability to set the channel for ad-hoc networks, not a limitation in NM. wpa_supplicant does not have the ability to set a band limitation for infrastructure networks, which means we can't let users limit particular connection to only 802.11a for example) See nm_supplicant_config_add_setting_wireless() in NM for where the frequency gets sent to the supplicant, and build_supplicant_config() for where the channel/band get translated into a frequency. Also check nm_ap_new_fake_from_connection() to ensure that the frequency is getting correctly set on the fake AP. This all should work as long as the connection dict has the band/freq specified. Dan ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: gobbledegook in iwconfig after NM disconnect
On Sun, 2010-04-11 at 05:20 +0100, Richard Neill wrote: > Dear All, > > Not sure if this is a bug, but if I disconnect a wireless interface from > network manager, the result of iwconfig's ESSID becomes gibberish. > I have no idea where the value of > > \x90\xef\xfd\x...@dub5\xab3q8\xe2\xcf\xdc\x8db+\xa3\x9f\x1d\xaa1\x82\xa4\xfa\xdczsl > comes from, or what it means! The Linux Wireless Extensions API (the older API that programs use to control wifi) doesn't have a clean way to say "Disconnect and stop doing anything until I tell you." This random SSID is done by wpa_supplicant when disconnecting from the network to ensure that the driver/firmware does not attempt to automatically connect to anything else until it's told to do so. It's not anything that NM is doing, and it's normal. The new kernel wireless API (cfg80211) has fixed this problem by properly including a "disassociate" command that idles the device. If you've told NM (or the card) to disconnect, you probably did mean for it to disconnect. Without this random SSID the card (depending on the driver or firmware) may well attempt to connect to some SSID that you didn't want it to. Dan > Here is an example: > > 1. While NM is connected to the network > > $ iwconfig wlan0 > wlan0 IEEE 802.11abg ESSID:"26PPWG602" >Mode:Managed Frequency:2.427 GHz Access Point: > 00:0C:F6:7E:FA:18 >Bit Rate=54 Mb/s Tx-Power=15 dBm >Retry long limit:7 RTS thr:off Fragment thr:off >Power Management:off >Link Quality=65/70 Signal level=-45 dBm >Rx invalid nwid:0 Rx invalid crypt:0 Rx invalid frag:0 >Tx excessive retries:0 Invalid misc:0 Missed beacon:0 > > > 2. Immediately after disconnecting the Wifi, via the NM gui > > $ iwconfig wlan0 > wlan0 IEEE 802.11abg ESSID:"\x90\xef\xfd\x...@dub5\xab3q8\xe2\xcf > \xDC\x8Db+\xA3\x9F\x1D\xAA1\x82\xA4\xFA\xDCZsl" >Mode:Managed Frequency:2.427 GHz Access Point: > Not-Associated >Tx-Power=15 dBm >Retry long limit:7 RTS thr:off Fragment thr:off >Power Management:off > > > 3. After running ifconfig wlan0 down > > $ iwconfig wlan0 > > wlan0 IEEE 802.11abg ESSID:off/any >Mode:Managed Frequency:2.427 GHz Access Point: > Not-Associated >Tx-Power=15 dBm >Retry long limit:7 RTS thr:off Fragment thr:off >Power Management:off > > > ___ > networkmanager-list mailing list > networkmanager-list@gnome.org > http://mail.gnome.org/mailman/listinfo/networkmanager-list ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: invalid path nma-gconf-connection
On Sat, 2010-04-10 at 12:20 +0200, Sébastien Fillaudeau wrote: > Le 09/04/2010 02:04, Dan Williams a écrit : > > On Thu, 2010-04-08 at 20:13 +0200, Sébastien Fillaudeau wrote: > > > >> Hi, > >> > >> For my work, i try to modify the mn-applet to order connection by there > >> priority. > >> > >> The priority for the nma-gconf-connection can be get by the path of this > >> connection. > >> > > What behavior are you trying to get here with the priorities? NM > > doesn't use priorities for a specific reason, usually because they're a > > pretty bad way to interact with users. It's not often that you have > > more than one visible wifi network in the same area that you switch > > between frequently; that would usually indicate bad network planning :) > > There may be other ways we can adjust the behavior. > > > > But I don't think your patch is the right way to do this. You don't > > really want to be modifying nm-connection.c or really libnm-util at all, > > and you don't need to. > > > > Just abuse the 'timestamp' option to be your priority, which NM will > > already use to sort the connections. Disable the code in nm-applet that > > updates the timestamp periodically, and then add some bits in > > nm-connection-editor that change the timestamp based on what the user > > wants the priority to be. > > > > Dan > > > > > >> I have a probleme when i tried to get te connection path. It seems to > >> have illegal caracter at the beginning of the path. > >> > >> Maybe, any modification of char in gchar could cause it? > >> > >> The get_path function give a char *. > >> > >> If you know something about comparable error can you help me. > >> > >> I search a lot about this errors and i don't found any help. > >> > >> I attach the patch of my modification, maybe you could help me better if > >> you see the code i generated. > >> > >> > >> Thank for any help > >> > >> Sebastien > >> ___ > >> networkmanager-list mailing list > >> networkmanager-list@gnome.org > >> http://mail.gnome.org/mailman/listinfo/networkmanager-list > >> > > > > > > > > > I can take my situation to make an example. For my internet connection > my FAI is free. > They propose to share your internet connection by a specific wifi network. > And if you share you can connect to other one wich share. > In my home i don't want to connect to this second limited connection. > But in my appartement for my study, i connect to it. > So if we can set priority in applet it could be useful for case like > this, don't you think? > > For you NetworkManager should'nt explicitly modify the priority? > > about the code : nm-connection is the representation of system connection? > so it is on this class i need to interacte with it ? NMConnection is the representation of all connections, both system and user. But you shouldn't need to modify it if you're using the timestamp for the priority. > timestamp is a good example but it is not a element that user by using > applet can modify No, but my suggestion was to make it modifiable by the user. NM just uses the highest value there to pick the first network to try, so timestamp is really a good mapping of what you're trying to do. Dan ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
is there any way to get NM to store WEP keys without using a keyring?
Dear All, Am trying to get a set of laptops preconfigured for various wifi networks and WEP keys. I've set the laptops to autologin, but now gnome's keyring is prompting for a password every time, which rather defeats the purpose; it also means that an unattended or remote reboot will not come back online. How can I get NM to treat WEP keys as "world-readable" on this computer? Thanks, Richard ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
RE: Web login
> > Just got it - it is simple as that. Try to open any page you're sure > > isn't accessible without authentication. It will redirect to login > > page. If there is clear shot that is indeed login page, go ahead. If > > not, fall back to current default - do nothing :) > > > bool requiresWebLogin() > { > return (http://www.google.com > == http://www.yahoo.com/ > == http://www.hotmail.com/ ) > } > > Something like that. > I think Dan was against this sort of functionality. I think he referred to some netgear routers that autoconnected to some hosts on startup or something. A sort of unintentional DoS. Search the archives. Some at least slightly more likely URLs then google, yahoo and hotmail are "http://www.canonical.com/nmtest"; for ubuntu and "www.redhat.com/nmtest" for fedora etc etc. /Mattias - lurking here for like 3 years ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: No connect pop-up for PPTP VPN
On Wed, 2010-04-14 at 13:44 -0400, Matthew Saltzman wrote: > When I connect to my VPNC VPN, a pop-up notifies me of the successful > connection. When I connect to my PPTP VPN, there is no similar > notification--the orbit animation stops and the lock appears, but that's > the only indication that the connection is complete. That's because only vpnc has the ability to send a "banner" along with the connection. Neither PPTP or openvpn have this ability. But the applet should really just show something even if there isnt' a banner. Dan ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
No connect pop-up for PPTP VPN
When I connect to my VPNC VPN, a pop-up notifies me of the successful connection. When I connect to my PPTP VPN, there is no similar notification--the orbit animation stops and the lock appears, but that's the only indication that the connection is complete. -- Matthew Saltzman Clemson University Math Sciences mjs AT clemson DOT edu http://www.math.clemson.edu/~mjs ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: NM 0.8 can't connect to PEAP
On Wed, 2010-04-14 at 11:19 -0400, Matthew Saltzman wrote: > On Tue, 2010-04-06 at 19:34 -0400, Matthew Saltzman wrote: > > > I found and fixed the continuous asking for the password even when > > > "Always ask" wasn't checked. > > > > That's not in NM-0.8.0-4.git20100325.fc12 I suppose. I'll look > > forward > > to testing the next update in F12 updates-testing. > > Indeed, I grabbed NetworkManager-0.8.0-6.git20100408.fc12.x86_64 from > updates-testing, and that seems to fix the password problem. I had a similar problem that got fixed by changing the username. My credientials are for the university infra-structure and instead of using the normal login: XX I've used: xxx...@domain.tld And somehow it went through smoothly. > > > > > Thanks! > > > > > I could not reproduce the missing CA > > > certificate issue though. > > I couldn't either with this package (though I've only tried once so > far). I also have a colleague here (Bill Moss) who insists that he's > never had that problem. Must be some bizarre interaction effect with > something, I guess. (I also had a hard-disk crash in the interim that > took out my home directory, so who knows what else has changed.) ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
[PATCH] [nm-applet] Fix compilation using -DGSEAL_ENABLE (bgo#615653)
Hello, This patch substitutes all direct access to Gtk+ object fields with accessor functions in order to fulfill the UseGseal goal as described at http://live.gnome.org/GnomeGoals/UseGseal It also requires to bump the required Gtk+ version to 2.18 -- Mirsal From 53e95da9a186dcc21cbcaccfee4e506deab65847 Mon Sep 17 00:00:00 2001 From: Mirsal Ennaime Date: Tue, 13 Apr 2010 15:46:13 +0200 Subject: [PATCH] Fix compilation using -DGSEAL_ENABLE This patch substitutes all direct access to Gtk+ object fields with accessor functions in order to fulfill the UseGseal goal as described at http://live.gnome.org/GnomeGoals/UseGseal It also requires to bump the required Gtk+ version to 2.18 --- configure.ac |2 +- src/applet-device-gsm.c |8 src/applet-device-wifi.c |3 ++- src/applet-device-wired.c |2 +- src/applet-dialogs.c | 20 +--- src/applet.c | 15 +-- src/utils/nma-bling-spinner.c | 21 + src/wireless-dialog.c |4 ++-- 8 files changed, 41 insertions(+), 34 deletions(-) diff --git a/configure.ac b/configure.ac index a928b6a..97351fa 100644 --- a/configure.ac +++ b/configure.ac @@ -77,7 +77,7 @@ PKG_CHECK_MODULES(NMA, libnm-glib >= 0.8 libnm-util >= 0.8 libnm-glib-vpn >= 0.8 - gtk+-2.0 >= 2.14 + gtk+-2.0 >= 2.18 libglade-2.0 gmodule-export-2.0 gconf-2.0 diff --git a/src/applet-device-gsm.c b/src/applet-device-gsm.c index 939a4a7..a88168c 100644 --- a/src/applet-device-gsm.c +++ b/src/applet-device-gsm.c @@ -723,16 +723,16 @@ ask_for_pin_puk (NMDevice *device, else if (!strcmp (secret_name, NM_SETTING_GSM_PUK)) w = gtk_label_new (_("PUK code is needed for the mobile broadband device")); if (w) - gtk_box_pack_start (GTK_BOX (dialog->vbox), w, TRUE, TRUE, 0); + gtk_box_pack_start (GTK_BOX (gtk_dialog_get_content_area (dialog)), w, TRUE, TRUE, 0); dev_str = g_strdup_printf ("%s", utils_get_device_description (device)); w = gtk_label_new (NULL); gtk_label_set_markup (GTK_LABEL (w), dev_str); g_free (dev_str); - gtk_box_pack_start (GTK_BOX (dialog->vbox), w, TRUE, TRUE, 0); + gtk_box_pack_start (GTK_BOX (gtk_dialog_get_content_area (dialog)), w, TRUE, TRUE, 0); w = gtk_alignment_new (0.5, 0.5, 0, 1.0); - gtk_box_pack_start (GTK_BOX (dialog->vbox), w, TRUE, TRUE, 0); + gtk_box_pack_start (GTK_BOX (gtk_dialog_get_content_area (dialog)), w, TRUE, TRUE, 0); box = GTK_BOX (gtk_hbox_new (FALSE, 6)); gtk_container_set_border_width (GTK_CONTAINER (box), 6); @@ -749,7 +749,7 @@ ask_for_pin_puk (NMDevice *device, g_signal_connect (w, "changed", G_CALLBACK (pin_entry_changed), ok_button); pin_entry_changed (GTK_EDITABLE (w), ok_button); - gtk_widget_show_all (dialog->vbox); + gtk_widget_show_all (gtk_dialog_get_content_area (dialog)); return GTK_WIDGET (dialog); } diff --git a/src/applet-device-wifi.c b/src/applet-device-wifi.c index bc82afd..19c239f 100644 --- a/src/applet-device-wifi.c +++ b/src/applet-device-wifi.c @@ -58,7 +58,8 @@ show_ignore_focus_stealing_prevention (GtkWidget *widget) { gtk_widget_realize (widget); gtk_widget_show (widget); - gtk_window_present_with_time (GTK_WINDOW (widget), gdk_x11_get_server_time (widget->window)); + gtk_window_present_with_time (GTK_WINDOW (widget), + gdk_x11_get_server_time (gtk_widget_get_window (widget))); } static void diff --git a/src/applet-device-wired.c b/src/applet-device-wired.c index 7f10e57..b1649cd 100644 --- a/src/applet-device-wired.c +++ b/src/applet-device-wired.c @@ -569,7 +569,7 @@ pppoe_get_secrets (NMDevice *device, w = gtk_dialog_add_button (GTK_DIALOG (info->dialog), GTK_STOCK_OK, GTK_RESPONSE_OK); info->ok_button = w; - gtk_box_pack_start (GTK_BOX (GTK_DIALOG (info->dialog)->vbox), + gtk_box_pack_start (GTK_BOX (gtk_dialog_get_content_area (GTK_DIALOG (info->dialog))), glade_xml_get_widget (xml, "DslPage"), TRUE, TRUE, 0); diff --git a/src/applet-dialogs.c b/src/applet-dialogs.c index 5b891d0..b2140c7 100644 --- a/src/applet-dialogs.c +++ b/src/applet-dialogs.c @@ -564,7 +564,8 @@ applet_info_dialog_show (NMApplet *applet) g_signal_connect (dialog, "delete-event", G_CALLBACK (gtk_widget_hide_on_delete), dialog); g_signal_connect_swapped (dialog, "response", G_CALLBACK (gtk_widget_hide), dialog); gtk_widget_realize (dialog); - gtk_window_present_with_time (GTK_WINDOW (dialog), gdk_x11_get_server_time (dialog->window)); + gtk_window_present_with_time (GTK_WINDOW (dialog), + gdk_x11_get_server_time (gtk_widget_get_window (dialog))); } static void @@ -697,7 +698,8 @@ applet_warning_dialog_show (const char *message) gtk_window_set_title (GTK_WINDOW (dialog), _("Missing resources")); gtk_widget_realize (dialog); gtk_widget_show (dialog); - gtk_window_present_with_time (GTK_WINDOW (dialog), gdk_x11_get_server_time (dialog->window)); + gtk_window_present_with_time (GTK_
Re: NM 0.8 can't connect to PEAP
On Tue, 2010-04-06 at 19:34 -0400, Matthew Saltzman wrote: > > I found and fixed the continuous asking for the password even when > > "Always ask" wasn't checked. > > That's not in NM-0.8.0-4.git20100325.fc12 I suppose. I'll look > forward > to testing the next update in F12 updates-testing. Indeed, I grabbed NetworkManager-0.8.0-6.git20100408.fc12.x86_64 from updates-testing, and that seems to fix the password problem. > > Thanks! > > > I could not reproduce the missing CA > > certificate issue though. I couldn't either with this package (though I've only tried once so far). I also have a colleague here (Bill Moss) who insists that he's never had that problem. Must be some bizarre interaction effect with something, I guess. (I also had a hard-disk crash in the interim that took out my home directory, so who knows what else has changed.) -- Matthew Saltzman Clemson University Math Sciences mjs AT clemson DOT edu http://www.math.clemson.edu/~mjs ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list
Re: Web login
On Tue, 2010-04-13 at 14:57 +0300, Peteris Krisjanis wrote: > Just got it - it is simple as that. Try to open any page you're sure > isn't accessible without authentication. It will redirect to login > page. If there is clear shot that is indeed login page, go ahead. If > not, fall back to current default - do nothing :) Logging in is really a second step to providing a simple boolean about whether the network is in such a state to begin with. Right now a fundamental frustration is the lack of notification and api to make work-flow more intelligent. Get that in there and then we can do the more interesting auto-logging on functionality. We have some designs cached for how notifiers would work in such a situation, but NM api is holding them back. Best Regards, Martin Owens ___ networkmanager-list mailing list networkmanager-list@gnome.org http://mail.gnome.org/mailman/listinfo/networkmanager-list