summaryrefslogtreecommitdiffstats
path: root/perl-install/network
Commit message (Collapse)AuthorAgeFilesLines
* fix bogus use of old netwok module instead of network::network (missed whenThierry Vignaud2003-11-141-2/+2
| | | | | cleaning up this)
* fix program logic: write up/down programs for adsl speedtouch even ifThierry Vignaud2003-11-121-1/+2
| | | | | | firmware is already there or if one say "let see this later" which result in up/down scripts to be removed on drakconnect startup
* fix bogus use of old netwok module instead of network::networkThierry Vignaud2003-11-121-2/+2
|
* reuse is_dynamic_ip() to not bother asking about zeroconf specific stuff whenThierry Vignaud2003-11-121-1/+1
| | | | | there's not any dynamic interfaces
* (network::tools::is_dynamic_ip) consolidate test for any dynamic interfaceThierry Vignaud2003-11-121-1/+6
|
* (is_wireless_intf) fix #6312: list acx100_pci as a wireless networkThierry Vignaud2003-11-081-1/+1
| | | | | cards driver too (fix impossible to set wireless parameters for it)
* fix lan always marked as detected even when no detection was performedThierry Vignaud2003-11-081-6/+6
| | | | | | | | | | | | or when there's no ethernet devices btw, generalize this "fix" to all net connection types. rationale: since we've already check if we've devices, there's no point in doing another different check in order to add the string "- detected" so let's just consolidate the previous test, this will prevent such bug to come back again
* fix last part of #5315: do not ask for confirmation at install timeThierry Vignaud2003-11-081-1/+1
| | | | | | rationale: while this is reasonable within standalone drakconnect, it's quite less needed at install time
* perl_checker fixesThierry Vignaud2003-11-071-3/+1
|
* (conf_network_card_backend) use getNetDriver() rather than relying onThierry Vignaud2003-11-071-1/+1
| | | | | | ethX alias for mapping the ethX interface to the driver that created it thus fixing quite old detection bugs
* oops, network::tools is still needed by configureNetworkIntf...Thierry Vignaud2003-11-071-0/+1
|
* (write_interface_conf) setVarsInSh() already take care of not writingThierry Vignaud2003-11-071-2/+1
| | | | | | | undefined parameters, so it's safer to always write wireless parameters if set else we drop user set wireless parameters when our wireless modules list is not uptodate
* one more wireless moduleThierry Vignaud2003-11-061-1/+1
|
* (configure_eth_aliases) let the autoprobing load modules tooThierry Vignaud2003-11-061-1/+0
|
* let only remove aliases on bootstrapping (ie only in harddrakeThierry Vignaud2003-11-061-1/+0
| | | | | caller); it's safer
* (configure_eth_aliases) introduce it to autoconfigure all ethernetThierry Vignaud2003-11-061-0/+10
| | | | | aliases (MdkMove, harddrake service, drakconnect)
* (configureNetwork) alter remove_alias() callers for new APIThierry Vignaud2003-11-061-1/+1
|
* simplify pcmcia matchingThierry Vignaud2003-11-061-5/+2
|
* save a useless ifconfig forkThierry Vignaud2003-11-061-1/+0
|
* document dams logicThierry Vignaud2003-11-061-1/+1
|
* (conf_network_card_backend) fix unseen/unconfigured usb wirelessThierry Vignaud2003-11-061-2/+1
| | | | | | | | | devices due to francois trap for aliased interfaces (new detect_device::getNet() filter out aliases which were the reasons for the test) btw, if no alias was found, there's a bug somewhere
* list a few more modules as wireless ones (should really be splited outThierry Vignaud2003-11-061-1/+1
| | | | | from network/usb in list_modules.pm
* always list sagem_dhcp in list, hidding it due to missing translationsThierry Vignaud2003-11-061-3/+1
| | | | | is insane
* escaped strings fixesThierry Vignaud2003-11-062-3/+3
|
* (configure) fix untranslatable strings: switch from ask_from_list_()Thierry Vignaud2003-11-061-7/+7
| | | | | to ask_from_list()
* unexport not used elsewhere connected2 functionThierry Vignaud2003-11-041-1/+1
|
* perl_checker fixesThierry Vignaud2003-11-041-3/+3
|
* fix always detecting ethernet cards even when none is presentThierry Vignaud2003-10-301-1/+2
|
* fix anthill bug #50: ensure /etc/ppp/pap-secrets is not world readableThierry Vignaud2003-10-301-2/+4
| | | | | since it contains password/user mapping for dialup
* fix installer when network::ethernet wasn't loaded beforeThierry Vignaud2003-10-291-0/+1
|
* (isdn_get_info) simplify it trough MDK::CommonThierry Vignaud2003-10-271-3/+1
|
* simplify isdn detection:Thierry Vignaud2003-10-272-4/+3
| | | | | | | - no need to pass a ref around; if it's bound to be overwritten, just return a new hash - let isdn_detect only care of which fields we want for isdn stuff
* - make isdn detection more readableThierry Vignaud2003-10-271-1/+1
| | | | | - ensure previously detected isdn stuff is cleared on redetection
* handle usb isdn adaptersThierry Vignaud2003-10-221-2/+4
|
* add entries for usb isdn adaptersThierry Vignaud2003-10-221-0/+10
|
* fix steping back on choosing isdn card stepThierry Vignaud2003-10-221-1/+1
|
* fix isdn detectionThierry Vignaud2003-10-221-1/+1
|
* fix #6159: fix detection when a local name server is faking theThierry Vignaud2003-10-201-7/+7
| | | | | | connection because of its cache by checking at least a packet is ack-ed
* roolback hp fix:Thierry Vignaud2003-10-171-1/+1
| | | | | | | | after further investigation at Houston, the bcm440 bug was not related to MII_NOT_SUPPORTED (ethtool/mii support is fine in this driver). instead, it is related to half/full duplex detection.
* - simplifyThierry Vignaud2003-10-172-24/+15
| | | | | | | | - remove spurious module imports & requires - remove never used arg (this enable to simplify caller) - fix detection logic: do not mix old detected stuff with new one if one step back in drakconnect after plugging some network devices in
* fix for HP (current module does not work if MII_NOT_SUPPORTED to set to yes)Francois Pons2003-10-161-1/+1
|
* - translate all server namesThierry Vignaud2003-10-141-5/+5
| | | | | - harmonize server names
* (default_interfaces)Thierry Vignaud2003-10-141-16/+8
| | | | | | | | | | | - simplify: ask_from already optimize the one item case for us - make it show up a pull-down menu rather than forcing one to type in the interface - fix it: $conf{net_interface} was *not* set on multiple interfaces case (read) fix it (conf was uselely initialized)
* readd strict & diagnostics pragma removed by florinThierry Vignaud2003-10-131-2/+2
|
* simplify "do not remove wireless parameters from ifcfg when not in wizard ↵Thierry Vignaud2003-10-082-8/+4
| | | | mode" fix
* network interface configuration step: using network interface fromThierry Vignaud2003-10-081-3/+4
| | | | | | interface/driver list is more robust that relying on possibly empty int variable (better #3276 fix)
* fix #3276: when steping back in drakconnect wizard, do not overwriteThierry Vignaud2003-10-061-3/+3
| | | | | first card parameters with last one's
* check both nfs servers version 2 and version 3, and remove duplicates (bug ↵Pascal Rigaux2003-10-061-11/+17
| | | | #6055)
* cleanupsThierry Vignaud2003-10-031-2/+0
|
* only read /etc/sysconfig/network-scripts/ifcfg-*, not old.ifcfg-* andThierry Vignaud2003-10-031-1/+1
| | | | | the like...