Hi,
I don't see how this commits helps ? The function code is exactly the same in both files and the db.c dependency in vty_interface.c hasn't been removed. Am I missing something here ?
[ See commit 424c4f0e2927d5a7538b31c69113c6e4f861d2c9 on the git for context ]
Sylvain
On Wednesday 06 January 2010 08:58:42 Sylvain Munaut wrote:
Hi,
I don't see how this commits helps ? The function code is exactly the same in both files and the db.c dependency in vty_interface.c hasn't been removed. Am I missing something here ?
Well... this is all not set to stone but the idea is the following:
libbsc.a contains all functionality that is in the domain of the BSC. This includes the vty_interface.c.
libmsc.a contains all the handling of GSM04.08. In our case even the HLR/VLR (db.c). In our case this is everything that is not in the BSC.
bsc_hack links to the libbsc and libmsc and adds vty_interface_layer3. Now the dependencies don't really matter as the link will just resolve them.
For the on-waves/bsc-master branch I'm only using libbsc.a. This means your commit added a dependency on db.c again. On top of that it will really help us keeping the BSC/MSC separation in place which is beneficial to stuff like channel handling..
does this clear things up?
Hi,
libbsc.a contains all functionality that is in the domain of the BSC. This
includes the vty_interface.c.
libmsc.a contains all the handling of GSM04.08. In our case even the HLR/VLR (db.c). In our case this is everything that is not in the BSC.
bsc_hack links to the libbsc and libmsc and adds vty_interface_layer3. Now the dependencies don't really matter as the link will just resolve them.
For the on-waves/bsc-master branch I'm only using libbsc.a. This means your commit added a dependency on db.c again. On top of that it will really help us keeping the BSC/MSC separation in place which is beneficial to stuff like channel handling..
does this clear things up?
I understand the idea of separation (the naming scheme _layer3 could be better :). But what I don't see is how that commits restores it. Because it doesn't remove the db.c dependency: from vty_interface.c you still access the DB.
Sylvain
On Wednesday 06 January 2010 09:35:10 Sylvain Munaut wrote:
Hi,
I understand the idea of separation (the naming scheme _layer3 could be better :). But what I don't see is how that commits restores it. Because it doesn't remove the db.c dependency: from vty_interface.c you still access the DB.
hehe, that is just a human error... :)
z.
-
Holger Freyther -
Sylvain Munaut