Fixed
Status Update
Comments
js...@android.com <js...@android.com>
mi...@gmail.com <mi...@gmail.com> #2
доколе?
mi...@gmail.com <mi...@gmail.com> #3
+1
There should be possible to open source in any installed (or minSdk for project) android API version, because compileSdk is not only one used for run the app.
Why it takes so long to release source code to SDK when the sources was already pushed to GIT?
There should be possible to open source in any installed (or minSdk for project) android API version, because compileSdk is not only one used for run the app.
Why it takes so long to release source code to SDK when the sources was already pushed to GIT?
gu...@gmail.com <gu...@gmail.com> #4
Fuck this issue, always happened from android 23. Fuck this. Not friendly to developers
rg...@google.com <rg...@google.com> #6
yep, no 26 sources no funny ;(
all the bunnies gonna cry ;((
all the bunnies gonna cry ;((
gu...@gmail.com <gu...@gmail.com> #7
gu...@gmail.com <gu...@gmail.com> #8
+1
pe...@googlemail.com <pe...@googlemail.com> #9
While debugging, we consider this behavior WAI. That's because stepping through code that doesn't match SDK version of the device is not a good experience. However, we can consider fallbacks to older sources when navigating to definitions while editing your code.
ar...@rfc2549.org <ar...@rfc2549.org> #10
@9 Android needs IPv6 NAT support for VPNService and IPv6. Unfortenately the Android kernel that are used to toaday lack IPv6 NAT.
pe...@googlemail.com <pe...@googlemail.com> #11
@10 thanks for the information. This could have taken into consideration before changing a framework for building VPN clients that used to work just fine. Personally I would never have missed NAT over VPN. I'd strongly advice to never open a VPN on Android device level for other devices via tethering!
ab...@gmail.com <ab...@gmail.com> #12
Thanks for the advice.
we...@gmail.com <we...@gmail.com> #13
Is it possible to make an xposed module that could fix this issue?
ar...@rfc2549.org <ar...@rfc2549.org> #14
@13 I have have xposed you probably have root; if you have root you can also use the workaround of @6
ec...@gmail.com <ec...@gmail.com> #15
This issue has been around for over a year. Why has it not been resolved
je...@gmail.com <je...@gmail.com> #16
I experienced this issue on Android 4.4.4 on a Nexus 5 with the strongSwan app.
Now after upgrading to Android 5.0, both IPv4 and IPv6 work inside the tunnel, even when I am on an IPv4–only network.
Now after upgrading to Android 5.0, both IPv4 and IPv6 work inside the tunnel, even when I am on an IPv4–only network.
rg...@google.com <rg...@google.com> #17
Jeremy, you confirm that this issue is fixed on 5.0 for you?
je...@gmail.com <je...@gmail.com> #18
As far as I can tell, yes.
IPv6 inside the tunnel was broken on 4.4.4, but works on 5.0. Same version of the strongSwan app, and the same server configuration.
You may want to collect more data before marking this one as closed though. :-)
IPv6 inside the tunnel was broken on 4.4.4, but works on 5.0. Same version of the strongSwan app, and the same server configuration.
You may want to collect more data before marking this one as closed though. :-)
sl...@gmail.com <sl...@gmail.com> #19
IPv4 and IPv6 are now routed inside an OpenVPN tunnel. Like Jeremy, was not working since 4.4.x and now works fine in 5.0.
su...@gmail.com <su...@gmail.com> #20
So apps like drony, psiphon, shadowsocks, are working okey on upgraded 5.0 devices?
Not that is related to ipv6 routing but something is changed on 5.0 that makes problems.
Not that is related to ipv6 routing but something is changed on 5.0 that makes problems.
rg...@google.com <rg...@google.com> #21
Supp.Sandrob, what problem do you mean in #20?
Thx Jeremy. Glad we got something right on this. We did reach out to a number of VPN makers to verify things this round - maybe it helped.
Thx Jeremy. Glad we got something right on this. We did reach out to a number of VPN makers to verify things this round - maybe it helped.
Description
2. Connect to a VPN providing IPv6 connectivity
3. All IPv6 routes (apart from those directly on the interface) are unreachable.
I think this bug is realted to MTU. I think what happens is:
The route is lookuped up in the main route table. It is not found in there giving destinationen unreachable.
Therefore the packet is never sent, never gets the fwmark mark and rule
from all fwmark 0x3d lookup 61
cannot match the packet.