Very interesting. Thanks for your review!
Is that still on sale, do you know?
And would you go so far as to say the hardware is possibly actually
identical to the VMG1312-B10A, with just a different software load ? You said more businessy?
I might be really interested provided I could get Johnson, to whom I owe
so much already, to do a port of his custom firmware. If the unit has some advantages, I would have to get him one to test on, if he were willing. It’s kind of taking the pee out of him; don’t want to abuse his generosity further.
What kind of line were you using, and what performance spec ? I’m very interested by what you said about faster upstream. For me a DLink DSL-320B-Z1, which I compare against my ZyXEL VMG1312-B10A on my ultra-long 3.0/0.5 Mbps sync rates ADSL2 line (~7300 m), is slightly faster on the upstream, by 5-10% but the ZyXEL has a small advantage on downstream, maybe 0-10% advantage to the ZyXEL B10A over the DLink out of ~2.9 Mbps d/s. I have no idea why there is this upstream advantage in the Dlink. The Dlink’s usual game is to run at an SNRM which is way below target and thus to get a speed advantage that way. This is certainly true for the downstream but might be true in the upstream as well, which I would have to take another look at to refresh my memory. But then the question simply moves to why does it run too hot, below the target in upstream then (ditto on downstream). The ZyXEL is still faster even though the DLink runs at too low an SNRM, but it still is faster downstream. This is because it has Adsl2 PhyR L2-RETX error correction (like G.INP) on the downstream (only) since luckily the exchange is protocol compatible with this proprietary protocol and this causes a massive improvement in reliability and speed. It can run at very low SNRM without any problem at all.
I’m extremely interested in your report of improved upstream as this is something I’m desperate for. I don’t know the context: it’s an extra 0.5Mbps upstream, but out of what/compared with what ? If I had 0.5 Mbps extra u/s * 4 lines then that ~2Mbps less bonding inefficiency would rock my world, as I only have ~1.3-1.55 Mbps u/s
in total anyway, and that’s 4 lines combined. So very interesting.