**** BEGIN LOGGING AT Tue Sep 18 03:00:00 2018 Sep 18 10:15:47 you don't need to subscribe Sep 18 10:16:01 but if you seriously care about getting something fixed, it's the way of getting it done. Sep 18 13:01:14 Trying to understand what specifications of the router makes my macbook pro link wifi at 300mbps or 450mbps. Connecting the laptop to a UBNT AC-LITE connects 5ghz at 450mbps Sep 18 13:01:48 Connecting the same laptop to some other IPQ40XX based router connects at 300mbps. Why is that? That is it the IPQ40XX is missing that UBNT has? Sep 18 13:03:59 barhom: 2 streams vs 3 streams (of theoretical 4 for 11n) Sep 18 20:53:58 jow: please re-assess https://bugs.openwrt.org/index.php?do=details&task_id=1840 Sep 18 20:54:18 jow: I've submitted a patch which will hopefully be accepted tomorrow by lynxis Sep 18 20:54:41 jow: https://paste.pound-python.org/show/2xheq2Rc7Rl79fq74lEq/ Sep 18 20:54:46 jow: or by you today :-) Sep 18 21:59:19 you would have had this much faster if you'd just spent the energy posting it properly yourself. Sep 18 23:26:14 https://github.com/openwrt/mt76/issues/192, I would say one of the last issues remaining on the mt7603e chip before I call it stable. If you get the same problems please report. Sep 18 23:31:16 I was just running binwalk on certain openwrt images just to see how the layout of the final compressed images. I downloaded some 18.06.1 release images specifically for two routers, D-Link DIR-825, TRENDNet TEW-673GRU. The top level layout looks as follows: ```uImage Header | LZMA Compressed Data | SquashFS Filesystem```. Sep 18 23:32:08 number_three_: why not jsut look at the makefiles to see how they're laid out? Sep 18 23:33:24 I guess yeah, the main question I had was when I decompress the LZMA data which contains the embedded kernel image there seems to be ~2-3MB data blobs with low entropy before the kernel image. Sep 18 23:33:40 Could you tell me what purpose do these data blobs serve? Sep 18 23:34:26 I could look at the Makefiles but figured before I try to reverse-engineer the entire build process, it might make sense if someone knew the purpose of these data blobs in the LZMA data already. Sep 18 23:35:03 barhom: Try it as a client before you pronounce it stable Sep 18 23:41:30 requested pricing for the GW6404 dev kit. curious :) Sep 18 23:42:10 stintel: Cavium. That's going to be scary. :) Sep 18 23:42:22 cpu model : Cavium Octeon+ V0.1 Sep 18 23:42:29 but it's getting slow! Sep 18 23:42:45 and iirc octeontx target was added to git master a while ago Sep 18 23:43:04 I meant the price Sep 18 23:43:20 I've never seen a Cavium CPU in anything not painfully expensive Sep 18 23:43:27 oh :) Sep 18 23:43:32 http://www.gateworks.com/product/item/newport-gw6400-single-board-computer Sep 18 23:43:38 I saw Sep 18 23:43:55 my APU2 turns out to be too slow when running suricata Sep 18 23:44:00 in inline IPS mode Sep 18 23:44:10 so looking for other options Sep 18 23:45:08 that fitlet2 people were talking about in #openwrt today looked interesting too Sep 18 23:45:14 but not sure if I want to keep supporting intel Sep 18 23:45:44 and that gw thing can be poe powered <3 Sep 19 01:17:19 stintel: where exactly is the bottleneck? **** ENDING LOGGING AT Wed Sep 19 03:00:00 2018