**** BEGIN LOGGING AT Tue Jul 24 03:00:01 2018 Jul 24 03:52:05 * brolin_empey bought a 28-inch 4k UHD video monitor for use at home, am using it to write this message. Jul 24 04:23:25 brolin_empey: model 4 looking on the net Jul 24 04:37:53 sunshavi: It is a Samsung LU28E590DS/ZA. Jul 24 10:16:58 i don't use 2-factor authentication for any of my accounts :-/ Jul 24 10:18:25 but i want to. however i do not want to use SMS. i only have an N900. Jul 24 10:19:44 any ideas on solutions/applications/etc. Jul 24 10:22:54 sicelo: what are you talking about? Jul 24 10:23:11 (I'd guess gmail, but ...?) Jul 24 10:26:43 yes. seemingly some of the support authenticator apps instead of sms Jul 24 10:27:13 I use gmail with mutt, you can add a password per application Jul 24 10:27:29 and use it exclusively with the specified application Jul 24 10:34:11 thanks bencoh. i will try that for gmail. will check if something similar is available for the others (facebook, github, twitter, etc.) Jul 24 10:48:55 If by "authenticator" you mean the time-based OTP authentication used by "Google authenticator", I just run a Java implementation of that on my computer. Jul 24 10:49:11 that will surely also run fine on N900. Jul 24 10:51:13 http://blog.jcuff.net/2011/02/cli-java-based-google-authenticator.html Jul 24 10:52:04 Yeah, it seems to just be a Main class for use with the code Google uses in their Android application. Jul 24 10:59:21 nice. Jul 24 10:59:43 i had been looking at oath-toolkit. no idea if that would do Jul 24 11:00:10 oauth*? That's quite a bit different to authenticator. Jul 24 11:00:44 though maybe some systems let you pick and choose which authentication mechanisms can be used, dunno. Jul 24 11:01:40 https://github.com/malept/oath-toolkit Jul 24 11:01:45 I've been wanting to use OAuth (and hence SMS 2FA) for mutt/gmail, but haven't got round to looking enough into it. Jul 24 11:02:00 afaik, I'd probably have to just hack mutt. Jul 24 11:02:24 Oh, okay, not OAuth then. Jul 24 11:22:18 bencoh: per app password seem to need my phone number after all. or i am missing something? Jul 24 11:23:17 sicelo: are you sure it's not just to get the password the first time? Jul 24 11:24:25 i guess yes, that. had been wishing/hoping for something to not use phone number ever :) Jul 24 11:24:33 but yeah, does look like i am dreaming Jul 24 11:31:20 sicelo: err, could be, yeah Jul 24 11:31:41 but that's just to register iirc Jul 24 11:38:21 looks like oath-toolkit is also the right tool for 2FA needs. wonder if it will compile for n900 Jul 24 22:42:25 back to my earlier questions about 2FA, looks like there already is an OTP tool for N900 Jul 24 22:42:29 http://maemo.org/packages/view/otp/ Jul 24 22:42:47 no idea how well it works (or how secure it is) Jul 24 22:53:26 TOTP i assume? Jul 24 22:53:44 ah yes Jul 24 22:54:10 there is oathtool aswell, at least in debian Jul 24 22:54:36 RFC 6238 TOTP is what 'google authenticator' uses Jul 24 22:54:56 you can just copy the seeds out of it , feed them to oathtool and get a key to use on websites for 2FA Jul 24 22:55:26 yes, i have seen oathtool, and one of the days i should try compile it for maemo through sb Jul 24 22:55:51 i don't have google authenticator anywhere. Jul 24 22:55:52 > oathtool --totp -b $CODEYOUGOTFROMTHEWEBSITE Jul 24 22:56:19 it gives you a QR code to scan usually, the code from that is what you'd use to make a new key each time Jul 24 22:56:28 and 'top' n900 doesn't seem to be working. well ... at least facebook says codes are invalid Jul 24 22:56:51 maybe you need to set it to TOTP instead of HOTP? Jul 24 22:57:08 its not default on oathtool either Jul 24 22:58:02 it only allows switching between HOTP & MOTP, and with HOTP, you can switch between 'time-based' (which i assume is TOTP) and 'event-based' Jul 24 22:59:20 weird, if you have a debian box available somewhere you can test oathtool there Jul 24 22:59:34 apt install oathtool Jul 24 22:59:35 :) Jul 24 23:00:02 yes, left my laptop at work. already installed oathtool on it .. just need to put it to use Jul 24 23:00:22 i hope it will build nicely for maemo as well Jul 24 23:00:35 quite sure, i doubt it has much dependancies Jul 24 23:01:34 maybe i should already look in older debian repos .. a few packages work ok from there on maemo Jul 24 23:25:54 >> Optimally, all RAM should be used, for cache if nothing else. << With Windows 7 Enterprise 64bit, if Task Manager shows high RAM usage (like, 13GB out of 16GB), it means "ultra-slow computer, restart some of those programs". But then, this slowness may also coincide with high (like, Jul 24 23:26:51 1725914275) "Page Faults" (whatever they are) for processes. Jul 24 23:28:54 (oathtool - at least all the ones on debian repos need higher libc6 than maemos ..) Jul 25 00:02:19 Assuming "page faults" means "major page fault", that means you're copying things in and out of swap, which gets bad when you've run out of non-freeable RAM. Jul 25 00:03:04 I suspect Windows probably doesn't include disk cache in its "used memory" count, but theoretically, that extra memory should still be used. Jul 25 01:20:40 Maxdamantus : Is it possible to have 0 "major page faults"? Jul 25 01:53:44 Oksanaa: sure, if there's no swap. Jul 25 01:54:28 and maybe no disk; not sure if memory mapped files' page faults are considered major. Jul 25 01:54:52 (swap/pagefile) Jul 25 02:45:05 Why do things copied in and out of RAM get bad when computer runs out of non-free-able RAM? **** ENDING LOGGING AT Wed Jul 25 03:00:02 2018