**** BEGIN LOGGING AT Fri Dec 28 02:59:58 2018 **** BEGIN LOGGING AT Fri Dec 28 03:14:53 2018 **** BEGIN LOGGING AT Fri Dec 28 05:26:53 2018 **** BEGIN LOGGING AT Fri Dec 28 05:33:15 2018 Dec 28 11:19:03 a key has fallen out of my hwkbd Dec 28 11:19:39 what's a good glue to put it back on? i don't want tomake a mistake and use something that might make matters worse Dec 28 11:20:07 order replacement kb inlet? Dec 28 11:20:44 not practical in my case Dec 28 11:22:19 the key is not lose. just need the right glue to stick it back on Dec 28 11:23:01 *lost Dec 28 11:24:34 i see replacement kb for ~4 usd Dec 28 11:26:42 h3droid is only for allwinner H2+/H3 soc Dec 28 11:26:48 erm, -ECHAN Dec 28 11:43:21 sicelo: I would go with 3M products or some other syanoachrylate glue. Dec 28 11:44:28 or some two-component glue, but I think that would be messier. Dec 28 12:14:44 Trying to merge a list of txt files into one stripping the first line of each. However it would appear bash is processing wildcards not in filename order so output is not in date order. Anyone think of a solution without listing a string of filenames? Dec 28 12:15:14 bash required? Dec 28 12:15:37 nope desktop I think it's multithreading the tail command Dec 28 12:16:54 some scripting language might be nicer if it's complicated Dec 28 12:17:07 can't you sort them after merging? Dec 28 12:18:06 Tried just tail +2 and piping find into xargs tail +2, files seem to be processed in eights (8 core desktop) Dec 28 12:18:40 hmm, internet says globbing guarantees alphabetical order Dec 28 12:19:29 I expected bash to maintain my intentional file order of foofile00 foofile01 etc Dec 28 12:19:47 in php i would just populate array of arrays, drop first element in each subarray and iterate on final list Dec 28 12:20:30 (yes, i'm a fan of php-cli) Dec 28 12:20:52 I think I would use java for that task. Dec 28 12:21:30 (just because we have had exhaustive java courses) Dec 28 12:22:22 Just running straight from the shell, I suppose if there is no straight forward way I could list the array of 60 odd files with tab complete... Dec 28 12:22:55 lets hear what exactly you are trying to do? Dec 28 12:23:25 btw. sorting depends on locale set Dec 28 12:23:50 ah locale Dec 28 12:26:57 so for example "tail -q -n +2 *.txt >> conc.csv" Dec 28 12:27:17 hmm Dec 28 12:27:27 tail would concatenate then cut probably Dec 28 12:27:54 you probably need for each in *.txt; do tail -q -n +2 >>conc.csv; done Dec 28 12:27:58 also "find . -name "*.txt" | xargs -n 1 tail -n +2 >> concx.csv" Dec 28 12:28:13 erm. Dec 28 12:28:25 for each in *.txt; do tail -q -n +2 "$each" >>conc.csv; done Dec 28 12:28:47 unless i misunderstood your question Dec 28 12:30:58 nope perfect, you understood exactly Dec 28 12:31:25 the power of a loop, er Dec 28 12:32:53 check if it globs the files in the right order Dec 28 12:33:00 otherwise you could simple replace it with sort Dec 28 12:33:18 ie. for each in `ls *.txt|sort`; do .. Dec 28 12:33:36 (as long names dont have spaces or weird chars Dec 28 12:35:46 Output is in correct order :) Dec 28 12:36:00 yay for cmdline! Dec 28 12:36:26 \o/ Dec 28 12:59:14 KotCzarny: do not simplify every code you get :E https://github.com/kubernetes/kubernetes/blob/ec2e767e59395376fa191d7c56a74f53936b7653/pkg/controller/volume/persistentvolume/pv_controller.go Dec 28 13:00:42 i see a lot of missing elses :P Dec 28 13:05:12 the shuttle flies too fast - I can't see :E Dec 28 13:13:52 totalizator: my shuttles are usually compressed into high velocity bullets that pierce through anything Dec 28 13:16:54 I'll stick with my bike, thanks Dec 28 13:28:10 After Vajb mentioned merging first I came up with this solution :) "cat *.txt >> conc.csv && sed -i '/String/d' conc.csv" Dec 28 13:28:25 Don't know why i didn't think of that earlier... Dec 28 13:28:49 or just grep -v ? Dec 28 13:33:04 Possibly I seemed to gravitate towards sed -i. Seems cat works in file order tho whereas tail didn't, makes sense. Dec 28 13:35:20 glob sorts Dec 28 13:35:26 per definition Dec 28 13:36:05 why not pipe? cat *.txt | sed -i '/String/d' >conc.csv Dec 28 13:36:21 err without "-i" Dec 28 13:36:55 Useless use of cat ;) Dec 28 13:37:09 * KotCzarny mreoows Dec 28 13:37:25 prolly sed '/String/d' *.txt >conc.csv also works Dec 28 13:37:25 inz, cat is nice as a pipe starter Dec 28 13:37:35 It is usually more readable, yes Dec 28 13:37:51 easier to modify ending params than to move the cursor back further Dec 28 13:37:59 so it has its uses Dec 28 13:40:48 a bit offtopic, what triggers gsm network to send internet/mms configuration? Dec 28 13:41:43 location register server does that, on sensing a new IMEI on an account Dec 28 13:41:57 or on sensing a roaming user Dec 28 13:42:03 afaik Dec 28 13:43:01 DocScrutinizer05: so putting card into another phone, letting it go into network, then putting card back should work for sending config again Dec 28 13:43:18 unless they keep a list of seen imeis per user Dec 28 13:43:19 It was due to finding cat was processing the files in the correct order but it seems just sed works fine too. Dec 28 13:44:11 now i need another phone to confirm, hrm Dec 28 13:51:44 Here once SIM is seen on the network it doesn't prompt you to download config info again unless you request it from the network. Dec 28 13:52:03 KotCzarny: depends, prolly not tsince the "old" IMEI gets stored Dec 28 13:52:30 yes, keeping list of seen IMEIs seems to be exactly what they do Dec 28 13:52:48 s/SIM/IMEI/ ^ is most likely Dec 28 13:54:04 once that IMEI had been seen on that SIM/account, it won't usually get a second SMS Dec 28 13:54:44 but I'm not completely sure, done too few tests to have a sufficient sample base Dec 28 13:55:41 I also guess there must be some "aging" Dec 28 13:56:08 dont have another phone handy atm, but will check in an hour Dec 28 13:56:49 after all you could have nuked the settings by inserting another SIM Dec 28 13:57:31 but it could've been different network, which your home network wouldnt know about Dec 28 13:57:56 so it might be bound to sim+imei combination Dec 28 13:57:56 sp maybe they send a SMS when they didn'T see that account|IMEI tuple for >1 week or whatever Dec 28 13:58:32 so* Dec 28 13:59:02 SIM=IMSI=account, yes Dec 28 13:59:47 yup Dec 28 14:00:17 i just find it weird it wasnt standardized and left for user side Dec 28 14:02:05 most providers offer a web service to trigger semding of config SMS Dec 28 14:02:24 "register new phone" or somesuch Dec 28 14:02:45 the one i'm currently helping with doesnt Dec 28 14:02:54 hmm Dec 28 14:03:04 or at least they dont expose that information anywhere Dec 28 14:03:16 ask their hotline Dec 28 14:03:40 also, it's a virtual operator Dec 28 14:04:09 could also simply be a SSC/USSD, or a SIM with "cinfigure internet" to 666, or whatever Dec 28 14:07:38 well, either they operate on their own network ID (2 digits country + 3 digits Provider-ID) or you are using then genuine network operator's location register and thus should ask their hotline Dec 28 14:09:05 what does the phone show? "virtual" üprovider name, or genuine network operator name? Dec 28 14:09:06 nah, they use real operator network, numbers are probably routed inside the network by operators using shared database Dec 28 14:11:53 they say it shows suboperator name (ie. different than real operator) Dec 28 14:20:30 then they use their own database, either on own hardware or virtualized on genuine operator's Location Register Dec 28 14:21:12 You can either text a number or use the web portal to resend the config here. Dec 28 14:23:26 Also customer service can force the prompt to appear again. Dec 28 14:24:07 i'm looking for an automatic way, preferably without customer service call, people are allergic to it Dec 28 15:26:15 oh, so you ask how providers *should* do it? Dec 28 15:28:22 keep a list of IMSIs in Location Register user account. Delete those that had not been seen logged in for a week or two. Send a SMS to each new IMSI Dec 28 15:29:37 allow manual trigger of SMS sending. Ideally by SMS and by web interface Dec 28 15:31:12 rate-limit to max 10 config SMS / day Dec 28 15:31:41 to defeat all sorts of hiccups and exploits Dec 28 15:31:45 yum, manual trigger with some service code would be best Dec 28 15:32:31 SMS "new" to 666 Dec 28 15:32:34 currently its up for operator to setup such codes, which varies wildly between them Dec 28 15:33:17 or dial *NEW-INTERNET# Dec 28 15:33:29 and its even worse with MVOs Dec 28 15:33:39 better: *NEWPHONE#" Dec 28 15:33:51 or just INTERNET Dec 28 15:33:54 and MMS Dec 28 15:34:00 which some of them do now Dec 28 15:34:04 but not all Dec 28 19:11:06 are there any good tools to determine which bands a phone supports? Dec 28 19:11:32 like i have an e66 and i want to figure out if it's the US version or non-US version **** BEGIN LOGGING AT Sat Dec 29 02:53:33 2018 **** ENDING LOGGING AT Sat Dec 29 02:59:58 2018