**** BEGIN LOGGING AT Wed Jun 13 03:00:02 2018 Jun 13 03:00:39 will it be in repos when you get it ready? Jun 13 03:00:50 jonwil : Great :-) Fahrplan, right? Remind me to try it - I call 131500 from time to time, and it would be nice to have the app working instead of voice-calling every time a bus runs late. Jun 13 03:00:51 i mean the ff24 Jun 13 03:02:55 Will Gtk+Hildon apps benefit from TLS1.2 support in openssl? I often see MicroB complaining about lack of common encryption algorithms on https:// websites. Jun 13 03:06:27 microb doesn't use openssl at all Jun 13 03:07:22 ff24 is a long way from being done Jun 13 03:12:05 Oksana: I have no idea about the working of the 131500 parser but its not using SSL by the look of it. Jun 13 03:12:56 All the better, in the sense that I don't need to think about update reaching repositories. Installing Fahrplan now... Jun 13 03:14:13 Which Fahrplan version are you going with then? Jun 13 03:16:02 0.0.37 Jun 13 03:16:19 Yeah that should be fine for 131500 Jun 13 03:18:00 * Oksana wonders what is Translink - Australia Jun 13 03:18:01 As for OpenSSL and other apps, things need to be recompiled and link to OpenSSL 1.1 to get TLS 1.2 support in there Jun 13 03:18:08 Translink is the network I use Jun 13 03:18:10 the one I added Jun 13 03:18:13 QLD Jun 13 03:18:59 Ah, okay. I was wondering about Melbourne, will be travelling here for a day or three in July. Hopefully. If everything goes well. Jun 13 03:19:33 But I will most likely be walking more than anything else. Jun 13 03:22:35 * Oksana doesn't think that "launch GPS on start up" is working Jun 13 03:22:44 I went to Melbourne in 2010 for a LEGO show, it was fun and I would like to go back Jun 13 03:23:02 Yeah GPS isn't always working for me Jun 13 03:23:07 And I haven't been able to debug why Jun 13 03:23:32 Throbber animation is nice, but no idea how to provide a valid station. It always says "Station not found" Jun 13 03:23:55 NSW wasn't written by me Jun 13 03:24:01 so I have in idea why its not working Jun 13 03:24:37 Hmm... Link to code? Would like to have a look... Jun 13 03:25:20 C++ , right? Going by parser131500ComAu::getStationsByName Jun 13 03:25:36 Yeah its C++ Jun 13 03:26:07 * Oksana is rusty, but could at least look for something Jun 13 03:28:46 https://github.com/jonwil/fahrplan/ has the current code Jun 13 03:29:12 My guess is that whatever backend URL NSW is using has changed and its no longer compatible Jun 13 03:36:40 * Oksana sighs and goes to log in to GitHub Jun 13 03:37:06 It's not 131500.com.au anymore, it's TransportNSW.info, and it doesn't care about am or pm Jun 13 03:37:37 Ok so it looks like it needs a change to the URL its using Jun 13 03:38:14 A solid change. Let me see... Jun 13 03:38:59 It doesn't have separate variables like &itd_itdTimeHour for Date, Hour and Minute - it just has &dateTime=201806131340 Jun 13 03:41:25 * Oksana is forking this file, to submit pull request later. Hopefully. Jun 13 03:43:38 If you come up with a fix, I will push a new deb file (since I imagine you probably dont have a dev environment :) Jun 13 03:46:32 Would be nice, thank you :-) I just hope there isn't too much to change Jun 13 03:48:09 Its not hard to build a new deb for Fahrplan Jun 13 03:48:27 Although needless to say the new deb I build will need the new QT and new OpenSSL :) Jun 13 03:49:25  Jun 13 03:53:09 transportnsw.info seems to use https. While 131500.com.au used to be http. Would http->setHost(url.host(), QHttp::ConnectionModeHttp, url.port() == -1 ? 0 : url.port()); work still? Jun 13 03:53:47 Or do I need to change to http->setHost(url.host(), QHttp::ConnectionModeHttps, url.port() == -1 ? 0 : url.port()); ? Jun 13 04:01:57 Darn... Is html parsing really the ideal way to handle results of the journey request? Jun 13 04:16:52 * Oksana looks around for a better API, while simultaneously attempting to decode large amount of nested divs Jun 13 04:26:06 * Oksana curses quietly Jun 13 04:27:51 JSON is a far better way for an app to receive results from transport website, compared against bulky HTML. However, this json uses ancient formatting of the URL, which means... Lots of Ctrl+Z Jun 13 04:34:25 No idea Jun 13 04:34:50 I dont know how the website for transportnsw works Jun 13 04:35:11 As for SSL, take a look at what I have done in parser_translink.cpp to see how I did it. Jun 13 04:35:20 http://talk.maemo.org/showthread.php?p=1545412#post1545412 Jun 13 04:35:44 New fahrplan with the working Translink backend Jun 13 04:35:59 http://talk.maemo.org/showthread.php?t=100352 is the new QT Jun 13 04:36:24 Any idea how to send header with https link? For authentication? Jun 13 04:46:45 Have a look at how I did it via QHttpRequestHeader Jun 13 04:47:40 * Oksana nods and thinks that Translink, thankfully, also has json Jun 13 04:48:19 Yes it does json parsing Jun 13 05:44:34 Is there an intentional connection between Steve Jobs reportedly owning an air-cooled vehicle (Volkswagen Type 2) and wanting to avoid use of a fan in the computers sold by his company (Apple Computer Inc.)? Jun 13 05:45:55 Fans are noisy, dusty, and require maintenance. Jun 13 05:46:35 What's the difference between GET and POST, when using QHttpRequestHeader ? Jun 13 05:48:28 POST is for when you have data to attach (i.e. form post data) Jun 13 05:48:39 Which one you use depends on what the site itself needs Jun 13 05:49:12 If you have a browser with a web developer toolbox or something, you could sniff the connections being made and see whether its using get or post... Jun 13 05:49:23 A fan can be inaudible and not require maintenance for the expected service lifetime of the device containing the fan. Jun 13 05:49:26 Site says "curl -X GET --header 'Accept: application/json' ..." Jun 13 05:50:05 Fan doesn't sound like it's waterproof, or foolproof Jun 13 05:51:28 Is not a propeller basically a fan that moves water instead of air? Jun 13 05:52:14 I mean on a watercraft, not an aircraft. I thought it is at least somewhat analogous. Jun 13 05:52:30 Yes, but then you need to waterproof/insulate the connection between fan-chamber (exterior, air or water or dust) and interior (PCB and electricity), while still having excellent heat conduction. Jun 13 05:53:36 Otherwise, it does sound interesting - a lighter-than-water laptop which can use its fans for propelling itself in the water. Jun 13 05:53:57 Or phone ^ Jun 13 05:54:39 But then fan would need gears for switching between air and water. Because these two liquids are quite different. Jun 13 05:54:48 fluids * Jun 13 05:55:46 Its great to have the current OpenSSL release running on Maemo Fremantle Jun 13 05:56:22 Although the idea that someone had a while back to produce a shim library to map 0.9.8 to 1.1.0 isn't really possible. Jun 13 05:56:26 I wonder if Rudolf Diesel used a motorcycle because his motorcycle would be a Diesel motorcycle in a sense. Jun 13 05:57:22 Not sure which of the closed packages that are linking to the old OpenSSL present a security risk and which dont Jun 13 05:58:18 Some can probably be dropped without too many people getting annoyed (I doubt anyone is using a N900 to talk to an ActiveSync MS email server anymore...) Jun 13 06:00:52 What is loop.exec(); for? I think I got how header works with GET method. Jun 13 06:01:32 jonwil: I am using my n900 for exchange mail Jun 13 06:01:54 Ok so as-daemon is not obsolete then Jun 13 06:01:59 mhm Jun 13 06:02:17 I did use Mail for Exchange (Active Sync?) recently. I am using IMAP now, though. But I still sync with Microsoft-based Calendar and Contacts. Jun 13 06:03:52 IMAP works fine (I think), but I was very annoyed when an email I sent from N900 didn't show up in Microsoft-based email's "Sent" folder. It's like, if my N900 breaks down, then I will have no way to see which letters I sent? Why doesn't server keep a copy in "Sent" folder? Jun 13 06:04:05 In any case I have clearly shown that its possible to mix 1.1.0 and 0.9.8 on the one system without any problems. Jun 13 06:04:16 * Oksana nods Jun 13 06:04:48 The only question therefore is which of the packages linking to 0.9.8 9 (closed or open) represent a real security risk and therefore need to be dealt with. Jun 13 06:05:24 That's what the major version changes are for. Firefox clearly disregards this, since I cannot even remember which Firefox versions are major milestones (like, dropped support for XUL interface, NPAPI plugins, and so on) Jun 13 06:06:00 Is TLS truly useful considering that the encryption can be subverted (maybe not the best verb to use) by a government? Jun 13 06:06:52 brolin_empey : How so? Do they need server's agreement (backdoor) for that? Or is it a genuine weakness in TLS algorithm itself? Jun 13 06:07:42 * Oksana thinks that it would be nice to have GPG-or-something commonly used for emails, but not sure if Modest supports this Jun 13 06:08:34 I have yet to see any compromise of the encryption used with TLS or any other major internet security protocol. The weaknesses mostly have to do with the fact that the current model of certificate authorities used for SSL/TLS is totally broken and there isn't much interest from anyone to actually turn the proposals to replace CAs from proposals and presentations snd documents into actual... Jun 13 06:08:35 ...real-world code... Jun 13 06:10:49 So we end up with bandages on top like public key pinning and other stuff rather than moving away from CAs altogether in favor of something better. Jun 13 06:12:13 Freemangordon: Are your clones of connui-internet and connui-wlan complete? Do they contain anything Leste specific or could they work as a drop-in replacement on Fremantle? Jun 13 06:13:15 Oksana: Plug-ins for the NPAPI use native code, which, for consumer stuff in the 1990s and 2000s, probably mostly means x86 and maybe PowerPC, not ARM. If you want to use a closed-source plug-in not released for ARM then NPAPI support on an ARM computer, such as the N900, is not useful as far as I know. Jun 13 06:15:20 Now that I got Fahrplan going again, the next plan I guess is to keep working on Firefox 24. Unfortunately that's a much trickier job than OpenSSL, QT and Fahrplan was. Jun 13 06:16:57 I suspect that part of the motivation to replace NPAPI was because NPAPI plug-ins use native code. Jun 13 06:17:26 jonwil: not 100% sure, but the should be drop-in replacement Jun 13 06:17:50 But are they complete? Jun 13 06:18:01 no eapd and wps Jun 13 06:18:21 they're replaced with wpa-supplicant Jun 13 06:18:34 jonwil : What is loop? I don't understand how loop.exec(); works when there is no definition of loop anywhere. Unless it's a magical keyword, like for and wait and if? Jun 13 06:18:44 loop is declared in another file Jun 13 06:18:54 parser_abstract.h Jun 13 06:18:56 but it is Wizzup with the details as he was working in it lately Jun 13 06:19:03 Huh, okay. Thank you :-) Jun 13 06:19:22 Basically loop.exec will wait until one of the signals is signaled Jun 13 06:20:26 The connect lines in parserTranslink::parserTranslink are part of it Jun 13 06:20:36 httpRequestFinished then causes the loop to stop executing Jun 13 06:20:45 and progress returns to the line after the loop.exec statement Jun 13 06:20:46 jonwil: but yes, they are complete, we have working wifi (more or less) on leste with icd2 Jun 13 06:20:49 httpRequestFinished is also in parser_Abstract Jun 13 06:25:16 Oksana: if you copy what I have done for Translink and adjust the parameters and URLs and things as appropriate, it should work Jun 13 06:26:59 jonwil : Thank you, working on it. Why does https://github.com/jonwil/fahrplan/blob/master/parser_translink.cpp have "bool ok;" in lines 256 and 293? Jun 13 06:27:40 Probably just some leftover I forgot to remove Jun 13 06:31:48 38363 lines in one json... What is TransportNSW thinking? Jun 13 06:33:16 Thank god QJson makes it so easy to parse that stuff and grab the values needed Jun 13 06:34:01 :) Jun 13 06:34:34 at least they server that json with compression? Jun 13 06:34:38 *serve Jun 13 06:35:07 Oksana: It seems that “encryption” was a poor choice of term. I meant what jonwil was saying that the infrastructure of TLS has problems. Well, actually, even if the CA system was replaced by a different system, is not it still possible for a government to require network/service providers to provide access to the government? Jun 13 06:35:18 No idea. It's application/json, interactive example shown in https://opendata.transport.nsw.gov.au/node/601/exploreapi#!/default/tfnsw_trip_request2 Jun 13 06:36:12 brolin_empey : That's why peer-to-peer would be nice. But then, TLS can be used for P2P, as well, not only for client-server connections? Jun 13 06:37:32 Is there a pastebin capable of handling such an amount of data? Jun 13 06:38:06 Oksana: the "issue" isn't p2p vs client-server connections, but p2p trust vs the current star-like "chain" of trust Jun 13 06:39:32 bencoh : Yes, current star-like chain of trust is strange. I may trust a website, to an extent, but I have no idea about issues of certificates. Jun 13 06:40:37 issuers * Jun 13 06:50:47 {version, systemMessages, journeys: [{rating, isAdditional, interchanges, legs: [{duration, origin, destination, transportation, hints, stopSequence, infos, coords, properties},{}]},{}]} Jun 13 06:52:47 If the systems that have been proposed as replacement for CAs were adopted and if they had good encryption built in from the start, it would be very difficult to get at whatever data is flowing over the pipes even if you were able to somehow obtain the private keys from the server end. Jun 13 06:54:20 The big problem with CAs really is that its possible for any of the many CAs out there to create bogus certificates for websites Jun 13 06:54:26 and use them for man-in-the-middle attacks and things. Jun 13 06:55:18 A new system that prevented such a thing by more closely tying the certificate/key info to the actual owner of the domain would be good. Jun 13 06:56:44 that's the 'authority' part Jun 13 06:57:30 KotCzarny: Yes, good point. Jun 13 06:58:18 >> Waiting for your fork… << Brilliant, GitHub. Jun 13 06:58:32 oksana: ask for a knife too, just in case Jun 13 07:05:18 Many experts have said that the current CA system is broken. Jun 13 07:05:47 Some CAs do the right thing and ensure the certificate is legitimate. Jun 13 07:06:38 But many don't necessarily follow the rules. Or they get hacked (DigiNotar anyone?) Jun 13 07:07:51 Or you get someone like China forcing a CA in that country to generate certificates for their own nefarious purposes. Jun 13 07:07:52 KotCzarny : The same message, still. I think K-Meleon isn't particularly supported by GitHub. Jun 13 07:08:31 What exactly are you trying to do on github? Jun 13 07:09:40 Editing parser_131500comau.cpp. Aka "You’re editing a file in a project you don’t have write access to. We’ve created a fork of this project for you to commit your proposed changes to. Submitting a change to this file will write it to a new branch in your fork, so you can send a pull request. " Jun 13 07:09:58 * Oksana just went to SeaMonkey Jun 13 07:10:33 Replacing CAs with something like DANE or the EFF Sovereign Keys proposal (or the proposals out there for a web of trust where certificates and keys are signed by multiple entities and then your client figures out how trustworthy it is based on who has signed it) would be much better than the current CA system Jun 13 07:17:45 jonwil: I was going to mention the “People’s Republic of China” but I did not but now I do not have to. I do not contract “People’s Republic of China” to “China”, though, because the Republic of China still exists. Jun 13 07:19:01 You mean Taiwain? Jun 13 07:20:14 Maxdamantus: Yes, the Republic of China since 1949. Jun 13 07:30:15 jonwil : Is "QStringList trains" expected to contain "Train, Bus, Ferry" or "T8, E70, MFF" or "Macarthur to City via Airport or Sydenham, PrePay Only - City Wynyard to Manly (Express Service), Circular Quay to Manly" ? Jun 13 07:31:30 Going by "trains.removeDuplicates();", it's probably "Train, Bus, Ferry", otherwise duplicates would have been unlikely. Jun 13 07:31:36 Yeah Jun 13 07:33:17 Its definitely the mode name Jun 13 07:33:28 travelModeName:"train" is what the JSON for Translink returns Jun 13 07:33:40 and that's what the translink code uses to fill in the string array Jun 13 07:34:48 Problem is, modes are specified as integers. Which means, I need a... tulip or something to store the 1=train, 4=light rail, 5=bus, and so on, in. And then whichever easy way to convert integer into the word. Jun 13 07:35:12 Probably QStringList would work. Jun 13 07:35:59 But numbers aren't sequential. They are 1,4,5,7,9,11,99,100,101,102,103,104,105,106 Jun 13 07:37:00 QMap is the way to go here Jun 13 07:37:32 How is it different from std::map? https://stackoverflow.com/questions/1910733/how-can-i-map-an-int-to-a-corresponding-string-in-c-c Jun 13 07:37:53 Its much the same but since its all being done in QT, QMap makes sense Jun 13 07:39:12 QMap modes; ??? Jun 13 07:39:34 Yeah something like that would work Jun 13 07:39:58 NSW does have quite a few mode types Jun 13 07:40:58 * Oksana will continue tomorrow, hopefully. I have no idea whether QMap or QMap is expected here. Jun 13 07:41:11 QMap is the right one Jun 13 07:42:21 * Oksana nods and goes outdoors Jun 13 08:08:56 Anyone here any good with Python programming? Jun 13 10:47:09 In attempt to figure out ssl_error_no_cypher_overlap (shown by both MicroB and Fennec [aka Firefox 17]), I have stumbled upon https://cc.dcsec.uni-hannover.de/ . It's useful for testing a web browser, but how do I know which cyphers are supported by a web server? Say, https://scratch.met.edu/ . Jun 13 10:47:44 https://www.ssllabs.com/ssltest/ Jun 13 10:49:06 Thank you :-) Testing Jun 13 10:49:44 If the site is an intranet site, it won't work though, no idea for a tool to do the same Jun 13 10:50:41 No problem, I don't remember the last time I visited an Intranet site Jun 13 10:51:23 What does A+ grade mean? Jun 13 10:52:05 That they do not support weak ciphers etc Jun 13 10:52:11 click on the address to se more details Jun 13 10:56:29 MicroB supports TLSv1. Scratch supports TLS 1.0-1.3, not a problem. Jun 13 10:56:44 Cipher suites... Jun 13 10:58:48 Scratch supports ECDHE RSA with AES, or weak RSA with AES. MicroB supports ECDH, DH, or RSA. The only common point is RSA. Jun 13 11:01:32 MicroB supports RSA with AES 128/256 SHA1. Scratch supports weak RSA with AES 128/256 CBC SHA. Is CBC the crucial difference? Jun 13 12:19:22 Hm.. is omapfb known to be required to be built-in or something? Jun 13 12:20:35 when trying to load it as a module in 4.16 I just get "no such device" .. irc I got the same thing in the maemo branch of 3.14 some years ago and fixed it by having it built in. Jun 13 12:20:45 s/irc/iirc/ Jun 13 12:20:45 Maxdamantus meant: when trying to load it as a module in 4.16 I just get "no such device" .. iirc I got the same thing in the maemo branch of 3.14 some years ago and fixed it by having it built in. Jun 13 12:21:01 er, n900 branch* Jun 13 12:21:47 * Maxdamantus will probably just try a built-in version tomorrow. Jun 13 12:25:29 maybe check what leste config does Jun 13 13:14:45 Wizzup: i belive, uses the kernel default rx51_defconfig , except oddly i can't see arch/arm/configs/rx51_defconfig in the n9xx-linux repo Jun 13 13:17:05 https://github.com/maemo-leste/n9xx-linux/blob/pvr-wip-4.15.7/arch/arm/configs/rx51_defconfig Jun 13 13:23:23 CONFIG_FB_OMAP2=y Jun 13 13:23:25 Yeah. Jun 13 13:23:39 mainline rx51_defconfig has =m Jun 13 13:24:55 actually, no rx51_defconfig in mainline .. using omap2plus_defconfig Jun 13 13:26:25 FWIW, other configs are here: https://github.com/parazyd/arm-sdk/tree/master/boards/kernel-configs Jun 13 13:26:29 (droid4, n9, n950) Jun 13 13:26:40 ah, so maemo-leste still requires those pvr patches. Jun 13 13:27:57 Good to know, thanks. Jun 13 13:28:02 there is nothing that works without pvr patches :) Jun 13 13:28:06 no powervr* Jun 13 13:28:36 Well, plain fb works without it. Jun 13 13:28:57 right, and you can use maemo leste with plain fb and llvmpipe for h-d, but it'll be unusable ;) Jun 13 13:29:40 * Maxdamantus intends to not use maemo/hildon in the long run anyway. Jun 13 13:29:48 (and even more unusable on n900) Jun 13 13:29:56 Maxdamantus: awww, that's a mean thing to say :> Jun 13 13:30:44 I wouldn't use software like that on my desktop systems, so wouldn't want to (ultimately) on my phone either. Jun 13 13:30:51 Maxdamantus: good luck :P let me know if you find anything usable Jun 13 13:30:59 Maxdamantus: desktop==phone? Jun 13 13:31:13 Wizzup: wmii is okayish on droid4 :D Jun 13 13:31:23 Wizzup: not exactly ==, but I prefer things to be decoupled. Jun 13 13:31:23 until you start to actually use it ;) Jun 13 13:31:38 portrait wmii is a pain Jun 13 13:31:46 Maxdamantus: I think the IT world has been trying to figure this out, so I'm surprise you have :p Jun 13 13:31:50 landscape is fine :) Jun 13 13:31:57 and I don't need the sorts of fancy graphics you get with HD Jun 13 13:32:03 bencoh: yes, but apps don't work that well in general touch wise Jun 13 13:32:15 Maxdamantus: agreed, I don't care about the shaders and effects Jun 13 13:32:40 I use xmonad on my desktops, but would rather use something like wmii/w3m on n900 for memory reasons. Jun 13 13:33:01 Wizzup: yeah, real apps still need to be adapted somehow :) Jun 13 13:33:12 bencoh: we've been trying for what, 10+ years? Jun 13 13:33:17 I'm sure it'll change overnight :P Jun 13 13:33:57 Wizzup: riight :)) Jun 13 13:34:44 seriously though, wmii in itself is okay as a wm (since you have a keyboard), but yeah, it still misses a whole touch-oriented ecosystem Jun 13 13:34:51 * Wizzup uses wmii right now Jun 13 13:34:57 (on my laptop) Jun 13 13:35:05 (same on my desktop. and laptop) Jun 13 13:35:15 wow, another wmii user ;) Jun 13 13:39:28 Enjoying your multiple screen support? :p Jun 13 13:41:26 I have 3 here at work (with wmii) and I must say it's quite comfy :) Jun 13 13:41:41 :O Jun 13 13:49:19 bencoh: xinerama? Jun 13 13:50:53 probably Jun 13 13:51:18 at least XINERAMA shows with xdpyinfo Jun 13 13:52:04 and I don't see how else wmii would work, so ... :) Jun 13 14:04:46 well, do you have different workspaces per monitor? Jun 13 14:22:06 nope, I only have several workspaces, independent of monitors Jun 13 14:32:19 bencoh: hm, I can't have wmii display different workspaces on monitors Jun 13 14:33:03 you mean, monitor-tied workspaces? Jun 13 14:33:30 (ie having a different workspace per monitor?) Jun 13 14:33:39 yes Jun 13 14:34:04 yeah, that's unsupported Jun 13 17:02:14 * sicelo also uses ActiveSync dialy .. in fact, it's even a pity that N900 can have only 1 EAS account. I need two Jun 13 18:38:10 wow, just tested wmii .. i like it very much! have used i3wm too Jun 13 19:03:32 anybody using KDE5? could you please "switch user" to open a new KDE session and see if the old session keeps internet connection while you're "not there"? you'd likely notice that after some 3 minutes, you lost IRC login here when the old session actually gets disconnected from internet Jun 13 20:25:32 Haven't seen such a problem on KDE4, but then, we haven't got around to KDE5 yet. Jun 13 20:58:36 solved: allow connection for all users Jun 13 21:00:49 Nice :-) Jun 13 23:25:42 * Oksana is exploring sawfish window manager on Linux laptop, currently; while idea of tiling is nice, I am not a fan of keyboard-based control. And highly customized window handling sounds... Very interesting. Jun 13 23:58:58 jonwil : Do I need to felete filebuffer before exiting function getJourneyData ? Jun 13 23:59:06 delete * filebuffer Jun 14 00:10:21 getJourneyDetailsData sounds... complicated. Jun 14 00:15:32 Just copy whatever the Translink code is doing Jun 14 01:00:02 jonwil : What should getStationsByGPS do : turn longitude and latitude into format readable by trip planner, or retrieve bus/train stops closest to GPS location? Jun 14 01:00:17 It gets the bus stop closest to the GPS location Jun 14 01:00:29 If the nsw site has that function then use it Jun 14 01:00:48 Otherwise set the flag to false so the system doesn't even try to use GPS with that backend Jun 14 01:01:59 * Oksana nods, and proceeds to use NSW site function. Unfortunately, with my chosen location, it doesn't return a bus stop, but either way it's going to work with GPS. Jun 14 01:17:46 When getting locations by name or GPS... When multiple results are returned... When user is asked to select one of the possible locations/stops... Where do these nice and readable names come from? Because StopID (required by trip finder) is one thing, and human-readable name of bus stop is another Jun 14 01:20:03 thing. Jun 14 02:10:50 Is it like "NameAkaDescription@LocationId"? Jun 14 02:20:12 Yeah that's what I did for Translink Jun 14 02:27:54 How does arrivalStation.section('@',0,0) work? I understand destinationStation.section('@',1,1), https://doc.qt.io/archives/qt-4.8/qstring.html#section Jun 14 02:30:03 jonwil: I know some about Python but I mostly use it only as a calculator in recent years. **** ENDING LOGGING AT Thu Jun 14 03:00:04 2018