Saturday, 2021-01-09

*** zbenjamin_ is now known as zbenjamin02:16
*** frinring_ is now known as frinring07:50
*** greguu is now known as greguu[away]09:58
DUCK12does signal work on sailfish10:39
DUCK12without android support10:39
ggabrielDUCK12: last time I checked, there's something unmaintained in openrepos. Personally, I use the android "app"10:43
DUCK12thanks10:43
Gestrathere's this but I haven't tried it myself https://forum.sailfishos.org/t/whisperfish-the-unofficial-sailfishos-signal-client/333710:45
DUCK12thanks10:46
ThaodanThere is a client but you could be banned any second11:24
DUCK12https://www.securemessagingapps.com/11:28
DUCK12since i read this11:28
DUCK12i never thought about this apps much11:29
DUCK12not it's very difficult choice11:29
rubdosDUCK12, ggabriel, it's far from unmaintained :)12:02
rubdosFeel free to come have a chat over at #whisperfish12:02
rubdosThe amount of WF users has more or less doubled, yesterday.12:03
rubdosKinda crazy.12:03
DUCK12yeah12:04
DUCK12because of musk12:04
DUCK12wrote use signal12:04
rubdosThat + the WhatsApp announcement of privacy statement changes :)12:04
Teguuhh, why does Signal open-source their stuff if they do not like third-party clients?12:48
TeguI may have misunderstood something, though12:48
attahNo, that is what they seem to be doing... pretty weird12:49
attahthey want to have auditablilty, but not promise any backwards compatibility or be robust for slight behaviour differences between clients12:49
Teguhmm, I guess that's kinda tolerable but someone mentioned bans?12:51
attahthey seem to be pretty up-tight about it, so wouldn't surprise me12:52
attahand being a service they provide they are well wthiun their rights to do that12:52
attahso personally i can't be bothered with their BS12:52
Teguye12:53
Nico[m]<Tegu "uhh, why does Signal open-source"> So that you can verify, that they are secure13:12
olSignal is not needed anymore. Matrix is much better.13:44
ol<DUCK12 "https://www.securemessagingapps."> Info about Riot is outdated. It's called Element now and secures message and attachments. And infrastractire jurisdiction is absolutely irrelevant: you can have your server wherever you wand.13:46
Nico[m]Well, I agree that Matrix is preferable, since I'm actually writing this from Matrix ;p13:47
attahI thought Matrix was just another silly attempt to bridge all protocols into one app... so it is more than that?13:47
Nico[m]It is like the email for instant messaging13:48
Nico[m]And has bridges to other protocols, like you can create PRs using email on Github, you can send messages to IRC using Matrix. Or to Steam, Discord, Signal, Whatsapp, etc13:49
ol<attah "I thought Matrix was just anothe"> Actually, it's quite successful attempt. I'm writing this message using Matrix. Matrix to IRC bridge is the best IRC bouncer.13:54
attahhmmm13:56
Nico[m]Haven't signed into IRC for a few years now .-.13:56
Nico[m]Or Telegram for that matter13:57
attahi guess i need to crawl out from under my rock at one point or another13:57
attahI tried Telegram/Fernschreiber the other week, but afiled already at the basic instruction "hit me up on TG, @SomeName"14:01
attahwhat is the Matrix app on SFOS anyway? can't seem to find any14:01
Nico[m]Well, I'm working on Konheko, but the app is very barebones still14:02
olI use Element-Android in Alien Dalvik.14:02
Nico[m](And most of my time is spent on Nheko, which is a desktop client)14:03
olAnd Element-Web in Chromium on my laptop and workstation.14:04
* attah crawls back under the rock from whence he came14:06
Nico[m]How rude, I like my clients! D:14:07
olNico[m] (IRC): I still need edits and replies for my Matrix experience.14:08
Nico[m]Well, Nheko has replies, just not edits, because they are not in the spec yet and the current proposal sucks .-.14:10
attahi didn't mean to diss your clients, i see now it can read like that14:11
attahsorry!14:11
olThat's the problem. No matter how current proposal sucks, it's implemented in mainstream clients, so everybody expects it to be in other clients as well.14:11
attahI mainly meant to say, a) ugh, fragmentation (about the general situation), and b) i'm taking a walk before it gets dark14:12
Nico[m]attah: no worries, I was just joking around ;p14:12
Nico[m]ol: Yes, but there are a lot of problems with that. You can't just have the biggest client with actual paid developers implement crap and then force it on everyone else14:14
Nico[m]attah: It is not really fragmentation, if you make a native client instead of forcing people to use the electron client imo :314:16
olYou can. It's how things are done in real world. There are multiple examples of proprietary extensions becoming so pupular that they turn into de facto standards.14:16
Nico[m]<ol "You can. It's how things are don"> Sure, but I don't like it14:16
Nico[m]I'm not going to implement a hack, that breaks in every edge case, when there is an obvious solution14:17
Nico[m]Like edits don't support editing images, what you replied to, etc. Imo that should be possible. But the proposal says to just delete the event in that case and send a new one. In which case, why have edits at all?14:18
olI understand. But users don't care about how nice it looks under the hood, they just want edits in a way compatible with other clients.14:18
Nico[m]<ol "I understand. But users don't ca"> Agreed. But I just focused on other useful features instead, that I find more useful than edits :314:19
olIt would maje sense to implement current ad hoc solution instead of waiting for years until nicer protocol extension is implemented in mainstream clients.14:19
olI don't see what can be more important that edits.14:20
Nico[m]<ol "I don't see what can be more imp"> Call support14:22
Nico[m]You can just use redactions instead of edits14:23
olOK, this is important as well. Depends whether you're talker or texter.14:23
Nico[m]You can't really work around missing e2ee, calls, etc14:23
Nico[m]Edits are really just a small UX improvement14:24
olNo, I can't use redactions because after my message was replied to (no matter whether using reply feature or by sobsequent message) because it breaks conversation flow.14:24
olEdits are very useful to correct typos, for example.14:25
Nico[m]Fair enough14:25
Nico[m]<ol "Edits are very useful to correct"> Well, yeah, but that is mostly not necessary14:25
olFor me it's necessary very often.14:25
Nico[m]I can understand you with typos too :D14:26
Nico[m]Anyway, I'll implement them at some point14:27
Nico[m]I just don't like the way the Element team handled that whole thing, because they just enabled an experimental feature for everyone and now they say, it is too late to fix it14:27
olAll Internet develops this way. Reminds me about IPv4 with 32-bit addresses that are too late to fix. :-)14:28
Nico[m]Well, the matrix spec has unstable prefixes for a reason as well as labs flags14:29
fridlmueI need to download and extract some zip in my app using python and pyotherfile (it should be downloaded only temporarily). Which path do I use correctly in the python code to download to and to extract to?17:10
Nico[m]QStandardPaths::TempLocation ?17:11
Nico[m]Alternatively QStandardPaths::CacheLocation17:11
fridlmueI would downlaod and extract within the python code. So i need the path there. Would you hand over the path?17:13
fridlmueAnd do I manage the removal in TempLocations and CacheLocations myselfe17:13
fridlmue?17:13
fridlmueOr is that done by the system sometimes.17:14
Nico[m]Well, the path differs on the user and such, is there no qt function for that in python?17:14
Nico[m]And you can just delete the path of what you downloaded after extracting17:14
Nico[m]But the operating system may clear it at times as well, like reboot or low disk space17:15
fridlmueok.17:18
fridlmuetbh i don't know if there is such a function on the python side. Thats why I ask.17:18
Nico[m]Hm, seems to be an extra package: https://pypi.org/project/PyStandardPaths/17:21
fridlmueAnd i can include the lib then in the RPM somehow, can't i?17:25
fridlmue(thanks for the finding, btw!)17:26
Nico[m]Yeah17:26
Nico[m]You can probably also hardcode /tmp or ~/.cache/appname on Sailfish, but it is ugly17:26
fridlmuei'll try with the nice way ;-)17:32
Nico[m]:D17:32
*** vgtw_ is now known as vgtw20:15
thigg[m] has someone a way for python to resize images without introducing another dependency?20:24
attahthigg[m]: there are Qt bindings right? So QImage?20:32
thigg[m]hm, thats an anyoing roundtrip from python to qt, but should be doable... hmm20:33
attahannoying workarounds are my speciality https://github.com/attah/harbour-seaprint/blob/731db2dd1e2e7ece5a3b4299ec963420d740aeb4/src/convertworker.cpp#L23120:34
Nico[m]attah: agreed20:35
rubdosTegu, attah, they seem to tolerate third party clients these days.  I've been quite public with Whisperfish and mentioned it, e.g. on their Github and when they posted on Reddit; they seem to ignore third-party clients instead of banning and shouting now.21:06
Tegurubdos: yea, I briefly read the whisperfish thread and noticed that21:06
attahthat's nice21:06
rubdosAlso, they seem to be *very* much into backwards compatibility. There was very little to update in the quite old Whisperfish code to get things to work again.21:06
rubdosWe are behind on schedule (groupsv2, reactions, blablabla), and still all the old stuff works.21:07
rubdosWe're progressing faster than they invent new stuff though, but we're still a bit behind :-)21:08
attahWell, that's just what i could gather from the info that was out there a year ago... and FWIW, i have at least once had my Android Signal tell me it was outdated enough to refuse to work21:08
attahgood work!21:08
rubdosYeh, but that's because they don't want to bother with out-dated clients; their story is completely different on protocol-level!21:08
rubdos(which is really awesome imo)21:08
attahi just assumed that was for protocol reasons, but i guess there can be other reasons21:09
rubdosI mean, you *do* lose data if a newer client sends something new-and-fancy to an old client21:10
rubdosbut it won't crash or do anything unexpected21:10
rubdosthey mainly want to avoid that21:11
* rubdos has added `whisperfish` to his highlight list.21:11

Generated by irclog2html.py 2.17.1 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!