Signal, thank you for not collecting my data. But I won’t use you
ruky.me
ruky.me
This article reminds me of that — "we need x" and then a big void instead of discussion about why we don't have x or what we might do to get x.
We don't know why only one of the big providers started interoperating with XMPP. But we do know why that one stopped again: A terrible wave of spammy contact requests. They were able to fight intra-gtalk spam much more effectively than spam that originated via XMPP.
Someone who wants app authors (and IM providers) to support an XMPP-like protocol could investigate and write about how an XMPP-like protocol could be made unfriendly to spam. That would be interesting and IMO useful, completely unlike this article.
I think you're mistaken to believe this over the fact that Google was going through one of its turbulent internal transitions. The whole of Google Talk got replaced by Hangouts (which is now facing its own complex demise: https://arstechnica.com/gadgets/2019/01/the-great-google-han... ) and the whole Google+ shift undoubtedly played a part.
Spam is not the problem. I have a Gmail account, and I receive tonnes of spam daily. I've had spam in Hangouts, I've had spam in Google Docs. Meanwhile independent and voluntary public XMPP server operators have quite a good handle on spam on the network today. I refuse to believe that Google's resources were unable to solve this, rather than that they were unwilling to solve this and took it as an excuse to wind down their participation in an open network.
I'm all for more research into anti-spam solutions, any open network has to deal with it. But so far none of the well established open networks (phone, email, physical mail) have yet solved the problem.
Consider Signal, since that's in the subject. Despite having my phone number on several web sites, I don't get spam on Signal, which I take to mean that the Signal maintainers are successful at fighting spam (and it doesn't matter whether they do so mostly as server operators or protocol designers). If you wanted to persuade them to interoperate/federate, then you have to provide arguments, ie. the burden of proof is on you.
Saying "none of the well established networks have yet solved the problem" isn't enough, because you'll be trying to persuade someone who has solved the problem. You could try to argue that spam is a worthwhile price to pay, or that it could be fought as well using your protocol as via Signal's own servers, or other things.
What you can not do is lean back. An interop mechanism is a server-to-server bulk message delivery mechanism, and so far I've not heard about any way to do bulk message delivery that doesn't atttract the interest of spammers. That spammers would try to spam Signal users via the interop mechanism is plausible by default, just as it's plausible that spammers would try to spam Gmail/gtalk users.
>If the subject software had a concept of priority, these articles would explain how to set priority using 'configure' but not whether priority=2 was higher or lower than priority=1.
It's amazing how often this is done. Two favourite examples of mine are:1: Those text-base Captchas which NEVER tell you whether the letters are case sensitive or not.
2: I get a monthly report from Yandex Webmaster which gives me some basic info on visits to my sites. It also gives my sites a 'score' which varies up or down a bit. Sometimes it's 10. Sometimes it's 8.
Not one single time in the several years I've been receiving these reports has Yandex ever thought to enlighten me as what this score means; whether it's good when the score goes up... or when it goes down... or what the score is out of. Is my site occasionally scoring 10/10?... or is it scoring 10/100 or /1000?
I wonder what they mean by that. There was a recent post on HN where many people said that it's one of the most polished private messengers (and I myself love it too).
As for the connections to Russia, this isn't an issue, since they store all their data encrypted in datacenters around the world, making it near-impossible for the governments to get decrypted data by court orders in all needed countries.
They also again mentioned the misconception that "telegram is not secure by default". See https://telegra.ph/Why-Isnt-Telegram-End-to-End-Encrypted-by...
Plus there is no option to select multiple chats, which is quite annoying
I don't see what's so difficult about that.
[and, like a previous commenter, I never delete chats either but I sussed that out in about 2 seconds by following the standard Android convention of tapping and holding on items in a list, to make multiple selections.]
I also find it odd you condemn Telegram for its 'poor interface' when it's generally recognised as being one of the most polished of the various chat apps you list.
I've recently come to Telegram after years of using Wire and becoming frustrated with its stagnated development and user-hostile dev team. Before settling on Telegram, I also kicked the tyres on Signal and several other secure messengers and Telegram was lightyears ahead of them all, in terms of features and usability.