A hundred different chat systems mentioned here. None of them compatible with one another. Well, I guess a couple of them have IRC gateways, but you have to actually set those up. Gee, wouldn't it be nice if you could pick and choose your UI? Pick and choose your "integrations", your "plugins", your client, etc without having to lose your entire userbase, history, contacts? Some sort of open protocol. Urgh, I've been…
> Some sort of open protocol. XMPP seems to be rather dead, for various reasons, and no real contender for a replacement has sprung up. I guess there always is, and always will be IRC…
Open Sourcers Race to Build Better Versions of Slack
171–180 of 397 posts
Re: Open Sourcers Race to Build Better Versions of Slack
#172What about reusing and polishing XMPP and existing decent software for it? About donating money to people who develop it all these years? No, we'll spread FUD upon XMPP or just ignore it but will proceed to use and abuse its legacy behind the closed doors. http://risovach.ru/upload/2016/03/mem/novyy--shablon_1085837...
For one thing, XMPP is complicated and difficult to implement.
HTTP and IRC on the other hand look easy at first, then you get past that and there's just more and more complications and edge cases and horror. Like all the various cache headers and behaviors in HTTP, or how all IRC networks have their own dialect that's subtly different.
I once tried to work on an IRC server interface, I read all the RFCs, wrote code, then tested with a client. Nothing worked, so I had to resort to replay what the client sent to a different network and reproduce whatever they did.
If you go read the XMPP RFCs, you should be able to write a client or a server that gets along pretty well with the rest of the XMPP world.
Re: Open Sourcers Race to Build Better Versions of Slack
#173A hundred different chat systems mentioned here. None of them compatible with one another. Well, I guess a couple of them have IRC gateways, but you have to actually set those up. Gee, wouldn't it be nice if you could pick and choose your UI? Pick and choose your "integrations", your "plugins", your client, etc without having to lose your entire userbase, history, contacts? Some sort of open protocol. Urgh, I've been…
Obligatory xkcd reference: ... ok, never mind, you probably all know which one I mean.
Re: Open Sourcers Race to Build Better Versions of Slack
#174Earlier quoted context omitted.
If you're worried about the upvotes on the article, they're definitely legit (as far as we can tell). I suspect it's as simple as that Slack is a hot topic (for or against) and people here like open source. And are passionate about chat software.
Thanks dang!
Re: Open Sourcers Race to Build Better Versions of Slack
#175A hundred different chat systems mentioned here. None of them compatible with one another. Well, I guess a couple of them have IRC gateways, but you have to actually set those up. Gee, wouldn't it be nice if you could pick and choose your UI? Pick and choose your "integrations", your "plugins", your client, etc without having to lose your entire userbase, history, contacts? Some sort of open protocol. Urgh, I've been…
Laziness.
Re: Open Sourcers Race to Build Better Versions of Slack
#176Earlier quoted context omitted.
> We need open protocols. You have that - SIP and XMPP. Please don't complain that they are complex and consist of a lot of documents, and have hard feature standartization path. It is the only possible way to openly evolve such systems - to publish serious writeups as documents on their own and have them discussed with criticism and then gradually adopted. And then you may end up with competing alternative implement…
The specs are not directly the problem in XMPP. They just amplify it. The real issue is XMPP does not directly solve the problems the authors of those chat apps deal with. Doing things properly requires writing XEPs and just a lot of hassle which a company building its own little Slack killer doesn't want to invest in, and doesn't have a reason to invest in either because none of the current ones support XMPP in the…
Re: Open Sourcers Race to Build Better Versions of Slack
#177Therefore, wouldn't it be nice if there was a website where product designers (not software developers) could post their product designs, so that the developers have something to work from?
Re: Open Sourcers Race to Build Better Versions of Slack
#178If only there was a well supported, standard, open and interoperable text messaging protocol (with multiple implementations) that companies could host themselves... Oh wait, IRC.
This comes up every time, and as usual you have totally missed the point . IRC is difficult to configure, difficult to host, difficult to use, and does not offer the same feature set as Slack. There is an obvious use case for something like Slack (obviously, because otherwise it would not have millions of users!) and covering one's ears while yelling "IRC! IRC! IRC!" is sort of wilfully ignoring that if it were a sui…
Re: Open Sourcers Race to Build Better Versions of Slack
#179Earlier quoted context omitted.
> Some sort of open protocol. XMPP seems to be rather dead, for various reasons, and no real contender for a replacement has sprung up. I guess there always is, and always will be IRC…
I use XMPP every day. All my friends are on XMPP. What makes you think XMPP is dead? What are "various reasons"?
Re: Open Sourcers Race to Build Better Versions of Slack
#180A hundred different chat systems mentioned here. None of them compatible with one another. Well, I guess a couple of them have IRC gateways, but you have to actually set those up. Gee, wouldn't it be nice if you could pick and choose your UI? Pick and choose your "integrations", your "plugins", your client, etc without having to lose your entire userbase, history, contacts? Some sort of open protocol. Urgh, I've been…
You're certainly right that the UX of most IRC clients is seriously lacking compared to eg. Slack. But one thing I've always wondered - is there any reason that IRC couldn't be used as the underlying protocol for these Slack-esque apps? There's no reason why the protocol itself has to handle UX features like emoji/embedding/etc. - just send a text token over the wire and have the client interpret it as a specially di…