The documentation has improved a lot since then, and we’ve just
recently hired someone to help here. But yes, we’ve been hearing this a
lot: “you guys do incredible work, but the docs suck”.
This is good to hear, but comes a little late.
The first version of Crossbar.io was release 2013-11-21 to PyPI. So this project exists for roughly 12 month (in the public).
So “late” is relative;) Anyway, I absolutely agree on this: the docs need to improve.
I am probably not good at writing easy, approachable docs. And I personally definitely can’t ramp up my workload more … I have reached the limits of what I can do.
Apart from authentication/authorization, what other features does
your custom router have?
Various custom logging, a modular plugin system (reusable), and many
months of infrastructure and integration that will need to be tossed.
=( I can feel your pain.
And I am feeling guilty for not having tried even harder to raise awareness to the following paradigm / strong opinion:
Routers should not contain application logic. Never.
Actually, app developers should not even think about writing or hacking on routers.
Are you hacking the PostgreSQL codebase to inject some application logic? (rethoric question)
It was a mistake to not splitting out all routing code from the very beginning into a separate project. Yes. In hindsight that’s clear.
But that’s how things evolve. In fact, if you go back in the code history of AutobahnPython, you will find WebSocket testsuite code! Which had long ago then be split out into a separate project (AutobahnTestsuite).
Now, logging sounds to me it should be in the router. So I’d also offer to work with you on getting that migrated.
“plugin system”: unsure what this means. If it is about app logic, then this is IMO the wrong approach to put into a router. The “plugin” system of Crossbar.io are standard WAMP components.
But, here is what you can do: just take the classes and code from AutobahnPython 0.9.3-3 and put that into your proprietory codebase. It’s Apache 2.0. Do what you want.
Of course that means you then have a fully custom router to maintain on your own. Sure. But nothing that existed was taken away from you.
If “uncalled” would mean “unneeded”, then: you don’t judge what’s
needed or not for this project.
No, but closer. I mean I believe that there is rarely a time when the
right call is to break backwards compatibility. Of course, it is quite
That would mean no project can evolve ever. Doesn’t make sense. No, now (<1.0) is the time to do so. It’s not either or. It’s shades of grey … need to find a balanced approach. And: I am still learning on how to manage a growing, open-source project. It’s tricky in ways … so I do make mistakes. Yes.
So what exactly have you contributed to this community? You didn’t
even care to announce that you use this stuff, that you wrote a custom
router, not to speak of trying to get your custom code integrated upstream.
I contributed a WAMP1 client (which you posted). I opened several
issues, and I spoke up from time to time on IRC. All this has happened
over the past year. My involvement is minimal, so I do not expect you to
remember me. I am sorry if I overstepped my bounds.
Oh, ok. Sorry for not recognizing! To be honest, I didn’t remember you (I am communicating about this stuff with a lot of people all the time over various channels) … but now I see. And will remember you;)
The reason I do not announce everything is that unfortunately I am not
allowed to discuss most of my work. I try to contribute where I can.
Never mind, I think I found your startup;)
I will take you up on your offer then. I ask that the project offer long
I agree. A statement on that is critical:
Let me ask a question: do you think the benefits of WAMP2 outweight
the break with WAMP1?
Well… Yes and no. Theoretically, yes. Practically, there had been no
complete, stable Advanced Profile WAMP2 implementation I can use. So, no.
WAMP BP already goes way beyond WAMP1.
WAMP AP as a spec is neither stable nor complete. So there can’t be such an implementation. Having said that, Crossbar.io is the most complete implementation I am aware of.
There is always a stretch between moving forward, inonvating, and
keeping the state of affairs.
Stability is a requirement for me. After the switch to WAMP2, I assumed
(I guess “hoped” is the better word), things would settle and begin to
Protocol-wise, WAMP2 BP is stable. There won’t be a WAMP3. In general, WAMP2 AP is work in progress. Selected features of WAMP2 AP are stable.
AutobahnPython is still <1.0. But see the issues I filed above …
Let’s make it concrete: I am willing to assist you porting your
custom authentication/authorization code to Crossbar.io.
Your kindness is truly appreciated. I think the biggest problem is
infrastructure changes and how my router was integrated. I will post
later with an example of my authentication stack and maybe we can see if
it has a place in crossbar.
Sure. Also note, that you can keep your stuff proprietory: e.g. if you implement your custom authentication as a WAMP component, you can hook that up to Crossbar.io without touching CB code, and only using AutobahnPython. The latter is Apache 2.0, so you can go closed source. There is no need to contribute anything into the CB codebase for such things.
Thank you for the quick reply.
You received this message because you are subscribed to the Google
Groups “Autobahn” group.
To unsubscribe from this group and stop receiving emails from it, send
an email to email@example.com
To post to this group, send email to autob...@googlegroups.com
To view this discussion on the web visit
For more options, visit https://groups.google.com/d/optout.
You received this message because you are subscribed to a topic in the Google Groups “Autobahn” group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/autobahnws/bCj7O2G2sxA/unsubscribe.
To unsubscribe from this group and all its topics, send an email to firstname.lastname@example.org.
To post to this group, send email to autob...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/autobahnws/54A8618F.2000208%40gmail.com.
For more options, visit https://groups.google.com/d/optout.