OWIN is dead

[Sebastien Lambla]

Statement:

Help has not been coming forward for neither BuldFunc, MidFunc, or Microsoft doing anything beyond glaring happily at their own onanistic creation in IApplicationBuilder.

This is the last straw for me. Is there any sign whatsoever that Microsoft’s attitude of embrace and extinguish didn’t kill us? I see nothing. Nothing at all. A few OSS projects like mine with very low adoption, and no one caring.

This is a last ditch attempt at a last ditch participation in Microsoft communities to see if there is any cojones in moving this forward.

Is anyone caring? At all? We can’t do this on our own, and Microsoft only cares as long as you nod and bow to IAppBuilder, or go away and create an alternative doing EXACTLY THE SAME ORIGINAL SIN with a different name because some decided to take their toys somewhere else.

After months of silence, I’m sick and tired.

Are we going anywhere? Is anyone caring? Are we moving in any direction?

I’m happy to drop the whole thing, but I need people, in an out of the b0rg, to make a choice. We fight or we flee, or somehow collaboration gets back on the table.

Anyone? At all? Anywhere?

Seb

[Ryan Riley]

On Wednesday, January 14, 2015 at 2:22:52 AM UTC-6, SerialSeb wrote:

Statement:

Help has not been coming forward for neither BuldFunc, MidFunc,

Sorry about that. I've been busy working on better OWIN support for F#, which includes an implementation of web machine on top. We've also been working to get OWIN support in another F# web stack called Suave. I'm still game for helping with MidFunc, though I need a lot of ramp up to know how to correctly write specs. I'm less interested in BuildFunc, as I've noted before, but happy to help move it forward. I just can't drive some of those things. Can you submit a PR of what you have?
 

or Microsoft doing anything beyond glaring happily at their own onanistic creation in IApplicationBuilder.

This doesn't bother me much. This is like Freya's opinions of mapping AppFunc to Env -> <'T, Env> used in Freya. As long as you can still hook up your OWIN stuff, I don't see this as a deal killer.
 

This is the last straw for me. Is there any sign whatsoever that Microsoft’s attitude of embrace and extinguish didn’t kill us? I see nothing. Nothing at all. A few OSS projects like mine with very low adoption, and no one caring.

Very few people have ever cared about OWIN. I've been trying to get more people interested by talking about it a few times in the last year, first at Austin Code Camp and more recently at CodeMash.
 

This is a last ditch attempt at a last ditch participation in Microsoft communities to see if there is any cojones in moving this forward.

Is anyone caring? At all? We can’t do this on our own, and Microsoft only cares as long as you nod and bow to IAppBuilder, or go away and create an alternative doing EXACTLY THE SAME ORIGINAL SIN with a different name because some decided to take their toys somewhere else.

I don't think this is the case. They created IAppBuilder and carried it forward, and it serves as a hook going forward for their frameworks. That's fine. We'll all have our own such mechanisms. The big deal is in making our stacks big enough to be noticed by those building re-usable middlewares. And most people just won't care. I'm okay with that. I've had enough good success promoting OWIN within the F# community that I'm satisfied. I'd love to see it get bigger, but that will be a conversation with people like Thinktecture.

After months of silence, I’m sick and tired.

Are we going anywhere? Is anyone caring? Are we moving in any direction?

I’m happy to drop the whole thing, but I need people, in an out of the b0rg, to make a choice. We fight or we flee, or somehow collaboration gets back on the table.

Anyone? At all? Anywhere?

I'm sticking with it. I'm finally seeing more pick-up. I think it would be good to publish the MidFunc spec and keep working on the most common cases. I know Damian is also doing a lot of work with OWIN and creating some interesting things, but he seems to have limited time, as well. A good, solid, core spec is my goal, as well as promoting compliance as much as is possible or, failing that, providing guidance as to how to map things that are close to OWIN into OWIN.

Cheers!
Ryan

[darrel...@gmail.com]

I continue to hold the same position as before, that I will ensure any middleware I create will conform to the OWIN specs.  I still believe OWIN plays a significant role the .Net web ecosystem.  Microsoft’s new stack supports OWIN components even if it doesn’t use the AppFunc directly internally.  I can put up with that.

I still make every effort to avoid taking a dependency on the IAppBuilder/IBuilder stuff, simply because I don’t buy into the approach.   But that’s just me.  And it shouldn’t matter what any of us do in the privacy of our own app frameworks and HTTP Servers, as long as we continue to interface via OWIN.

It’s not a perfect world that we have ended up in, but it could be a whole lot worse.

I was under the impression that the MidFunc signature was decided upon and I missed the BuildFunc discussion.  I don’t really see the need for a BuildFunc, but I’ll catch up on the archives and throw in my 2c if I have an opinion.

Darrel

Sent from Surface

--
You received this message because you are subscribed to the Google Groups "OWIN working group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to net-http-abstrac...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Ryan Riley]

On Wednesday, January 14, 2015 at 2:30:57 PM UTC-6, Darrel Miller wrote:

I was under the impression that the MidFunc signature was decided upon and I missed the BuildFunc discussion.  I don’t really see the need for a BuildFunc, but I’ll catch up on the archives and throw in my 2c if I have an opinion.

MidFunc is decided but not specified. Seb has a draft for the MidFunc spec and was also working on a BuildFunc (app initializiation / builder pattern) spec. I'm fine leaving this all at the MidFunc, as well, but I think it would be good to get the MidFunc spec published.