Never been a fan of Merb 2 == Rails 3 either. FWIW, not a fan of the rails merger at all.
2009/11/7 Pavel Kunc <[email protected]> > > Hi all, good to see you on mailing list again. > > This really amused me: > "I also think that it might not be worth delaying 1.1 to add unicorn > support." > > Good joke. The 1.1 can't be significantly delayed by adding unicorn > adapter. Because it already has delay about a year or so. Really, the > 1.1 promised many things and almost none of it was delivered. One of > the crucial part of 1.1 is 1.9.1 support. Tell me how you want to > provide out of box support on 1.9.1 if you defaults to Mongrel which > doesn't run on the 1.9.1? > > The idea of Merb adapters is that user can use server he wants by just > using the switch. Latest patch in 1.1 even allows you to set your > adapter in the init so you can use thin and don't need to type merb -a > thin all the time. > > Current 1.1.0.pre version is ready to ship to the public for testing > and once Martin solve last issue with merb-stack app, we're going to > push it to Gemcutter. Developers and users are frustrated with lack of > production release, bugfixes, and communication. > > On the removing whole master/worker stuff and maybe also the class > reloading which can be now solved outside Merb I totally agree with > Nicholas. My opinion is that it really shouldn't go to the 1.2. It > should not go to any of 1.x versions because it breaks the API. And > one thing we tried hard to keep was API during work on 1.1.0.pre. We > wanted to remove many things from Merb but because of spec10 must pass > for 1.x we can't. > > So for me it really is what you want to do with Merb? Because what > would be the best IMHO, is to release 1.1.0. as we have it now with > Unicorn. To provide good working release on 1.9.1 and compatible with > spec10. And start working on Merb 2.0 (and yes I never liked statement > Merb 2 == Rails 3). In Merb 2.0 we can easily break API, we can drop > spec10 and use current spec as frozen API state. Than we can remove > bootstraping, change router, bring merb much closer to Rack and all > integrate all the cool stuff which just emerge lately. > > If all this sounds a bit angry, it is. And also it's the point of view > of someone who don't need/want to migrate to R3. > > Pavel > > On Nov 6, 10:23 pm, Matt Aimonetti <[email protected]> wrote: > > +1 Passenger 3 is also coming up with lots of goodies. > > > > - Matt > > > > > > > > On Fri, Nov 6, 2009 at 1:55 PM, Nicholas Orr <[email protected]> > wrote: > > > > > Yeah I'd aim for 1.2 and getting rid of the bootstrap stuff. > > > > > Merb doesn't need to handle forking master/worker stuff anymore... > > > There are tools available that can do this. > > > > > Imho it make sense to make merb work with rack up like sinatra. That > > > way if someone else comes up with a new way to run rack apps merb > > > works out of the box, rather than having to implement special support > > > for in merb for the new tool. > > > > > Nick > > > > > On Sat, Nov 7, 2009 at 5:48 AM, Matt Aimonetti < > [email protected]> > > > wrote: > > > > Whatever, changing such a critical aspect of Merb right before a > major > > > > release doesn't seem wise at all. > > > > My suggestion: release 1.1, and play with unicorn in a separate > > > > branch/trunk. Once it's ready and you know it's reliable and better > than > > > the > > > > existing solution, then push a new release. > > > > > > - Matt > > > > > > On Fri, Nov 6, 2009 at 10:40 AM, Ezra Zygmuntowicz < > [email protected]> > > > > wrote: > > > > > >> Unicorn is by far less buggy then the current > > > forking/master/worker > > > >> stuff in merb. I say go for unicorn support. > > > > > >> -Ezra > > > > > >> On Nov 6, 2009, at 10:32 AM, Yehuda Katz wrote: > > > > > >> > I agree > > > > > >> > On Fri, Nov 6, 2009 at 10:22 AM, Matt Aimonetti < > > > [email protected] > > > >> > > wrote: > > > >> > I think that making unicorn the default adapter is not a very good > > > >> > idea. > > > >> > It still has lots of bugs and should be handled with caution. > > > > > >> > I also think that it might not be worth delaying 1.1 to add > unicorn > > > >> > support. > > > > > >> > - Matt > > > > > >> > On Fri, Nov 6, 2009 at 3:56 AM, Pavel Kunc <[email protected]> > > > >> > wrote: > > > > > >> > Unicorn support could come earlier than 1.2. I'd like to bundle it > > > >> > with the 1.1 if possible because it would allow out of box 1.9.1 > > > >> > experience. I'd also made unicorn as default adapter. So I think > we > > > >> > can try for now to write unicorn adapter and just plug unicorn to > merb > > > >> > the same way as other servers. > > > > > >> > We can than decide what to do with the bootstraping. > > > > > >> > Pavel > > > > > >> > On Nov 6, 3:17 am, Yehuda Katz <[email protected]> wrote: > > > >> > > I would be in favor of removing the Merb bootstrap stuff in > favor of > > > >> > > unicorn, which is basically the same code but made more generic. > > > > > >> > > We could do that for 1.2 for sure. > > > > > >> > > -- Yehuda > > > > > >> > > On Thu, Nov 5, 2009 at 6:29 PM, Nicholas Orr > > > >> > <[email protected]> wrote: > > > > > >> > > > I had a go at running merb via unicorn and had the same > > > >> > experience. > > > > > >> > > > My guess was that because merb does it's own bootstrap stuff > then > > > >> > > > launches (forks?) thin/mongrel - this is where running with > > > >> > unicorn > > > >> > > > falls over. > > > > > >> > > > As you can see in the unicorn_rails start script it is very > > > >> > different > > > >> > > > from the unicorn script and basically satisfies rails way of > doing > > > >> > > > things. > > > > > >> > > > One thing I didn't try was starting with a working passenger > > > >> > config.ru > > > >> > > > - that may have resulted in a working solution... > > > > > >> > > > A goal of 1.2 might be have merb behave more rack like so one > > > >> > could > > > >> > > > use rackup - then merb could slot into everything > (middlewares, > > > >> > rack > > > >> > > > stack) else rather than the current situation that everything > > > >> > can slot > > > >> > > > into merb... > > > > > >> > > > Nick > > > > > >> > > > On Fri, Nov 6, 2009 at 12:51 PM, scottmotte > > > >> > <[email protected]> wrote: > > > > > >> > > > > Hi guys, > > > > > >> > > > > Any thoughts on why using unicorn to run merb would quickly > > > >> > load up > > > >> > > > > the merb app, close it down, then respawn and try again - > and > > > >> > then > > > >> > > > > repeat that over and over again. > > > > > >> > > > > The config.ru file is the same I'd use for thin, and I'm on > > > >> > the latest > > > >> > > > > Merb 1.1 from wycats github. > > > > > >> > > > > #my config.ru > > > >> > > > > require 'merb-core' > > > >> > > > > Merb::Config.setup(:merb_root => ".", :environment => > > > >> > ENV['RACK_ENV']) > > > >> > > > > Merb.environment = Merb::Config[:environment] > > > >> > > > > Merb.root = Merb::Config[:merb_root] > > > >> > > > > Merb::BootLoader.run > > > > > >> > > > > use Merb::Rack::Static, Merb.dir_for(:public) > > > >> > > > > run Merb::Rack::Application.new > > > > > >> > > > > # if you want to try unicorn yourself. > > > >> > > > > The command to setup and run unicorn is: > > > >> > > > > sudo gem install unicorn > > > >> > > > > cd yourapp > > > >> > > > > unicorn # alternatively to try on a rails app it's > > > >> > 'unicorn_rails' > > > > > >> > > > > I'm asking this in the unicorn forum too. > > > > > >> > > -- > > > >> > > Yehuda Katz > > > >> > > Developer | Engine Yard > > > >> > > (ph) 718.877.1325 > > > > > >> > -- > > > >> > Yehuda Katz > > > >> > Developer | Engine Yard > > > >> > (ph) 718.877.1325 > > > > > >> Ezra Zygmuntowicz > > > >> [email protected] > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "merb" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/merb?hl=en -~----------~----~----~----~------~----~------~--~---
