At 05:03 PM 3/15/2004, Noel J. Bergman wrote: >All in all, I think it is a fine idea. One thing caught my attention, and >that was the comment: > >> The initial code contribution would consist of: >> >> Support the Microsoft ASP.NET System.Web.Hosting interface from >> within the Apache httpd server (Windows platform port). >> >> Later code could include support for the ASP.NET interface from Mono, >> and ultimately support an Apache.Web.Hosting interface with complete >> bindings into the Apache httpd Web Server 2.0 - supporting an array >> of platforms. > >The phrase "Support the Microsoft ASP.NET System.Web.Hosting interface" >raises concern with me, due to Microsoft's .NET patents and copyrights.
Note that I say specifically *support*. Don't confuse this with *reimplement*, which is what the mono project accomplished. >I have no problem with the idea of supporting mod_cli, to coin a vendor >independent name, but please review: > > http://www.onlamp.com/pub/a/onlamp/2004/03/11/mono.html > >in depth, specifically the "Two Stacks" section. My concern is related to >the impact of Microsoft IP on whatver portion(s) of the proposed project it >touches. In this case we plug into the MS CLI implementation, and summon an MS hosted System.Web.Hosting container, thunking it at our mod_aspdotnet instead of at IIS. We implement a System.Web.Host named Apache.Web.Host. >If we are only implementing a plug for Microsoft's socket, and would have >other plugs for mono or other CLI engine, so that it is no different from >having an MPM that calls Win32 APIs and another MPM that calls Posix, that >seems fine. Can you clarify? Yes this is the server 'pipes' if you like to call them that, we have two Apache.Web classes, Host and Request, that instigate and answer an http request within MS's ASP.NET framework assemblies. >Other ASF projects looking at working with .NET / Mono / CLI. I expect that >people are going to have to look closely at what can and cannot be done, and >get some sort of common guidance documented. Yes - I'm somewhat concerned that plugging mod_aspdotnet into the mono CLI would be a potential IP violation. Simply having the mono implementation around might be an IP violation. I'd be facinated to hear what Novell has done so far in this space, in determining IP constraints on mono. >IANAL, but as a general rule of thumb, I think we are safe if we are writing >adapters from our code into theirs -- standard client code. I think we are >on shaky ground whenever we implement one of their service interfaces. The >mono folks appear to agree. Perhaps CLI-using ASF projects should talk to >the mono folks about the second, non-Microsoft, stack. I thoroughly intend to involve the mono and haydn folks. Give me a list to drag them into :) If we can't 'support' mono from mod_aspdotnet, we should be able to support mod_cli as a very mod_perl-esque intrinsic Apache interface. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
