Has already happened ;) [LWN.net]

Has already happened ;)

Posted Jun 24, 2005 7:48 UTC (Fri) by pkolloch (subscriber, #21709)
In reply to: Maybe it could also be a wakeup call for Mono.. by pkolloch
Parent article: Microsoft Puts Roadblock in Front of Open-Sourcing Avalon and Indigo (eWeek)

I just clicked on the MonoIndigo link and saw that they already responded in a way I very much like: They renamed the project to Amber and do not want to clone the API anymore. Instead they want to provide similar functionality in a library that runs in Mono and the original .Net environment.

It would be cool, if they succeeded in creating something sexy which is more popular than Indigo -- even on Windows. But there's of course a tiny bit of wishful thinking ;)

(Log in to post comments)

Has already happened ;)

Posted Jun 24, 2005 9:07 UTC (Fri) by NAR (subscriber, #1313) [Link]

do not want to clone the API anymore. Instead they want to provide similar functionality in a library that runs in Mono and the original .Net environment.

And thus breaking compatibility with the Microsoft libraries...

Bye,NAR

Breaking Compatability

Posted Jun 24, 2005 10:20 UTC (Fri) by pkolloch (subscriber, #21709) [Link]

Yes, obviously, and I outlined above that I do not care.

Has already happened ;)

Posted Jun 25, 2005 6:33 UTC (Sat) by lacostej (subscriber, #2760) [Link]

Except if they provide a compatibility layer for Microsoft libraries, something that allows to code for the compatibility layer and use Microsoft libraries + layer or mono lib underneath :)

With a transition tool to ease migration. Not sure how technically this is feasible, but that would be fun, turning the problem the other way around.

Has already happened ;)

Posted Jun 24, 2005 17:57 UTC (Fri) by mmarq (guest, #2332) [Link]

"... Instead they want to provide similar functionality in a library that runs in Mono and the original .Net environment. "

Couldn't they just wrap the existing Linux stacks, opening the door for other OSes in Open Source arena to establish a solid atandard?.

The idea is instead of providing the same functionality of Windoze to OSS/linux space, provide the "superior" functionality of OSS/Linux to *old* Windoze environments which could be achieved by wraping and shell replacing those environments like in colinux ( http://wiki.colinux.org/cgi-bin ) and or in bb4win ( http://www.bb4win.org/news.php )...

It seems to me that.NET environment was not really conceived to be cloned, the managebility features of C# were security oriented first, and programm easyng second, like a bait, a carrot at the tip of a poll hanging in front of your eyes. It was brilliant and has agglutinant power some how. But its modularity ensures that MS will always be a step ahead if they want too.

C# core is OK, but for one thing C# is not good for writing OSes and other close to hardware stuff, or even common server programms, even if there is a CLI CPU architecture, condemned to be "inferior", for sure, in the same sense that the JAVA CPU was.

I belive MS wanted all, not only the programms and programmers on their side(commercial and "freeware"), but also CPU architectures and IO expansion buses and protocols on their side also(hardware industry).

I belive they still want all, thought is evident that transforming X-box into a prevasive computational architecture will not do the bid, and CLI will not jump them on the CPU industry.

Pursuing MS path try to clone them will only prove to be a titanic adventure condemned to failure, but also promove fragmentation on OSS part, because the trend unfortunatly still is for everyone to go a separeted path.

Belive MS wil gain control of mono, or they will never allow it to copy enough to be a competitor. MS bid is that the Avalon/.NET will be a totally different environment than Win32/Windows. .NET modularity will provide easying up compatibility to kernel and to shell/GUI. I can see Novell enthusiasm , Avalon, Indigo on top of linux! on a pure WebServices architecture... but OSS dont really need that, because if it proves to be a not very good path, MS gets in trouble has it has been, but much smaller cloning contenders like Novell can implode!... it only needs that applications will run with a consistent look and feeling, thats all.

IMO the best path for cross pollination is to go with basic kernel cloning on top of a emulation engine like QEMU http://fabrice.bellard.free.fr/qemu/qemu-tech.html
providing WINE/Mono API/ABI on top of it, be it on *user mode* or *full system mode*; i.e., get MS completely out of the loop, embracing and extending as much as possible... Would love to see their faces upon tasting their own remedy!