[Mono-list] Mono and MS Singularity

Alessandro Bottoni alessandro-bottoni at libero.it
Thu Oct 26 02:31:47 EDT 2006


Andreas Färber ha scritto:
> Concerning the thought of a managed OS, wouldn't that either mean you'd
> have to write it in assembler and thus architecture-specific or you'd
> need CLI-specific hardware? I thought I heard for Java there was some
> research to create a processor that could directly execute Java
> bytecode, don't have any links though.

Myabe it was JOP: http://www.jopdesign.com/ (FPGA based).

> Anyway, there are a number of both experimental and commercial operating
> systems out there, each with their unique pros and cons - 

I'm having a look at those, for example:
http://www.jnode.org/ (http://en.wikipedia.org/wiki/JNode)
http://www.vitanuova.com/inferno/index.html
(http://en.wikipedia.org/wiki/Inferno_%28operating_system%29)

They are quite interesting. In particular, Inferno is quite mature and its
language, Limbo, is quite similar to C#.

I know of a few others, as well:
http://www.eros-os.org/ (uperseded by coyotos and capros)
(http://en.wikipedia.org/wiki/Extremely_Reliable_Operating_System)
http://www.coyotos.org/ (used by HURD)
http://www.capros.org/ (successor of EROS)
http://www.bluebottle.ethz.ch/ (based on Oberon)

They are all, more or less, experimental and they are almost completely unable
to run on a machine that is different form the one used for their development
because of the lack of drivers.

> even if Novell
> as a widely known company said its OS would in a future version be fully
> managed, then that would still mean that they would have to "re-invent"
> everything from the ground up (resource management, threading etc.) just
> like any other OS (e.g. Linux) and even if they did and supplied drivers
> for virtually every device as well as a set of apps then that would
> still not assure a commercial success in the sense of a large user base.
> (think BeOS) Same argument as for Singularity. However, this is in no
> way intended to stop you from proving it can be done. :-)

Yes, it would be a huge effort and it would not have any guarantee to be
appreciated by the market. BeOS demonstrated how a good OS can be ignored by
the market. Anyway, if MS will go ahead on this road, the FLOSS world will have
to follow it. Inferno, Mono and CoyotOS could be good starting points.

Thanks for your time.
------------------------
Alessandro Bottoni


More information about the Mono-list mailing list