.NET for freebsd

Terry terry at mohimba.com
Thu Jan 30 02:55:13 GMT 2003

Quoting Robin Melville <robmel at innotts.co.uk>:

> The think that I find most worrying about future development along .NET 
> lines is the strongly stated aim of applications as 'services'. There 
> is already a dangerously monopolistic situation in terms of operating 
> systems. If applications no longer reside locally and one is obliged to 
> rent access to them (which is how I understand the thrust of 
> Microsoft's implementation of .NET) then the monopoly gains control 
> over your minute by minute use of applications. I may be a little 
> paranoid but that's not a situation I relish.

To address one end of the spectrum first consider that Microsoft is steering the
market segment it controls in any way that proves (or has the potential to
prove) most profitable for them. This is not a bad or non-ethical practice, it
is dictated by the hard economic reality. Companies exist to be profitable and
need to be ever-expanding. From this perspective, it is clear to any observer
that renaming applications to `services' and getting end-users to pay for them
on a pay-per-use or any similar scheme is a great source of revenue and it
eliminates a lot of piracy headaches for them. Micorsoft acquired Hotmail,
launched the MSN portal, strengthened its instant messaging software and
introduced PASSPORT. Even if this was not a well-crafted plan to create a wide
platform of various interoperable services which `joe average user' wants
desperately, it soon turned to look like one even to Micorsoft executives. Given
that in some loose sense this platform is already in place, yet people are
accustomed in getting all these everyday online services for free, how would you
get them to pay? `Application renting' (using the term loosely) is nothing new,
it has been attempted before (namely from Sun and network/thin client computing)
without much success. But this time two new factors come into play: 
1)The XML family of technologies is here and it sets widely agreed-upon
standards for I/O communication between applications. This not only means that
Windows-alien applications can `talk' to Windows native ones using these
standards (something that I suspect wouldn't concern Microsoft much since it is
dominant in the desktop market) but most important more vendors will (most of
them hopelesly in my opinion) try to `milk the cow'. This can lead to a wide
acceptance of Microsoft's platform (.NET), attract a wealth of highly
sophisticated coders from the UNIX world that never before bothered with Windows
platforms and open up the gates to enterprise markets where Microsoft is not so
competitive at the moment (server market). Needless to say, the companies that
hit an interesting market niche will follow Netscape's fate.  
2)Microsoft has already a foothold in the `market' it is trying to turn from a
non-paying to a regularly paying one so its chances are substantialy higher than
any other company's which has attempted something similar before

Given the above, picture a situation where the companies and users that are
already on Windows platforms will turn to .NET for development of services and
consumption of services. This will actually be a valuable upgrade for them
because if they don't move on either to UNIX platforms or to .NET they will be
left out on the great benefits of XML technologies. This is important, please
note that the benefits derive from XML technologies and _not_ .NET which is a
Microsoft (application framework) implementation based heavily on them (with the
addition of byte-compiled laguages and so on). Sun for example offers similar
implementations using their J2EE platform (in my opinion more advanced at the
moment than anything Microsoft talks in .NET. Anyone who wants a detailed
technical analysis is welcome to email me for details) as there are a number of
Apache related projects, Python projects and the list goes on. Now, having the
Windows based crowd into the picture lets take a look at the UNIX crowd. 
Suppose that a wealth of developers jumps on the .NET platform using .mono for
example. What is the benefit for them? Naturally they will have to learn all the
Microsoft-devised, Microsoft-invented, classes, methodologies, quirks. This is
not a benefit in itself. Then they will develop applications that can talk to
Windows ones. This seems like a benefit. Well for the most part it isn't since
.NET applications want WSDL descriptions and SOAP objects  to communicate. This
can be done using _any_ language. Sure, some languages have ready-made libraries
that can aid quick development but this doesn't mean that only C# in .NET or
.mono or DotGNU for that matter has them. There's a wealth of them in PHP and
Python and Java and Tcl to name some of them. So as far as web services are
concerned the benefit for the UNIX crowd has been none other than having to
learn the Microsoft application framework and embrace the great technical
tradition Microsoft has (not to mention innovation). With web services stripped
out of the picture we have the virtual machine benefits left. You write your
code once, it runs on any machine that has a virtual machine running on it (I am
greatly simplifying things here but this essentially is the benefit). Hurraayy!
Microsoft invented portable code, lets all rush and learn C# because  no other
language ever offered this benefit before and while we are at it, make sure we
learn all the Microsoft-devised class names and please make sure you name your
executables .exe because all great software enginners and scientists do.
The substantial benefits above warrant that we should all forget all about UNIX,
science, and downgrade to Microsoft's interpretation of new technologies
(invented elsewhere as always) because they indeed spent bucketloads of money on

thank you


More information about the Ukfreebsd mailing list