A Shallow Introduction to the K Programming Language. About two years ago I was introduced to a programming language that I really didn't like: it didn't have continuations, I didn't see any objects, it had too many operators, it didn't have a large community around it, it was strange and different, and it looked like line noise, like Perl, and I don't like Perl. However, I gave it a try. I had to learn that continuations may not be there, but first-class functions are; it may not have a normal object system, but that is because the language doesn't need it and gets it power by cutting across objects; all the operators are the functions that make up its standard library; it's community may not be large, but it is incredibly intelligent; it only looks strange until you understand its concepts; and well, it will always look like line noise, but you will stop caring because this also make the concise code easier to read. K has since become my language of choice. [kuro5hin.org]

one scary looking language with more power than imaginable.

Someday I really hope I can use k and kdb in physics.

10:45:48 AM  #  

trump cards up the big boy's sleeves:

MS and IBM both have huge trump cards that they could play that would cause big ripples. here they are:

MS: aggressively release the CLS, .NET framework, and .NET languages for *all* major operating systems. the UI parts probably wouldn't be ported, since their target would be server-side systems, but all the rest would be ported (including the SOAP stack, web server, transactions, etc.). they could charge for this software, just like they charge for windows, and make sure that the windows version is optimized to run better than the other platforms. this effectively neutralizes one of java's remaining differentiators, which is its portability. microsoft's mantra would become "develop on windows, deploy everywhere".

IBM: release an open-source CLR, open-source compilers for both Java and C# that target the CLR with compilers for eclipse, and an open-source framework that is different from microsoft's. this neutralizes microsoft's ability to corner the marketplace for CLR-based infrastructure, and is consistent with IBM's direction of open-sourcing stuff to engage the community to combat microsoft. it also frees IBM to innovate in the language arena and accelerate the development of alternative frameworks that are better than those in the J2EE stack.

my guess is that both trump cards will be played sometime during the next six months. what do you think?

[what's next?]

add SWT and Sash on IBM's side. IBM will have to produce desktop no matter what.

10:41:18 AM  #  

Embedded Databases: The unglamorous database option that works.. Dr. Dobb's Journal: Embedded Databases: The unglamorous database option that works. [Hack the Planet]

An interesting take on building home-brew datastore at Cisco.

10:17:34 AM  #