After reading the post that Brad Wilson wrote i began thinking of his ideas of having data performance related to our usage of a language.
the “If you implement an O(n^2) algorithm in C++ and I implement an O(n) algorithm in Ruby, I’m going to kick your ass for any significant number “n”. ” beyond been extremly funny is comparing apples to oranges, yes they are both round.
so i dont see the comparison there. i can see the differance in running a C# algorythm and a C++ algorythm and claming that the C# is running faster because there is an optimized implementation of some memory management in the .Net framework. or because there is a better helper class allready implemented and optimized for a mathematical equasion. it could allso be the CLR and its JIT caching mechanism that is expediting the running of the algorythm.
beyond that poing i totally agree that the power in a framework of any sort even Rails is the RAD capability. our customers are paying for our time and not for our looks or ability to run it a few milisec faster. they want their applications running flawlessly yesterday. if it will take a few secounds more then thats fine we are doing IT not RT.
i want to enter a simple relm of implementation and not to be dwelling with memory management and CPU cycles when i am getting a record from a DB.
this is not the world though and we all have to take good care of how we use our data and the way it is stored and restored.
the third part of the post is treating the data aspect of programming and it is the most interesting in the aspect of innovation. the ActiveRecord of rails is sounding like a solution i was looking to implement to a legacy tool we have here. think about the simplicity we are achiving by using this tool. as an O/R mapping enthusiast i can see the light now. it seems like a more elaborate run time coupling to data objects. this relates to the earlier idea of time saving at development and the price cut achived by that.
i belive that in the act of Data access with only select update and insert there is a limitation since the optimization gain achived by running data queries in the SQL Server or Oracle or DB2. though it limits you to a server i even wrote a system that generated the ANSI SQL and went with it to any DB vendor.
in this i think that sepecializing a programmer in to the SQL realm is a great power boost to the project. the SQL can be a great tool for making your application a faster and simplify the interface to the DB.
though it does make sense to run a simple select to get the data and a simple insert and update to insert the data.this methodolegie will chalange the developer and the coding technique to deal with data in a non standard way and to make the data as objects will not overrun the dificulties in getting a programer to comprise the new technique.
there is a really significant enchancement in performance achived with a good DBA although it pops the project price up. the boost he can get from the DB when properly alligned is enormous.