Profit in the Middle Ages
I have been struck recently by the idea of profit. "red herrings and white elephants", a sort of entymological-lite book on the origin of english phrases put me on to it.
The middle Ages saw the rise of the modern economy, middle classes in suffcient numbers, beginnings of democracy and a capital system not directly linked to the land. It also saw the rise of the term "bakers dozen". This means 13 not the more usual 12. And it apparently arose from bakers selling to a middleman would give 13 loaves not 12. The middleman would then sell the loaves to the public direct, using the 13th loaf as profit.
Now profit of 1/13th (7.7%) is quite a bit less than todays "average" profit, which according to memory is 10-15% (1/7th) (10% is enough survive 15% is enough to grow). (20% is what R. Murdoch aims for in all his businesses)
So what has changed in 400 years to double the amount of profit a business can or should make?
When I think of it I'll get back to you.
An alternate view on this that may throw all my clever calculations in the bin:
http://www.straightdope.com/mailbag/mbakersdozen.html
Static can be better...
Its not often that I wish Python was a statically typed language.-----------------------------------------------------------------
I have been playing with making my python - database interfaces a little more, well interesting.Specifically Tim at work showed me apachedb. It takes a database defintion and creates from that definition a setof java class files that represent each table, so you can (presumably - i have not used it) suck in data, create an object then "save" it to database and so on.
Aha I thought lets give this a go in python.
So firstly I start with making an object directly from a recordset::
class _recordset: def __init__(self, dict): for key in dict.keys(): self.__dict__[key] = dict[key]
def getRSasobj(cursor, SQL): ''' for a given cursor, and SQL, return RS as an object
improvements: somehow cast the items as they should be in ''' cursor.execute(SQL) desc = cursor.description rs = cursor.fetchall() #we now have the data as a tuple, and the col headings as a tuple names = [] [names.append(col[0]) for col in cursor.description] objlist = [] for row in rs: #zip the columnnames and the row data, then pass into a dictionary. #Now pass dictionary to mk a simple rs object, and append to list of objects objlist.append( _recordset( dict( zip( names, row) ) ) ) return objlist
The above just makes an instance of _recordset holding whatever data was in the returned row. (I guess it should be _row then).For example I can get x = getRSasobj(c,"Select Id, name from tblfoo;") and then see x[0].name
Ok its not brilliant but it works and when I refactor it it may be usable elsewhere
However i now want to try this apachedb thing.
SO first off - the plan
the plan--------(This is very mysql specific but hey!)
- get all tables in dbase "show tables"- get the table def for each table "DESCRIBE tbltask;"- Create a class module that holds the defintion Here is where I am coming unstuck.
I can create a class module fairly easily ( "class generic_%s:" % tablename ) but when I come to defining the attributes in a class (lets say the self.id and self.name) I hit the static problem. In Java (and i bet in apachedb) one would just declare (pseudocode - I canot rememebr the Java) ::
declare ID as INT or Datetime or Varchar(32) or whatever
But how od i do this in python?I am going to have to build seperate checks in ...similar to:: self.ID = parameter_ID if type(self.ID) != 'int': raise MyErrorClass
This seems overwrought...
Whilst digging through the proverbial "legacy code", my collegue Tim pipes up.- "Oh thats bloody helpful"
- I have an error message: "Error: it cannot be resolved."
Suggestions of using "it cannot be resolved - again" as the verbose option are thrown around the room.Thats how sad we are.
However all is not lost. Tim's lightening brain soon finds the root cause of the problem.
"Oh, there actually is a variable called 'it'."
Such is the lot of a programmer.
