Python advice on CME datafiles...

[bali_survivor]

Quote from nononsense:

Great words of wisdom. You couldn't sum it up better prt.

I'm not sure wether $soft can do any better. They are simply too lazy and dumb scrooges compared to the army of nimble smart developpers working and improving the OpenSoftware tools like Python.

MicroBrown Matter(starts also with an S and ends with a t and the two letters between the S and the t are also a common greeting ) has the legacy of attempting to "test" quality into their product. Hence a beta rollout to 80000 users. But having such a large beta test group implies automatically that they are involved in "spaghetti coding". Who can forget word version2 being released with over 8000 "known" problems? Or Windows ME that in MS wisdom does not need to adress more than 512Mb of RAM (although there is a fix for that). And going back: Gates uttering that the ordinary user would never need more than 640kB RAM (DOS)?

And working in that fashion is a very inefficient way of doing things.

Maria

 

[prt_systems]

Quote from bali_survivor:

..... Hence a beta rollout to 80000 users. But having such a large beta test group implies automatically that they are involved in "spaghetti coding". .....

And working in that fashion is a very inefficient way of doing things.

Maria

The betas are are much about marketing as they are about "testing" ... I look at Operating Systems, Desktop environments and servers as dialtone: they are simple commodities required to run a business. However "dialtone" needs to be ultra reliable and trusted. Open Source is cheap - no fee - transparent since you have the source code, which implies very low risk, and in my opinion, tested to a higher standard and with a better algorithm.

I use $soft if the situation requires it but for all new systems development - server deployed and desktop - my first choice is open source.

 

[bali_survivor]

Quote from prt_systems:

The betas are are much about marketing as they are about "testing" ... I look at Operating Systems, Desktop environments and servers as dialtone: they are simple commodities required to run a business. However "dialtone" needs to be ultra reliable and trusted. Open Source is cheap - no fee - transparent since you have the source code, which implies very low risk, and in my opinion, tested to a higher standard and with a better algorithm.

I use $soft if the situation requires it but for all new systems development - server deployed and desktop - my first choice is open source.

I do not like to be held to ransom by Mshit with their "activation codes". You rebuild a PC and have to reinstall the software (new HD, etc etc) and then you will have to ring up for reactivation of office. "You are running illegal copies, no we do not give a new code to you...". What a bunch of w&**rs! If they would price their software reasonable then the open software would not have a chance. But no, they have to hold the rest of us at ransom. I do not want to have an upgrade / patch etc forced down my throat. After all my requirements do not change a lot, a letter that I wrote in word 2 years ago I can still write in word 2. I do not need all this "internet enabled" shit or "collaberation" stuff.

It only opens up the PC for hackers and hence I have to take even more steps to make my environment secure. The cost of software maintenance is far outstripping the cost of hardware. Look at the purchase price of the hardware versus the software! And that is not taking in account maintenance either!

Maria

 

[prt_systems]

Quote from bali_survivor:

....
... The cost of software maintenance is far outstripping the cost of hardware. Look at the purchase price of the hardware versus the software! And that is not taking in account maintenance either!

Maria

When we compute the true hold cost of running a particular software platform we include this cost. $soft never talks about this cost. It is substantial and when correctly measured and included in the true hold cost of $soft systems the cost dwarfs that of open source stacks.

 

[Enginer]

Quote from bali_survivor:

[BAt this moment my needs are for some simple interfacing but I am wondering: is there any charting out there that is using Python?

Thank you
Maria [/B]

Troy Milhase wrote a set of Python modules for accessing Interactive Broker's feeds. It triggered a background project for me to learn Python programing. I am especially interested in Troy's charting in Python, but i cannot say I have them running yet--a fur piece to go for an old man...

"What is ProfitPy?
-----------------------------------------------------------------------------
ProfitPy is a set of libraries and tools for the development, testing, and
execution of automated stock trading systems[1]. The main web site for the
toolkit is here:

http://profitpy.sourceforge.net/ "
--from the ProfitPy Readme file

 

[brokershopping]

More good info... Thanks!

As I look into python packages, one potential pitfall I see is that there is often a hierarchy of dependance on other packages. Does this become much of a problem when upgrades are issued?

 

[nononsense]

Quote from brokershopping:

More good info... Thanks!

As I look into python packages, one potential pitfall I see is that there is often a hierarchy of dependance on other packages. Does this become much of a problem when upgrades are issued?

If you look at a system like Linux, it is amazing how much Python goes into different components of a distribution. If there would be any weakness like you are asking about, a linux distribution would simply not be usable at all.

However, if you are going to use Python as a developper, you are basically in the same boat as a developper using any other language: this is pretty much up to the user.

Some questions exist as to the suitability of Python for large-scale projects. Few languages qualify, in fact only C and C++, perhaps Delphi. AFAIK, no examples exist of Java and C# when you talk 'LARGE' scale. Google might perhaps qualify for Python???.