if it exists use it? or write it?
 

ok, one last question!

how do you feel about using open-source/free libraries for a big commercial project?

what im asking is would you hesitate to use libraries like Boost for a commercial project? or would it be safer to just write all the code from scratch?

suppose Boost were to withdraw the licence (hypothetically) that would cause problems with the licencing etc of the software developed using it?

i feel its probably best to just write the code from scratch (i.e. reinvent the wheel) because you have total control over it that way?

what do you think?

thanks!

:D

Sure, if you want to waist all that time. I am sure that if you are using a version that was before licensed as open source and usable then their would be no problems.

its not about wasting time, its about making it easy to extend your software, modify etc...

but i suppose nowadays its best to just use available libraries since software dev has become a fast paced business these days (generally i suppose anyway) ???

Depends on a number of factors.

Personally if there is an open source option available that meets both the functional and non functional requirements of my system then I would almost certainly use it.

On occassions I have found interoperability issues or performance issues that has required me to either write my own code or alter the existing code libraries. This is all 100% dependant on the proposed projects deadline, which often leads in trade-offs that I consider the best option at the time.

You also need to consider if ou have the resources to maintain your own libraries as good as existing open source libraries such as boost.

However again, this is all dependant on the scope of the project which in my opinion goes on a case by case scenario.

i guess thats right...

it feels nice if you write your own code from scratch, and not derived from available libraries...

still i guess the smart programmer will use available resources...

You do understand the libraries available like boost and such took ALOT of time to put together and for a beginner coder (I am guessing you are from your posts) you be better off using libraries that are there until you think that you are efficient enough in the language you code in to makes your own.

Usage of existing libraries would be ideal in a tight time frame. Although, you have to be careful in which libraries you use when developing commercial applications. If there is a doubt that will be a conflict in the license with your deployment, then you should re-evaluate the options. Also, it is a good learning strategy to reinvent the wheel, sometimes the wheel is smoother when you write it yourself. If the library is complex or there is a time issue, then use the tools that are already available.

Reusability is key to being an effective designer, regardless of the discipline. If something isn't usable to start with, it's reusability is obviously questionable.

There is a difference between acquiring knowledge and production. A modicum of thinking and research will probably point the way in most instances.

I feel that you are missing on on the learning process if you just use libraries like boost (no doubt it took a ton of effort to develop, and for a single person to write a similar, and just as good library would take forever), but i suppose you still learn both ways, suppose you dive into writing your own library, you might have a couple of questions along the way, but if you first start by working with libraries like boost and the STL, you will gain some knowledge of how one might implement a similar library, so its a learning process both ways!

im not a programmer by profession (actually i work in visual effects, and i do alot of scripting - which led to my fascination in programming!), im still learning all this stuff, and ive learnt so much from just writing my own libraries by coming across mistakes ive made, that made me understand more...