-2

How long does it take to extend the functionality of a pos terminal?

I don't have any idea how long does it take to develop / extend functionality of a POS terminal. I have a friend who is asking me for this and we came across the "How Long Part".

My friend already has a running pos terminal that we developed with MQX Operating System by freescale (we are using K60 MCU) we bought it at aliexpress.

We were planning to extend the current source for NFC / contactless functionality (we also bought the reader from aliexpress)

Is developing it in 6 months time for 4 C++ developers sufficient enough?

artless noise
  • 21,212
  • 6
  • 68
  • 105
RGAT
  • 486
  • 3
  • 13
  • The answer is dependent partially on difference between the current functionality and required functionality, neither of which you have described. It is also dependent on the performance of the developers, which in turn is multi-factorial. The question has no direct answer, so is not really suitable for StackOverflow. – Clifford Oct 30 '13 at 05:34
  • the required functionality is to add an NFC module. "to extend the current source for NFC / contactless functionality" I also have described the current functionality which is a POS Terminal. I already got an answer from a different source. He answered that it is feasible since their team extends Car GPS systems in less than 6 months. I don't know why you have said that it is not suitable for StackOverflow but am seeing some Arduino answers over here. So I think it is much more proper to say that there is none that has MCU development experience that saw this question. – RGAT Nov 06 '13 at 00:52
  • If you have to ask the question, it rather suggests that your team lacks existing experience, so the experience of the guy that says he does something similar in 6 months is worthless *unless* he is working for you. How does he know that it is at all similar or that the existing code is at all well designed or maintainable. It is unsuitable for StackOverflow for the reasons I described and the more general reason stated in the close notice. – Clifford Nov 06 '13 at 09:30
  • Thanks for your reply Clifford, we are already making it and is nearly done. We are expecting to test the prototype module by next month to the existing pos terminal, I have also asked the guy for consultations since he has experience. The existing source code for the main pos terminal has passed certifications for level 1 and level 2. And is well documented. – RGAT Nov 07 '13 at 06:02

1 Answers1

2

To quote a classic:

[Y]ou’ve got to decide if you can deliver a Fortran compiler from scratch in seven days. You’ve got a staff of two programmers to help you.

Can you do it?

“Well, I suppose, it depends,” you say.

On what?

“Um, will my team be able to use UML-generating tools?”

Does that really matter? Three programmers, seven days, Waterloo Fortran IV. Are UML tools going to make or break it?

“I guess not.”

OK, so, what does it depend on?

“Will we have 19 inch monitors? And will we have access to all the Jolt we can drink?”

Again, does this matter? Is caffeine going to determine whether you can do it?

“I guess not. Oh, wait. You said I have a staff of two programmers?”

Right.

“Who are they?”

Does that matter?

“Sure! If the team doesn’t get along, we’ll never be able to work together. And I know a few superstar programmers who could crank out a Fortran compiler by themselves in one week, and lots of programmers who couldn’t write the code to print the startup banner if they had six months.”

Now we’re on to something!

If your team has worked together before, if you set your features to match your deadline (not the other way around), and if you have a keen sense when you've gone off the rails, probably. Presume you can, multiple your estimates by four, and carry on!

(Seriously, I can't answer this technically becuase it's a people question. But I hope it's inspirational! Good luck.)

bishop
  • 37,830
  • 11
  • 104
  • 139
  • There is no way to do this in six months. The certification of the different EMV levels will take months. You have a level 1 and level 2. Things rarely pass the first time as the labs have a vested interest in failing you. This is outside the scope of the development staff. Developing a Fortran compiler is easier. Perhaps not a full feature highly optimizing Fortran compiler. But one that can produce code can be made in a month or less. Level 2 libraries are huge as each card has different protocols. There are also (at least) two types of level 1 'base band' protocols. – artless noise Apr 25 '19 at 16:22
  • @artlessnoise Well, perhaps they could leverage existing libraries to handle the wire protocols. Perhaps they could exclude the certification time from their calculation as it is, as you say, outside of scope. Or perhaps that certification is already done as per a comment on the OP. Perhaps they could implement fewer niceties of the protocol or of any frontend they had envisioned. There is always some kind of measurable output they could achieve in a given timeframe. Whether that constitutes an MVP for purposes of this question are opinion based. Hence its closure NEARLY SIX YEARS AGO. :) – bishop Apr 25 '19 at 21:24