I am asked for esmitation of JSF appln

[Shiv Sidhaarth]

Hello Ranchers,

Have you been asked for estimation?

I have no experience in JSF. I have to do dual activity of learning and implementing the JSF appln. What are the factors to be considered for estimation and which portion generally hav hiccups or take enormous amount of time?

Kindly help me!

Rgds,
Sankar

[Shiv Sidhaarth]

I did guesstimation as I don't have exp in JSF. After putting together all effort, it has come to 6500 days....Any comments on this?

Is this too high for a JSF project?

I can say there are about 50 screens..Complexity is quite high as data has to be grabbed from mainframe. Our task has been made simpler as we bought third party components to get the data from mainframe. Still it is coming to 6500 days.

I welcome your feedbacks

[Gregg Bolinger]

17.8 years? Yea, I'd say that is way too high. I completed a project while learning JSF in about 6 months. It had about 15 screens. I was also learning Hibernate and Spring at the same time. And I did it alone.

[Rajeev Ravindran]

hi enperu , are you really serious ?? 6500 days for a JSF project ? what will happen if sun come up with something much better than JSF after 6000 days ?

Thanks,

[Raphael Germano Boschiero]

hi friend's

i'am work in a company where mi work is get the informations at mainframe and transform for JSF. any problem send a mensage to me my email is


[email protected]
MSN:[email protected]
Thank's

Raphael Germano Boschiero

[Khopkar Amarish]

I think he means 65.00 days

[Shiv Sidhaarth]

Guys,

Even higher versions of JSF itself may be far far better than current one. So, as you all think, its not feasible or possible to do the project for 17 years by single person. And one person cannot work for all 365 days. So 17 yrs also is too less my dear. Subtract 52 weekends, holidays, leaves...216 days per year is considered and it comes to 30 man years.

If we have 10 persons, then we could finish by 3 years. We may have 3 releases. But it is in the premature stage to confirm anything as we all are infants in this new project....Thanks Raphael for your id. I have noted down so as to contact you for any help.

Thanks,
Sidhaarth

[Gregg Bolinger]

enperu

No offense, but if your team of 10 can't complete a 50 page project using any technology in less than 3 years, I'd consider getting a new team. Things are not that complicated these days. At least web applications aren't.

[Shiv Sidhaarth]

Hi greg,

Sure, I will go back and see what's wrong with the estimation.

Thanks,
Sidhaarth

[Raphael Germano Boschiero]

Ok


Raphael Germano Boschiero

[Varun Khanna]

We did prepare a low level estimation template for our JSF application earlier. Here is how we estimated tasks per transaction :

1. Number of JSF controls present in page and complexity of each (we defined 3 levels of complexity and each level had predefined time limit)
2. Number of associated classes such as backing bean & helper classes (if any). Here again we had 3 level of complexities.
Note : Avoid doing estimation for simple setter/getter methods in bean for JSF components since that time should be included in the step 1 itself.
3. Number of action classes and again had level of complexities for same.
P.S. You may keep some buffertime too, in our case it helped since we spent lot of time in customization of our application.

This was the basic approach we used and more or less it worked fine for us.
[ January 02, 2006: Message edited by: Varun Khanna ]

[Shan Karawita]

Hi Sankar,

Good luck for your JSF project .

Rgds,
[ January 20, 2006: Message edited by: Shan Karawita ]