| Author |
Message |
    
magmasystems
Citizen
Username: Magmasystems
Post Number: 231
Registered: 1-2002
| | Posted on Monday, April 12, 2004 - 10:42 am: |     |
Curtis,
In general, the class is independent of programming language. In RUP, you would define an "abstract" class and pass it off to the developers to implement in the programming language of choice, whether it be C++, C#, Java, or Smalltalk.
The architect doing the design should have a very good understanding of object-oriented programming. Meaning that he should understand things like inheritance, interfaces, virtual functions, polymorphism, etc. The RUP design should reflect an entire object-oriented hierarchy.
Now, to modify my very first statement, to some degree, the RUP will reflect the programming language. For example, C++ has something called Multiple Inheritance, where a class can be derived from multiple "base" classes. Java and C# do not have this... instead, classes can IMPLEMENT one or more INTERFACES. More tech-speak, but it will be important in the RUP design.
=====================================
Another methodology that people use now is something called Extreme Programming, or XP for short. This involves rapid turnaround and almost immediate feedback from the user. Some shops who were formerly rooted in RUP are now turning to XP for the reason (as Brett mentioned) that the whole RUP cycle can be tedious and time-consuming.
Marc
All Things Computered
www.millburnweb.com
|
Brett
Citizen
Username: Bmalibashksa
Post Number: 942
Registered: 7-2003
| | Posted on Monday, April 12, 2004 - 10:45 am: | |
Marc,
When you say
IMPLEMENT one or more INTERFACES.
do you mean
IMPLEMENT one or more INSTANCES.
or do C and Java differ there? |
magmasystems
Citizen
Username: Magmasystems
Post Number: 232
Registered: 1-2002
| | Posted on Monday, April 12, 2004 - 11:05 am: | |
Brett,
The INTERFACE and the INSTANCE are two different things.
In C++, C# and Java, an Interface has no code in it. It just contains the "signatures" of the functions that a class must IMPLEMENT in order to IMPLEMENT the interface.
I have not programmed Java for quite a while, but I believe that in Java, the keyword 'implements' is used when a class implements an interface, and the keyword 'extends' is used when a class is derived from a base class.
Also, as far as INSTANCES go, you INSTANTIATE an INSTANCE of a CLASS. (This is what is known as an object.. an instantiation of a class).
Curtis, if you aren't disuaded yet from doing this stuff, then you'll be fine at your job :-)
Marc
www.millburnweb.com
|
Brett
Citizen
Username: Bmalibashksa
Post Number: 943
Registered: 7-2003
| | Posted on Monday, April 12, 2004 - 11:10 am: | |
I see what you were saying. |
Dave
Citizen
Username: Dave
Post Number: 6820
Registered: 4-1998
| | Posted on Monday, April 12, 2004 - 11:19 am: | |
Curt,
Ignore most of this stuff. :-)
All you need to know from the programmers (or preferably, be able to TELL the programmers) is what variables are being passed from the form to the method (aka function) and what you want returned from the function. You then shape the output in a way that makes sense to human beings. |
magmasystems
Citizen
Username: Magmasystems
Post Number: 233
Registered: 1-2002
| | Posted on Monday, April 12, 2004 - 11:32 am: | |
Dave,
I would also like to tell Curtis to ignore this stuff. The RUP has no bearing on whether he can come up with a good UI or not.
However, since the Internet bubble burst, there are many, many UI designers looking for work. By UI designer, I mean people who are well versed in HTML, Flash, maybe some light Javascript, and can come up with an attractive UI.
If Curtis' job description calls for a knowledge (however cursory) of RUP, then it may very well be this knowledge that distinguishes him from the hundreds of other UI designers applying for the same position.
Marc
www.millburnweb.com
|
Brett
Citizen
Username: Bmalibashksa
Post Number: 944
Registered: 7-2003
| | Posted on Monday, April 12, 2004 - 11:36 am: | |
I was just answering the questions that he had. I agree he shouldn’t be versed in all of the things that we’re throwing at him. I’ve also never hired a GUI designer; I need my people to work over a broader spectrum then that. |
woodstock
Citizen
Username: Woodstock
Post Number: 568
Registered: 9-2002
| | Posted on Monday, April 12, 2004 - 11:56 am: | |
I have an old version of the RUP from back when Rational was independent.
http://mark.klomstock.com/rup/ |
magmasystems
Citizen
Username: Magmasystems
Post Number: 234
Registered: 1-2002
| | Posted on Monday, April 12, 2004 - 11:57 am: | |
Brett,
Where I am consulting now, we have a full-time employee whose job it is to do UI. She usually takes my nice, well thought out prototypes, and in her UI tool, changes things to a point where I have to spend several days fixing the underlying code.
The only thing that I have seen that may be more of a waste of space/money is "data modelers". All of your developers should know how to build normalized databases with proper indexing. Everytime I speak to full-time "data-modelers", they basically nod their heads and go "uh-huh".
Marc
www.millburnweb.com
|
Curt Wayne
Citizen
Username: Cswayne
Post Number: 120
Registered: 10-2002
| | Posted on Monday, April 12, 2004 - 12:10 pm: | |
All,
What Marc said is exactly the case. I'm an HTML hacker looking to redefine myself as a U/I designer and the company wants what I have with an understanding of RUP thrown in to boot.
I've been working around programmers for 20 years so I can usually follow along in conversations such as this. All info is eagerly accepted. However what I need most is the broad overview of the process and then have that overview broken down into it constituant pieces and explained again. Dont want too much huh?
an an emphasis on what my role as UI designer would be if that can be figured out. Examples of others designers work would also be great if any one knows where I can find them.
But, this is great! I just hope I can learn enough by the time the interview rolls around.
Curt
http://www.cswayne.com
|
Dave
Citizen
Username: Dave
Post Number: 6821
Registered: 4-1998
| | Posted on Monday, April 12, 2004 - 12:27 pm: | |
Your role is to make changes on changes so Marc gets more consultilng hours.
Why anything is coded before a UI is signed off on is odd. Why software is created BEFORE a users' manual is written is also odd. Everything seems backwards in the land of programming. The right order should be 1) write a manual about how the software will work, 2) create screenshots of how it will appear, 3) write the software. |
magmasystems
Citizen
Username: Magmasystems
Post Number: 235
Registered: 1-2002
| | Posted on Monday, April 12, 2004 - 12:32 pm: | |
Curtis,
I am not sure if this site will cover what you may be asked in a UI designer interview, but you can try:
www.techinterviews.com
Always good to see the kinds of questions that other UI designers were asked...
Marc |
Brett
Citizen
Username: Bmalibashksa
Post Number: 945
Registered: 7-2003
| | Posted on Monday, April 12, 2004 - 1:08 pm: | |
Dave,
That's an interesting point but what happens if you're converting a Fat Client over to JSPs? The back end is already there and the front ends need to be written, same holds true for the Lite pages that need to be created for wap devices.
I think you also have to look at this in context of what’s being developed. A website that’s being built from scratch is a little different that a Legacy system that has 8 million lines of code and is growing every day.
Marc,
That UI designer would be lynched in this office. I can’t tell you how many times someone has come in complaining of rework. Luckily Struts has changed that buy keeping the UI and Data separate.
|
Tom Reingold
Citizen
Username: Noglider
Post Number: 2798
Registered: 1-2003
| | Posted on Monday, April 12, 2004 - 2:28 pm: | |
Dave,
The approach you describe is food for plenty of thought. I learned to program while reading books of Kernighan, Plaugher, Ritchie, Pike, et al. They advocate writing a manual before the software.
I think it's idealistic and purist and impracticable. As you write code, you learn that some things are not as important as you originally thought, or they are harder than you thought. You learn that other features are required that you hadn't anticipated.
Of course, if you give in to these findings, you end up writing and unorganized mess, so perhaps a blend of the two approaches is best.
Donald Knuth advocated PROVING your algorithm or model before coding it. What a nice ivory tower he lives in. I don't think anyone follows his advice. |
Curt Wayne
Citizen
Username: Cswayne
Post Number: 121
Registered: 10-2002
| | Posted on Monday, April 12, 2004 - 2:53 pm: | |
Hey Prissy Pants,
I wondered how long it would take you to join this discussion.
Why dont we model the "Maplewood Parking Spot Finder Ap" in UML/RUP???
Curt
|
Tom Reingold
Citizen
Username: Noglider
Post Number: 2800
Registered: 1-2003
| | Posted on Monday, April 12, 2004 - 2:55 pm: | |
Jeez. I shouldn't have butted in (or should I say pantsed in?), because I barely qualify to discuss the topic at hand. It's been a long time since I wrote large amounts of code for a living. |
Dave
Citizen
Username: Dave
Post Number: 6823
Registered: 4-1998
| | Posted on Monday, April 12, 2004 - 3:30 pm: | |
Here's the barebones version for parking....
Class FindParkinginMaplewood {
function illegalLeftTurn
function infiniteLoop
function doublePark
} |
Brett
Citizen
Username: Bmalibashksa
Post Number: 946
Registered: 7-2003
| | Posted on Monday, April 12, 2004 - 3:35 pm: | |
I thought it would just throw a "Null Pointer Exception" |
Curt Wayne
Citizen
Username: Cswayne
Post Number: 122
Registered: 10-2002
| | Posted on Monday, April 12, 2004 - 3:51 pm: | |
Tom,
I'm sorry, I didnt mean to cast any dispersions...
I'm serious about the parking app. I think what I need is a real life situation to model to have this all make more logical sense.
and you know that the parking thing is near and dear to my heart.
Curt |
Tom Reingold
Citizen
Username: Noglider
Post Number: 2803
Registered: 1-2003
| | Posted on Monday, April 12, 2004 - 3:54 pm: | |
Oh, it's quite all right. I'm rolling with the punches. I'd be glad to sit down with you about the parking app. It's intriguing, but I really don't feel Maplewood has a parking problem. I know most people disagree with me. Still, the app could be a great invention where there are real parking scarcities. |
Cynicalgirl
Citizen
Username: Cynicalgirl
Post Number: 526
Registered: 9-2003
| | Posted on Monday, April 12, 2004 - 7:39 pm: | |
Right on, Dave! What you said.... |
Curt Wayne
Citizen
Username: Cswayne
Post Number: 123
Registered: 10-2002
| | Posted on Tuesday, April 13, 2004 - 12:56 pm: | |
Hey Cynicalgirl,
You know all about RUP don'tcha?
Now, back to the parking in Maplewood Ap...
Here are some notes I jotted down. Please help me put this into logical order to begin a "scope, vision, business case" document. and Tom I know you must have alot to add.
Ah hem...
The Satelite Scenario"
"Find a Parking Place In Maplewood"
User needs to park his car near the train station
FORM elements:
SELECT where:
1 block
1/4 mile
1/2 mile
1 mile from station
When: INPUT TIME
Permit: YES / NO
</form>
If Permit=true then search DB "all" If NO=true then search NON-PERMIT only.
HOW:
Satellite in geosynchronous orbit identifies available spaces within 1 mile radius of station .
Parking spaces are mapped using _______________ technology
Upadated in real time
Data stored in Database
BIlling: per use (req. ecomm app)
http://www.cswayne.com/rup-params.html}
|
Curt Wayne
Citizen
Username: Cswayne
Post Number: 125
Registered: 10-2002
| | Posted on Friday, April 16, 2004 - 1:36 pm: | |
Looks like this thread is dead...
Curt |
magmasystems
Citizen
Username: Magmasystems
Post Number: 239
Registered: 1-2002
| | Posted on Monday, April 19, 2004 - 8:09 am: | |
Well, you can let us know how you came out on your interviews, and whether any of the advice that we offered helped you out (or hindered you)....
Marc
www.millburnweb.com
|
Closed: New threads not accepted on this page