3 - 9/7/01 6:49:03 AM - barry: Good morning, everybody! As with last week, I'll start by entering answers to the most frequently-asked questions submitted. 
4 - 9/7/01 6:50:01 AM - barry: Q: How does risk management assessment vary with the size of the project?
5 - 9/7/01 6:52:58 AM - barry: 4 Risk assessment effort is correlated with size, but not absolutely. Some small applications have to do a lot of risk assessment. Examples are safety-critical aircraft control or medical devices, or the security kernel of an operating system. As these examples illustrat, risk assessment effort is also correlated with the ambition level of your level-of-service or quality requirements.
. - 9/7/01 6:54:29 AM - James K has entered the room
6 - 9/7/01 6:54:53 AM - barry: Q: Suppose two stakeholders are locked in conflict, and neither wants to back off. What does a project manager do to resolve this?
. - 9/7/01 6:59:34 AM - ahmad has entered the room
. - 9/7/01 6:59:40 AM - ahmad has entered the room
. - 9/7/01 6:59:54 AM - alsawi has entered the room
. - 9/7/01 7:00:27 AM - ahmad has timed out.
. - 9/7/01 7:01:04 AM - temp1 has entered the room
7 - 9/7/01 7:01:33 AM - barry: 6 There are several cases. If they are the manager's subordintes, the best thing to do is to transfer one of them to another project (but in a way that makes the person transferred look like a winner). If it is two customers, your best strategy is to escalate the problem up the customer management chain, since the conflict is not generating a win for customer management. If it is a subordinate-customer conflict, there are two main cases. If your subordinate is pointing out a disastrous customer position, your best strategy again is to escalate the problem within the customer's management chain. If the conflict is about a matter of taste, it's best to impress on your subordinte that in such cases "the customer is always right." 
8 - 9/7/01 7:03:17 AM - barry: Q: What do you do if some losses are unquantifiable, such as quality image or customer-supplier good will?
9 - 9/7/01 7:06:29 AM - barry: 8 You can try for quantitative proxies, like scale-of-10 ratings and weighted-sum figures of merit (covered in Chapter 15 of SW Engr. Econ.). Or you can use visualization techniques like polar graphs, red-yellow-green-blue ratings, etc. (covered in Chapter 18).
10 - 9/7/01 7:07:34 AM - barry: Everybody -- Feel free to submit followups to these answers or submit additional questions.
. - 9/7/01 7:12:31 AM - Frank has entered the room
11 - 9/7/01 7:12:36 AM - James K: Regarding conflict between stakeholders, as someone whose project (ongoing, repeated yearly) consists of mostly internal customers, one of the most common areas of conflict exists between the development team and the management team of our primary stakeholder. These conflicts mostly come in the form of expectations over schedule or requirements. I have yet to figure out a really good success model for that. Any thoughts?
12 - 9/7/01 7:13:05 AM - alsawi: Is scale-of-10 a standard scale in relative measuerment, cause in the my internship, the client asked me to do a scale-of-100, which was too fine grain for me.
13 - 9/7/01 7:13:53 AM - barry: Everybody -- Note that if you use the Normal or Threaded (better) agenda options to the right of the screen, the material from last week's session comes up first. The material from this week's session is toward the end. Also remember that when you come back to the main page, the cursor may not be located in the box for composing questions and comments, and you need to click on the box to get the cursor there and enable your typing to be recognized.
14 - 9/7/01 7:20:24 AM - barry: 11 The best you can do is to use objectively-calibrated estimation models like COCOMO II, SEER, KnowledgePlan, or PRICE S to show that the estimated resources are much bigger than the allocated resources, and use the models to explore ways to meet the available resoureces via the cost drivers like reducing effective size via reuse or deferring lower-priority features; adding more experienced people; relaxing high performance constraints; etc. Even better is to gather data on your own organization's experience and use this either for direct productivity comparisons or to calibrate the estimation models to your organization's experience.
15 - 9/7/01 7:22:56 AM - temp1: Could you please explain application of the concepts of discontinuous technology to software development and its impact on applicability of combined cost schedule strategy. Buy-into innovators dilemma makes concurrent competing on schedule, quality and cost possible due to replacement of one (outdated) technology by newer, more effective technology.? 
16 - 9/7/01 7:23:34 AM - barry: 12 Scale-of-10 is the most frequently-used level of granularity. You can always get a little more granularity by rating things as (for example) 6.5. This usually works better than rating it 13 on a scale of 20, or 65 on a scale of 100.
17 - 9/7/01 7:23:38 AM - Frank: When trying to give subordinates what they want in a win-win relationship, how do people know what they really want? 
18 - 9/7/01 7:26:09 AM - alsawi: Why is the emphasis is on meeting the schedule or SAIV, in other word, why reducing the features while maintaining the schedule is better than extending the deadline and completing the agreed features?
19 - 9/7/01 7:27:21 AM - temp1: With the business strategy that subscribes to lay-off of 10% of employees every year, technically from the bottom of the totem pole, how practical is assumption of Win-Win management? One of the most powerful reasons for job security is having status of one-of-a-kind experrt. That makes difficult sharing of technology and experiences between different employees difficult. While very attractive in theory (everyone working together towards the good of all), how practical is this theory in practice, given predominant business practices currently in place? Also, I did not have an option to enter my name while entering chat-room.
20 - 9/7/01 7:30:16 AM - barry: 15 Right. Often legacy constraints such as database conversion or impossibility of modularizing the legacy code to permit incremental replacement will constrain your solution options a lot. Other problems are predicting how long a current best-solution will remain being the best (for example, frequently early startups get overtaken by bigger late-entrants). The best you can do is to try to reengineer your legacy system and architect your new system for ease of evolution.
. - 9/7/01 7:32:39 AM - Techsupport has entered the room
21 - 9/7/01 7:34:06 AM - barry: 17 Frequently people don't know what they want. For example, they may think that the grass is greener on the other side of the fence, and then find out that it isn't. One strategy for this is to suggest options to people and have them think things over, talk to others, or try out some of the new tasks, before committing to a change.
22 - 9/7/01 7:36:04 AM - James K: 18 If you don't mind feedback from a fellow student, I would say that schedule might be important from the standpoint of accomplishing the basic mission for the organization (assuming an internal customer). If they can deal with slightly less functionality, but at least have the essentials, you are probably better off. Additionally, confidence in your ability to produce is eroded if you draw out the schedule too much. For a commercial application, I think the simple answer is loss of market share.
23 - 9/7/01 7:39:00 AM - barry: 19 It's not hard to log out and log back in with your real name. Arbitrary strategies like "lay off the bottom 10%" are frequently dysfunctional because they break up well-jelled teams, stimulate inter-employee competition rather than cooperation, and undermine the Security need deep down in people's Maslow need hierarchy. DeMarco and Lister have a lot to say about this in Peopleware.
. - 9/7/01 7:40:42 AM - Techsupport has left the room.
. - 9/7/01 7:41:49 AM - Steve has entered the room
. - 9/7/01 7:42:22 AM - Techsupport has entered the room
24 - 9/7/01 7:43:01 AM - barry: 22 Right. In new market areas, "first to market is first to the bank." Elsewhere, getting a core capability installed and used early generates savings earlier and provides usage feedback on the real priorities of the features not yet implemented.
. - 9/7/01 7:43:55 AM - DougT has entered the room
25 - 9/7/01 7:47:44 AM - barry: For new entrants, clicking on the Threaded Agenda is a good way to catch up. This week's chat is toward the end, after last week's.
26 - 9/7/01 7:50:14 AM - barry: Q: What are the most important tasks for a Theory W manager?
27 - 9/7/01 7:50:36 AM - DougT: It appears to me that we often do our best estimates to come up with a realistic cost and schedule, then for business reasons you're cut 30%. I've seen folks then pad their estimates knowing they'll be cut. Any good solutions to this dilema?
. - 9/7/01 7:51:59 AM - James K has left the room.
28 - 9/7/01 7:52:05 AM - alsawi: For an ongoing project that has started without Win-Win or having a shared vision, and the project has reached the architecture design stage, how can they go back to Win-Win negociation while maintaining the progress.
29 - 9/7/01 7:52:37 AM - temp1: Establish value models of each stakeholder involved and come-up as a result of negotiation process with the meaningful reconciliation and of those value propositions and agreement to cooperate. It is critical that each stakeholder has a sense of win for himself, since that committs him to ensure success of a project as a whole.
30 - 9/7/01 7:53:21 AM - Frank: 26 Communication between everyone is most important, in my opinion. Setting goals too. 
31 - 9/7/01 7:54:06 AM - barry: 26 Not issuing orders, micromanaging, designing, or programming (although its good to keep your hand into the technology to some extent). The most important tasks are negotiating among stakeholders, creatively packaging assignments and deliverables to achieve win-win outcomes, stimulating and coaching stakeholders to become more effective and cooperative.
32 - 9/7/01 7:54:59 AM - barry: 30 I think 30 and 31 are very compatible
33 - 9/7/01 7:55:26 AM - DougT: 26 I would agree communication. A big assumption is that everyone is professional and willing for a win-win scenario. That appears often not the case.
34 - 9/7/01 7:56:23 AM - barry: 29 And 29 and 31 are very compatible, if 29 was a response to 26
35 - 9/7/01 7:57:37 AM - DougT: 26 I have executed win-win scenarios successfully, and not meaning to be totaly negative there is just too many variables outside of our span of control. Will CMMI help spread the win-win philosophy?
36 - 9/7/01 7:57:44 AM - temp1: Successfull application of Win-Win models for software development requires that corporate culture in place does not undermine the basic premises of the Win-Win proposition. Frequently however the organizational culture rewards and encourages behavior that is contradictory to the Win-Win mind-set. People look at what is being rewarded and then act accordingly. In my opinion, it is very difficult/impossible(?) to establish W-W for software development, when rest of the business transcribes to counter-win-win practices. The bottom line is, is you want to be successful in software development, you need to address issues related to organizational culture in general.
37 - 9/7/01 7:59:45 AM - alsawi: I agree with 36, I have seen projects the win condition of some stakeholders it to fail the project!!
38 - 9/7/01 8:00:33 AM - Frank: My company, and my former company, have goals. Both want to be employer of choice and best in the business. We are to set our goals according to their stated goals. Is this win-win?
39 - 9/7/01 8:01:55 AM - barry: 33 True. In some case this may be due to personality conflicts. These can often be resolved by redeploying the people with conflicts to different parts of the organization that don't have to interact much. Sometimes, though, you can just get a "bad apple," who needs to be convinced that their future prospects will be better outside your organization, and let go.
. - 9/7/01 8:03:14 AM - temp1 has left the room.
40 - 9/7/01 8:03:51 AM - barry: 36 Right. Cultures can reward win-lose, which makes it very hard to make progress toward win-win.
41 - 9/7/01 8:03:55 AM - DougT: 37 & 38 In an systems driven environment that is hardware intensive, educating your cohort is the largest effort. I must comment that redeploying people is a shifting sand issue where if they are truly incompatabile, move them along outside the company.
. - 9/7/01 8:04:01 AM - A has entered the room
. - 9/7/01 8:06:48 AM - alsawi has timed out.
42 - 9/7/01 8:07:00 AM - barry: 38 This is win-win as long as there are ways for the employees to participate in interpreting the goals.
43 - 9/7/01 8:08:09 AM - A: Well, I've got kicked out by the system under my temp1 code and allowed to re-enter but I could only enter one letter. I guess wonders of technology. I'm sure every company we are working for, have a very lofty goals and objectives and strives to be the best in the business. However in day-to-day practical application sometimes those lofty goals might be suboptimized.
44 - 9/7/01 8:09:13 AM - Techsupport: 43- you usually get timed out if there is a period of inactivity from your side
45 - 9/7/01 8:10:52 AM - barry: 41 Education is often a big challenge. Your hardware-oriented manager may think he/she is making you a winner by adding people to your project to speed it up, unaware that in software Brooks' Law says that this generally slows the project down. 
46 - 9/7/01 8:11:01 AM - Frank: Is there a source I can read to better understand the homework?
. - 9/7/01 8:11:56 AM - alsawi has entered the room
47 - 9/7/01 8:12:14 AM - A: Prof. Boehm, could you comment on the role of the organizational infrastructure (broadly defined) in establishing the win-win environment. That infrastructure includes technical competence on one hand, people skills of employee, understanding the overall business environment issues (that helps in converging value propositions), technological skills etc.
. - 9/7/01 8:12:21 AM - James K has entered the room
. - 9/7/01 8:12:22 AM - James K has entered the room
48 - 9/7/01 8:12:22 AM - barry: 46 The reading assignments should be posted by today at the latest. I'll follow up with the TA on this.
49 - 9/7/01 8:13:59 AM - A: For us DEN students, last Wednesday lecture has not been posted yet on the web (or at least was not posted there yesterday late at night). Is it possible to post lectures the same day? Specially if they are impacting homework due on Wednesday? 
50 - 9/7/01 8:15:29 AM - Frank: 45 I still hear from friends, and have experienced, great new plans and new tools and technoligies. We never get training though. If all you know how to use is a hammer, you will pound that screw in with a screwdriver. 
51 - 9/7/01 8:16:06 AM - Techsupport: 49-i will co-ordinate with DEN about this
52 - 9/7/01 8:16:14 AM - barry: 47 Again, this is an area where lack of education will hamper win-win. Examples are government organizations whose contracting personnel are still stuck in a waterfall-model mode, thinking that linking progress payments to complete requirements specs rather than user-accepted prototypes is the way to get the project to win. Another example is that software people who master application-domain skills become much more valuable to the organization.
53 - 9/7/01 8:16:41 AM - Frank: 49 Homework is due Monday isn't it?
54 - 9/7/01 8:18:20 AM - barry: 50 Right. Another complication is newly fromed teams, some of whom prefer hammers, screwdrivers, and epoxy glue.
55 - 9/7/01 8:19:01 AM - barry: 53 Right. See my CS510 email for the latest revision of the schedule.
56 - 9/7/01 8:19:13 AM - A: We all understand critical role of education, technical skills, people skills etc. However in reality, when business is in trouble , educational benefits are frequently curtailed and frankly sometime they might be a huge drain on corporate resources. The other problem with education, is that its benefits usually are realized in long run, and that there has to be some assumption of employment stability in order for organization to benefit from the better skills of employees. Frequently, after getting trained by the company, employees - now with the better skills and therefore more marketable are leaving company, looking for a better, more paid jobs. That might make consistent committment to education difficult.
. - 9/7/01 8:20:56 AM - James K has left the room.
57 - 9/7/01 8:21:05 AM - barry: I need to go off to the 830 discussion session, so will log off now. Again, we'll webcast this so that remote students can benefit from the discussion.
58 - 9/7/01 8:21:22 AM - DougT: 54 I have heard of stability on teams prior to many consolidations, but at least what I've seen is people have lost their since of loyalty.
. - 9/7/01 8:21:23 AM - Frank has left the room.
59 - 9/7/01 8:21:43 AM - barry: 56 Right. Education needs to be coupled with pro-cactive retention initiatives.
60 - 9/7/01 8:21:49 AM - DougT: 57And companies
61 - 9/7/01 8:22:25 AM - barry: 58 One retention strategy is not to break up well-jelled teams. Again, DeMarco and Lister cover this well in Peopleware.
62 - 9/7/01 8:22:35 AM - barry: Bye
. - 9/7/01 8:22:40 AM - barry has left the room.
. - 9/7/01 8:23:08 AM - DougT has timed out.
. - 9/7/01 8:23:29 AM - alsawi has left the room.
. - 9/7/01 8:24:11 AM - somard has entered the room
63 - 9/7/01 8:25:57 AM - somard: Any one has any idea where to get the formula for the COCOTS that homework#2 requires ?
. - 9/7/01 8:27:29 AM - somard has left the room.
64 - 9/7/01 8:27:39 AM - Techsupport: the chat session has finished. plz see the live webcast of the discussion session through DEN website for further elaboration