allistic Home
Page
7/23/99: Most of the funciton calls have been tested and we are reviewing and formatting the results. The remainder of the calls will be done this weekend or early part of next week. We are still having problems with the fname.tpl and NFS.
7/16/99: We have completed many of the non-POSIX call parameter templates for DRC. We hope to have them complete and the functions tested by the end of next week. We have received the new kernel from DRC. However, we are having problems with NFS and the fname.tpl file.
7/2/99: Still waiting for new kernel from DRC. We are working to finish up the integration of the Ballistic project back into the Ballista project. We should have everything complete by the end of the week, but we still need to test the integrated system on LINUX, Digital UNIX, and Solaris.
6/28/99: The kernel we expect from DRC has been delayed until sometime the week of 7/5 which pushes back some of our planned tasks. Integration is on schedule to be complete by early next week. We are meeting this week to reset the goals and plan for the end of this cycle.
6/18/99: We only have one bug left to fix before we can rerun some POXIX calls. We have set some milestone dates for the remainder of this cycle. Build 1 which will be used to rerun the problem POSIX tests will be done on June 22 and Build 2 which will contain the integrated version of the system will be done on July 2.
6/11/99: We have determined the root cause of the problems with our tests and are working with DRC and the Ballista team to reslove them. Integration of the Ballistic system back into Ballista continues.
6/4/99: We have found some issues with the way that we are testing some of the POSIX calls and are attempting to resolve them. Also, we are working to present our results to DRC next week when TR and TG will be on site.
5/28/99: We have finished testing all POSIX calls for VxWorks. We are now looking into the tasks that are required to integrate the Ballistic package back into Ballista. Also, we are attempting to identify all tasks requried before the semester ends on August 6.
5/21/99: We are almost finished testing the POSIX calls for VxWorks
and will easily finish by Monday. There are 3 functions
left to test, we will start them after the meeting. Keith has
already sent some preliminary results to Tim.
4/16/99: The client visit went well. We have some feedback on the user's manual and we will start working on fixing the manual. Cycle 2 has ended, and the Post Mortem will be held next week. Cycle 3 will begin the week after next.
4/10/99: The user's manual is almost ready for the visit next week. This will be a preliminary version with the user's manual section complete, but the technical reference incomplete. Most system bugs have been fixed, with only minor bugs left. System has undergone almost all unit tests, all integration tests, and all traceability/verification tests.
4/2/99: Unit testing had begun before Spring break, however testing had to be stopped when both machines started having problems after spring break. The VxWorks-Solaris serial connection sometimes cannot be read from/written to. Also, there appears to have been an unauthorized reboot of the Solaris on 4/1/99. The Solaris is up and running, but stronger security measures will take some time to implement. Testing has uncovered some defects unrelated to the machine problems, and debugging is being done.
3/5/99: The design document has been inspected and is being reworked. The coding is almost finished, with the VxWorks test manager completed, the harness setup module 60% done (expected to be finished Monday), the Marshaler code and the Generator being finished sometime this week. Code inspections for all modules are expected to be done this week. Test plan and unit test cases/scenarios are being developed. We are expecting to finish testing the system by Spring Break, and to finish system rework the week after Spring Break.
2/21/99: The design document has been sent out for team review. It can also be found in the Document directory online. The Studio inspection will be held March 1. Charles is working on the rebooting remote power controller device. Bob is working with Kobey Devale, and Tim Gakopoulos on the User's Manual.
2/13/99: The design document is scheduled to finish on Friday, February 19, and will be sent out to be reviewed. Studio Inspection will be around Friday Feb. 26. Figuring out how to safely reboot the target machine is still an issue. Bob Buchko is our new technical writer, and he will be emailing Tim Gakopoulos and Tim Ryan about what should go into the User's Manual.
1/29/99: The design document is being re-worked to reflect the changes we've made to increase portability/reusability. Our new technical writers will be helping with this. The tech writers will also start working on a User's Manual for running tests using the new system. The team was unhappy with its lack of a firm end date for the cycle, so work was postponed while re-planning occurred. We now have an accurate Gantt chart in MS Project, and this helped the planning. Cycle 2 will end on 3/22/99.
1/22/99: The SRS has been approved by the Studio QA, and is in its finalized form. The architecture document was Studio inspected on 1/21/99, and should be up for QA review in a week or two. Technical writers have been added to the team. The design is almost complete, and the design document needs some reworking by the team and the new writer. Coding has begun on the host test harness. The problem of rebooting the target safely is being looked into by Charles.
12/8/98: The architecture document draft has been completed, and needs to go through a Studio Inspection. The SRS has been inspected and needs to be reworked. The SRS should be ready for sign-off when we come back for the Spring semester. The Design Document is almost complete, and needs to be reviewed. Prototyping and literature research are being conducted to reduce technical risks, such as insuring test independence (making sure one test does not influence a subsequent test) and monitoring target processes from the host using WindShell. Project end dates are projected for Mid-February.
11/23/98: The design document is underway, with flowcharts and function interface specifications. Modular prototypes are being designed to explore implementation choices. A mid-cycle 2 planning session will be held to look at the end date for Cycle 3. Mesa/Vista transitions have not had many problems. We are starting to look at new operating systems to test for Cycle 5 and beyond.
11/12/98: The Architecture document is going through peer review, and should be completed by 11/16/98. The SRS is finished, and needs to go through Studio Inspection. This is scheduled for either before Thanksgiving, or immediately after. The move to a new project management tool, Mesa/Vista, is proceeding smoothly. Meeting minutes are being moved right now, and as the CM capability is installed, other documents will move to Mesa/Vista.
10/29/98: Cycle 1 will be completed at midnight tonight. The SRS draft 1 and SPMP draft 1 will be ready for inspection during cycle 2. The Architecture document is 80% complete. The Design Document is 30% complete. Cycle 1 postmortem and Cycle 2 launch will be on Sunday, 11/1/98.
10/18/98: The date for the completion of Cycle 1 and the start of Cycle 2 is Friday, October 30, 1998. We are confident all cycle 1 tasks will be completed by this date. The Statement of Work was sent out by Michael to Tim on 10/12/98. Short (30 min.) design meetings and other document meetings have been scheduled every week. The test harness code from Ballista is being analyzed.
10/9/98: We estimate our first development cycle to end at the end of October. We are currently estimating the third cycle will be done sometime in early January. After we finish cycle 1, we will have a more accurate estimate for the end of cycle 3. The VxWorks box is up and running, and Keith has developed a "Hello, World" tutorial for the team. The Statement of Work is in its final draft stage, and will be ready to deliver soon.
10/1/98: We have finished our Development Strategy and Plan for the semester. Some difficulties with the TSP spreadsheet tool are being worked out. All team members should be tracking the amounts of time they spend on their assigned tasks.
9/23/98: We have done some analysis of the current Ballista architecture. The results of our analysis are the Architecture Diagrams which have added hyperlinks to increase understanding of the different abstraction levels. The SPARC is up and running, and can be accessed from the network. Usernames and accounts will be set up shortly. Planning has been sketchy, since we do not have much previous size/effort data for code or documents.