CS267 February 1995 Announcements

Fri Feb 24 19:05:29 1995

We will meet in 405 Soda from 5-7pm, Monday 2/27 for the short talks about this last assignment.

Thu Feb 23 18:09:18 1995

There have been a few questions about the various tradeoffs that exist in this assignment. For example, depending on your implementation, there could be at least the following tradeoffs: time-space, time-accuracy, space-accuracy, and space-improbable_event_coverage. For this assignment, it is OK to go with the simpler case as long as you understand the nature of the tradeoff, and clearly document it in your writeup.

Tue Feb 21 13:14:09 1995

Once again, we're going to stagger the due dates. If the oldest person (we don't want to be "ageist") in your group begins with the letters A-I, the assignment is still due Thursday morning in class. If the last name of the oldest person in your group begins with the letters J-Q, the assignment is due Friday morning at 9:30AM in my mail box in Soda. Letters R-Z, due Saturday 9:30AM in my box. Please put the assignments in my box; last time, I received assignments in 3 places.

Several people have asked for the X-code for better particle display. The code is now in display.c and display.h. You'll have to hack it to get it to work with your data structures, but it should be straight-forward. BTW, the code works for both mono and color monitors.

Tue Feb 21 04:51:32 1995

Please read: Comments on Assignment 2 and 3: Suggestions for Assn 4

Tue Feb 14 17:33:43 1995

Bernd Pfrommer (pfrommer@tiger.Berkeley.EDU) is looking for a group for the current assignment. If you wouldn't mind an additional member, please send him mail.

Mon Feb 13 17:37:35 1995

Don't forget that you can mail split-c questions to split-c@boing.

Thu Feb 9 00:13:06 1995

The ts-daemon is up on rodin but doesn't stay up on moore. I'll be watching things for a few more hours. Prof. Demmel will be in at 6:00am, so there will only be a couple hour period when no-one can restart the ts-daemon.

Wed Feb 8 21:13:54 1995

Don't forget that you can get timings both from rodin.cs and moore.cs, the 64 processor 8MB/proc CM5. I've checked over the last few hours, and it seems like moore.cs is mostly unloaded.

Wed Feb 8 16:49:28 1995

The TS-daemon has been dying almost every 5 minutes due to the large load. Admittedly, this is a CM5 bug, but in order to give the machine a bit of a break, we're going to stagger the due dates. If the last name of the youngest person in your group begins with the letters A-I, the assignment is still due Thursday morning in class. If the last name of the youngest person in your group begins with the letters J-Q, the assignment is due Friday morning at 9:30AM in my mail box in Soda. Letters R-Z, due Saturday 9:30AM in my box. We apologize for not accurately anticipating the load on the CM5.

Wed Feb 8 14:03:44 1995

Class account disk space has been increased from 3MB to 18MB.

Mon Feb 6 17:59:48 1995

rodin2.cs now compiles cmf and cmmd properly.

There might be a bug or two in the problem 1 or 2 code that you should find and fix. This shouldn't be very hard, and it will ensure that you've understood the code properly.

Mon Feb 6 00:07:00 1995

At the moment, rodin2.cs is not set up properly, and compiling and running cm5 programs sometimes will not work. Until we have the problem solved, feel free to compile on rodin or moore.

Thu Feb 2 16:32:58 1995

Several people have mentioned that they can't run the split-c S&F code even on a color monitor, i.e., it returns the message "You must use a color or grey-scale monitor." The problem seems to be that the split-c X library tries to allocate too many colors in the color-map and reports an erroneous error message. Best thing to do? Temporarily kill netscape, a color-map hog.

For those of you on a monochrome monitor, you'll need to temporarily run the code on a color monitor. The main part of this assignment will be generating speed-up plots, and you won't need the graphics then anyway.

Thu Feb 2 00:32:51 1995

Eric Fraser should have mailed you information about CM5 accounts. If he did not, please send him mail and/or contact Gwen Lindsay on the 7th floor of Soda.

A roster of 267 is now online. Some of the entries are not complete. Please check your own entry, and if something is missing or incorrect, send me mail.