Weekly planning Meeting, 6/20

[Marc Stiegler <marcs>]

Highlights:

Big congratulations this week go several people. Hill has enhanced 
the preAlpha backend so that data is stored to disk; as a consequence, 
it is much more useful for our frontend developers, since they 
can save test data. Hugh successfully hooked the Mac frontend 
with the Sun backend; he can create documents in the server from 
the application. And Roger has hooked up the Sun frontend to 
run to remote backends as well.

Congratulations also go to markm and dean for simplifying the 
ent yet again. Markm is delighted because we have more guarantees 
of log-bound performance under more extreme conditions. However, 
this is NOT what the congratulations are for: the congratulations 
are for shortening the time needed to implement. Dean tells me 
that version compare in the forward direction is just 4 lines 
of code.

Meetings: Markm, talk with robin about backfollow of endorsed 
objects of specified type. Markm, roland need to discuss xxref++ 
with respect to blasts. Hugh, roger need to discuss common code. 
Roland, michael need to discuss fester crashes. Dean, roger, 
hugh, do we need discussions of document titles?


Anyway, this week's task list is as follows:

A.     dean

x    6/20/89     presented the ent
x    6/20/89     simplified ent design again
x    6/20/89     figure out berts and sensors in ent
     6/27/89     break new ent into PERT pieces
     7/03/89     complete enhanced c++ translator

B.      hill

x    6/20/89     stubblefy bogus data structures to put onto disk
     6/27/89     episode scavenging
     7/03/89     session scavenging
     7/05/89     put blast shields everywhere
     7/10/89     session-persistent scavenging

C.       hugh

x    6/20/89     complete macapp mods
x    6/20/89     FE & BE talking
x    6/20/89     create documents in BE from FE
     6/21/89     air traffic control
     6/27/89     textedit remote to backend
     7/03/89     flat document viewer

D.       markm

x    6/13/89     solved concurrency in design
x    6/20/89     xrefcounts work
     6/27/89     adequate berts&orgls for doc&links
     6/27/89     spec document layer
     7/05/89     doc & links implemented
     7/12/89     polish off the basic semantics
     7/12/89     help marcs with febe documentation

E.       marcs

x    6/20/89     put blast shields on PERT
x    6/20/89     get a gleeful estimate of concurrency in tuning phase
x    6/20/89     hold PERT meeting, rearrange tasks
     6/27/89     finish June/July PERT chart
     6/27/89     capture semantics video in ascii
     7/14/89     rough draft of AMIX/Xanadu frontend comparison
     7/14/89     second round of patenting work
     7/28/89     rough draft AMIX/Xanadu integration plan
     7/14/89     racetrack

F.       michael

x    6/20/89     put bogus urdi in loclshare
     7/03/89     URDI complete for sun 4

G.       roger

     6/20/89     get FE working across net
     6/27/89     class definitions for textedit
     7/03/89     textedit working
     7/03/89     outline common portions of code with hugh
     7/03/89     implement draft document layer
     7/10/89     link editing
     7/10/89     comparison working with bogus
     7/17/89     walker FE in C++, version 2, with bogus

H.       roland

x    6/20/89     semiformal spec xlint++
     6/27/89     xlint operational
     7/03/89     automatic documenter
     7/10/89     concurrency/exceptions in stubble

I.       Bill

     6/27/89     backup system operational

J.       Lars

x    6/13/89     go to AEC expo
     6/15/89     go to europe for 4 weeks

K.       Joel

x    6/20/89     stand in for ted in hyperexpo
     7/01/89     draft mkt plan

L.       Unassigned

     Tumblers as subclass of partiallyOrdered
     pc frontend
     tutorial for users on the frontend
     reference manual for users on the frontend
     make sure memory reclaimed properly
     test scaffolding, code exercisers
     rollin, rollout a la Univac
     subaccounts, permissions on FE
     sys admin, i.e., archiving and partitioning disks
     do we need object-pascal stubble?
     programmer's example
     encryption stuff (postBeta, maybe post release)
          study stanford key distribution system, licensing
          find single key system we trust
          find encryption hash we trust
          find authentication we trust

--marcs