Edited & Last Updated 7/08/2001 by jow

Project Director's Log

People, often those quite close to some aspect of our projects, often ask what it is I do beyond what they're involved with.  I do not know how well I'll succeed or how long I'll continue but I thought to try to provide a few notes here, both for my own edification as well as for the possible interest of others.

My precision may be a bit loose here.  I often work in a near-continuous mode over weekends wherein I may take longer breaks but remain single-mindedly focused on and working through some set of immediate problems.  Whether such work belongs grouped with the preceding week or following week, I lump it in with whatever's most related.  I'll try to continue for at least a few weeks.

Week Ending 2001-07-29:  Am largely in clean-up mode and prepping systems, software bundles etc for 4-wk vacation in Maine.  // Requested DMZ connection at Jackson Laboratory w/ low priority in exchange for consults, demos etc.  // Received request from MCEER for use & control of MCU for Friday tests and conference on Monday.  //  Listserv for Startup Venture defined and prelim config tested & exercised.  //  Install new Remedy client and trim email subscriptions.  //  Prep development system suite, ancillary equipment, and VC systems for transport to Maine.  //  Get self out-of-town!

Week Ending 2001-07-22:  Round-the-clock work on the two new development systems continued.  Problems with SCSI and 1394 support thought to be related to presence or absence of appropriate fixes and SP's in W2K turned out to be a result of PCI-slot assignment for the boards.  With few visible signs of the probable problem source, this bad judgment call cost over sixty hours of tests, research, downloading and rebuilds.  Now proceeding with build-up of development systems.  //  Will try to get R 3.2 IP/TV up on W2K systems before I leave for Maine.  Bob Dixon is starting to stir the Megaconference pot and it looks like activity will heat-up soon.  //  We did get an operating budget from UB this fiscal year but it did not clear until the end of the year.  Parts & components ordered are now starting to arrive.  Emphasis will be on Sandbox enhancements for greater remote controlled functionality, a few new test/development endpoints, software upgrades. miscellaneous hardware additions and upgrades, and the first steps towards MPEG-2 stream/VOD components.  There has been insufficient time to chase the money that folks say they would like to donate for our larger MPEG and Real streaming appliance needs.  I'll put up a Web page with planning/development priorities ATP.  //  Returned VCON loaner to Joes Stewart at ECC & brief catch-up.  Expect more 323 activity and upgrade of their MCU.  Follow-up re startup venture group.  Follow-up re residential service trials. //  Another Startup Venture group mtg - at Daemen.  Setup mtg with Mark Deuell for prelim presentation, introduction & advice. //  Review comment on Carol's talking points outline.  //  Mtg with Mark went fine.  I2 leakage issue may be more problematic than thought but probably not show-stopper if UB/Lisa support value of UB share.  //  over wknd, completely tore apart Sandbox and reconfigured/compressed.  New PWR switch added.  New W2K box added for Apache trials, access methods, etc.  UPS installed in primary Sandbox rack.  All primary components now on power switches.  Preliminary testing of new configuration complete.  //   

Week Ending 2001-07-15:  Picked up a few Apache books today - yikes, at $50 a pop!  IP/TV R3 is based on Apache for Content Manager.  Had been using the Sun JavaWebServer that came with IP/TV R2.  Checked into Apache for Windows (win32).  Looks good.  Lots of power and many books including O'Reily and, of course, there's a Linux version -- better, in fact.  // The early part of the week was consumed with getting the presentation for the PRI meeting ready.  It turned out well we all thought.  The meeting was great!  Good launch for an advocacy group.  Discussions with Dave and Carol before, during and after continue to prove the value of attention to detail and iterative collective  refinement, I think.  The plan that's evolving for the next few months is sounding more and more solid and appropriate.  Maybe a presentation on SA in September with a European contact or two tossed in, Megaconference in October and the Digital Summit in November.  Not to mention Grand Rounds, which we'll lose if we don't invest in.  Maybe it's time to back it off a bit anyway.  //  Finished up a few draft streaming video scenarios last weekend for the I2  VidMid WG and took part in the conference call.  Most of it is and will probably remain beyond me but I'll assume some mutual benefit until I sense otherwise.  If the cost is not too high, it will help keep me tuned to a strategic value sector: emerging directory services.  //  Spent an enormous amount of time on integrating the HP IDE CDR/CDRW units into 2x new Seattle development platforms -- the BCG/Veraview unit and my own.  There should be some longevity with two units and a good core MB.  Now building W2K base system.  Expect to follow-on with updated Win98SE to preserve good peripheral access and to keep-alive DCF and the CDROM test utility at least.  //  Believe that W2K SP2 takes ECDC-5 down the tubes and vice versa.  Formal problem report in the works but unsure worth pursuit and may abandon.  Everything looks good with SP1.  May have to stay there and add in 1394 supplements, if any.  //   Small progress on Linux system.  Plunging into kickstart.  //  Secured aliases for various IP assets to enable easier access and to nail an extra IP for the stream machines.  

Week Ending 2001-07-08:  Two older systems waiting attention got it; one new PS and one removable drive frame fan dead and to-be-removed.  //  Primary development system getting extremely old but also unreliable on CDR burns.  Vacuuming restored reliability.  //  Replacement primary development system integration started on Seattle platform; target is W2K with both SCSI and IDE CDR and/or CDRW.  //  Web site development still consumes enormous time as organization and available tools become clearer.  //  Setup JavaWebServer from IPTV on a stand-alone WinNT platform (old VCON base) and started to explore administrative capabilities and supplemental delivered applets.  //  Continued to work towards restored and improved CDR mastering in order to backup the new Dudek Linux build.  Integrated and tested an HP 9100 series CDRW drive I had picked up a pair of into the BCG Seattle development system.  Many hours later, settled on having both CD drives on one IDE channel and both removable HD frames on the other -- MUCH simpler.  //  Minor prep for and ops support for EGR test session.  Egon and I had a chance to chat a bit.  //  Fourth of July holiday.  Taken.  //  Completed draft scenarios for VidMid WG.  //  Started clean ground-up W2K build cycle for Seattle systems.  //  As the new primary development system starts to take shape, the entire system cluster support in the kitchen need to be redone.  The new KVM, that had been standing by, was installed and a new system physical arangement was implemented.  The old primary development system now functions fine through the new KVM.  //  pcAW locked up again on the CM/Web site system.  Reset at CC205.  Need to move to power strip control.  Host had not completed logoff/close of remote session.  //  Numerous Web site organization approaches in-test.  //  The new HP 9100 series IDE CDR drivers do not integrate into the OS in ways that Adaptec's new (now Roxio) Easy CD Creator expects -- both cannot coexist on the same system.  HP can be successfully un-installed.  //  It appears that there is no longer a way to force direct writes instead of indirect image-writes.  //  Limited tests of multi-volume image production on the 9100 drive using Easy CD Creator 5 look promising.  Time gained with the faster CDR drive, however, may be lost due to slower generalized but safer buffering practices.  //  Started to move gear from CC205 to home storage.  Started to setup for creation of second Sandbox rack.  Still working over configuration/location of components.  Encoders local minimizes necessity to touch out of office and provides more extensible model.  //  Starting to organize for a presentation on 7/11 of the local PRI WG.  Maybe offer the same material as draft talking points to Jacobs et al.  //  A full 3xCD reload from a fresh burned image burn-tested AOK.  (pending reload now)  System used in 9100 series tests was the BCG Summer system from last year.  The next step will be to migrate this onto the new W2K primary development system.  //  Refreshed driver downloads for Logitech and Wacom for W2K build.  Using built-in drivers for all else so far.  //  Downloaded, installed and started test brain with new PersonalBrain release.  It could turn out to be a modestly useful Web information presentation paradigm.  //  GOTCHA!  DiPro at Rev 3 will not reload image of  its own creation on an identical drive!  Could have handled partitioning poorly?  No clue.  Off by about 8 KB.  Attempting reload on oversized drive; if it works it buys time.  Oversized drive reload worked after removing all existing partitions.  //  Having problems with Linux partitions.  Need to get on top of the latest Partition Magic.  //  Pick-up DiPro R4 with data migration.  //  

Week Ending 2001-07-01:  Daemen/BISSNET immediate production needs group meeting at PremCom. //