Welcome to Mike Grusin's

History repeating

Filed under: Engineering,Space — mgrusin at 8:44 am on Tuesday, May 26, 2009

Administration changes are always times of upheaval at NASA.  A panel, led by former Lockheed-Martin CEO Norm Augustine, is currently (5/09) reviewing the state of NASA for president Obama.

Mr. Augustine led a very similar review in 1990, and their report is interesting to read in light of the last 19 years of NASA history.  (It appears that many, if not all of their recommendations were followed.)  I particularly liked the conclusion, which puts risk and blame in perspective:

“We believe that the legacy our generation should leave to the future is that we pioneered the exploration of space, and thereby made important discoveries that will prove of benefit to all mankind.  However, space activity is inherently difficult — involving advanced technology and taking place over great distances.  It demands reliance upon machines, often very complex machines, which are designed, tested and operated by mortals.  It involves rewards which may be intangible.

As we labor under such challenges, we should insist upon excellence.  We should strive for perfection.  We should demand the utmost of those to whom we entrust our space endeavor.  But we should be prepared for the occasional failure.  If we as a nation are to place a greater premium on letting nothing go wrong, on not making errors, and on ridiculing those who strive but occasionally fail, than we place upon seeking potentially great accomplishments, then we have no business in space.”

Report of the Advisory Committee On the Future of the U.S. Space Program  (December 1990)

Time to get serious

Filed under: Engineering — mgrusin at 3:41 pm on Saturday, December 30, 2006

I’ve been more and more inclined to get away from Windows, for reasons I won’t go into here (that’s a post for another time). To this end, and spurred on by the need to replace my trusty but failing Toshiba laptop, I purchased my next laptop with Linux in mind.

I can’t totally chuck Windows; I own too much expensive software that requires it. But I would like to start using Linux for my day-to-day chores and try to migrate my larger tasks as I can. I was also seriously considering an Apple laptop (nice hardware and software), but eventually decided to vote for neither Windows nor OS-X and put my faith in the teeming (hopefully not ‘unwashed’) masses of open-source programmers. I always did like the underdog.

The machine: a Lenovo Thinkpad T60p. The good: almost everything, particularly the workstation-class graphics card, built-in DVD burner, and great screen. I would have liked to have gone with a smaller model, but those have been stripped down in the name of minimum size and power consumption, and I would have had to give up all of the above.

The bad: The keyboard feels great, which is nothing to sneeze at. But the layout is, in my opinion, nonoptimal. In my world, the Toshiba keyboard layout is about as perfect a thing as you’ll find, particularly the “home pgup pgdn end” column on the right side. The Thinkpad sadly doesn’t have that column, but what’s worse is that the delete key has been exiled to a small editing bank at the top right, way out of home-finger range. This is a key which I use nearly as much as the backspace, and Toshiba wisely put it next to the arrow quad where you can hit it in a millisecond. I knew the Thinkpad had this layout going in, and decided to adapt to it because the rest of the machine is so tight, but the lack of ergonomics is a little annoying. (I’ll have to look into remapping some of the keys…)

Also bad: the current state of desktop Linux. Nobody would like Linux to take off more than I, but frankly right now it’s a mess. I chose Ubuntu Linux, based on its reputation as an extremely user-friendly distribution. However I found out firsthand that the current version has many irritating problems such as broken laptop features like suspend/resume (some of which were working properly in older versions). Not to mention that the community documentation is highly fragmented, with lots of people advising lots of different ways to get it all working. I eventually patched things together enough to get most (not all) of the laptop features running, and the desktop with Beryl is flat-out gorgeous.  But I can’t imagine a “normal” person attempting this, nor can I imagine what will happen when they release the next version (do you upgrade on top of your patches and hope the patches don’t break the new software, or erase everything and start over with the new version, followed by another round of broken features and manual patches?)

I do appreciate the work the Ubuntu folks have put into making the desktop environment gorgeous and useful right out of the box – something sorely lacking from some other distributions I’ve used. But I may still chuck it and try loading Gentoo, which I’ve used before for server projects but never for a desktop. I do know that Gentoo has top-notch centralized documentation, and that the package system is extremely robust. I’m just not looking forward to the usual fiddling trying to get it looking and running well.  I’m getting too old for that nonsense.
The structure of Linux and Xorg still seem a bit fragile for mass consumption, but I think that’s a local architecture problem and not a symptom of open-source development. Architecturewise, I think Amiga and BeOS were on the right track (tight kernels and APIs based on the KISS principle); it’s too bad they were steamrolled by Microsoft and their own poor business decisions. But as I said, that’s another post.

I’ll write again as things progress. Happy 2007!

The Maker’s Bill of Rights

Filed under: Engineering — mgrusin at 3:43 pm on Sunday, November 26, 2006

From Make magazine (how many of these have you run into?)

  • Meaningful and specific parts lists shall be included.
  • Cases shall be easy to open.
  • Batteries should be replaceable.
  • Special tools are allowed only for darn good reasons.
  • Profiting by selling expensive special tools is wrong and not making special tools available is even worse.
  • Torx is OK; tamperproof is rarely OK.
  • Components, not entire sub-assemblies, shall be replaceable.
  • Consumables, like fuses and filters, shall be easy to access.
  • Circuit boards shall be commented.
  • Power from USB is good; power from proprietary power adapters is bad.
  • Standard connecters shall have pinouts defined.
  • If it snaps shut, it shall snap open.
  • Screws better than glues.
  • Docs and drivers shall have permalinks and shall reside for all perpetuity at archive.org.
  • Ease of repair shall be a design ideal, not an afterthought.
  • Metric or standard, not both.
  • Schematics shall be included.

http://www.makezine.com/04/ownyourown/