293 words
An interesting side effect of the current U.S.A. shutdown is that they actually make the effort to not just stop paying their staff but that they shut down their computer services as well.
Pretty much all of us have seen the shutdown notice you see these days when visiting, say, nasa.gov:
Let’s just hope they consider running stuff like GPS or not crashing the ISS as equally »essential« as spying on everybody over in the U.S. It may also be a good opportunity to take a look at the style of U.S. government sites which seem to have a visual language of their own.
The Library of Congress was offline as well (but seem to have a running website again now):
… a detail which seems to have plenty of fun consequences as they host numerous library standards and the XML Schemas of the formats belonging to them on their site. Really makes you think to which extent you can – or should – rely on online resources and how much effort it is to proof your software tools against failures like this one. Just imagine having software which relies on those schemas for some kind of validation and not being able to load them…
A curious further observation was the following: even somewhat obscure services like the NAL’s Z39.50 server stopped working (dutifully observed by nagios):
To me this suggests that either they just disconnected the complete networks of the respective government agencies or they made a real effort to shut down each and every one of them. It will be interesting to find out which of both … and in the latter case why they make this effort which could well be more expensive then just keeping everything turned on.
As noted by Jon Christian, “the determination of which services continue during an appropriations lapse is not affected by whether the costs of shutdown exceed the costs of maintaining services.”
Thanks for the info, which seems to set things straight from the bureaucratic POV at least.
The “shutdown” actually cost money, of course.
Uh, well, economic masterminds after all.
Interesting: The US government service my nagios has been monitoring recovered quickly.
I still wonder how exactly they turned off all these services so comprehensively (did they perhaps just re-wire DNS entries to some kind of FAIL server?) It also seems that the error messages they displayed were installed in one of the most harmful ways possible (302 redirecting to the error page instead of correctly 503ing and keeping people from crawling the wrong page).