Re: "At long last, Sir, have you no shame?"
- To: mathgroup at smc.vnet.net
- Subject: [mg18385] Re: "At long last, Sir, have you no shame?"
- From: Martin Kraus <Martin.Kraus at informatik.uni-stuttgart.de>
- Date: Wed, 30 Jun 1999 14:13:39 -0400
- Organization: University of Stuttgart
- References: <199906150543.BAA13337@smc.vnet.net.> <7ku02a$996@smc.vnet.net> <7l1d2g$dv7@smc.vnet.net>
- Sender: owner-wri-mathgroup at wolfram.com
John A. Sidles wrote: > > Andrew Watson <abwatson at mail.arc.nasa.gov> wrote: > > > What gives? Is this really something that has stumped the > > greatest computational minds of the end of the millenium? Is > > it really so low on the list of priorities that it has > > persisted for 10 years? Have you no shame? > > Things are slowly getting better at Wolfram, but progress > is slow. For example, did the Wolfram folks ever get > around to fixing/documenting in Version 4.0 another notorious > Mathematica behavior, that Eigensystem sporadically (about > 1 time in 10^3) returns complex eigenvectors, given a real, > symmetric input matrix? I received email from several > other physicists who had been bitten by this undocumented > behavior --- it afflicts several platforms. But all we got > from tech support were the same "sheepish chuckles" that > Andrew experienced. Net result, no documentation. This of > course guarantees that future users will continue to use > Eigensystem[] in a manner which sporadically & silently > breaks. This wasn't Tech Support's fault, the problem > originated in policies set higher-up at Wolfram Research. > > Seriously, what's going on at Wolfram Research which blocks > better on-line documentation of this kind of stuff? Hmmm, I guess we could quickly gather dozens of well known bugs in this thread. Before doing so: Is there anyone who would like to set up a database with online access for bugs (including workarounds) in Mathematica? Perhaps the existence of such a database would force WRI to include it into their support section in order to get it under their control. Not the best motivation for a better bug documentation by WRI but at least it might work. Greetings Martin Kraus