From: Raymond Toy
Subject: Re: Selling lisp to the Next Generation Space Telescope Project
Date:
Message-ID: <4nwvabzsau.fsf@rtp.ericsson.se>
>>>>> "Joe" == Joe Marshall <···@content-integrity.com> writes:
Joe> ·······@stsci.edu (John M. Adams) writes:
>> Those currently responsible for software strategy believe that using
>> lisp on this project would be a mistake along the lines of `lisp is
>> not a mainstream technology and therefore there would be maintenance
>> nightmare because it is difficult to hire lisp programmers'.
Joe> I have to agree. When manufacturing commodity items like space
Joe> telescopes, it is important for the end user to be able to do simple
Joe> maintenance without having to find an `expert'. Given the current
Somehow space telescopes don't sound like a commodity item. A launch
costs millions (hundreds of millions?) of dollars, and if you screw
up, even more millions for a broken telescope. And what about the
ground crew watching over the satellite? None of this sounds like a
commodity to me.
Joe> available. This is especially true in cases such as this where the
Joe> proper functioning of the software is unimportant (pratically
Joe> irrelevant. After all, the photons are going to go into the telescope
Joe> regardless of whether the software is operating). Of course, there
You mean like the photons that went into the telescope for the
original Hubble with a "broken" mirror? Or maybe the software can't
point the telescope where I want it to go. Those photons are pretty
useful, eh? :-)
Joe> may be a bit more downtime, but given that an experienced software
Joe> engineer can cost upwards of $100 per hour, a few days of downtime
Joe> every month is a minor inconvenience.
Last I heard, time on the Hubble was a very hard thing to get.
More downtime would make it even worse.
Joe> It is obviously quite risky to use any `non-mainstream' technology on
Joe> a project like this. I'd stick to Estes engines for the launch
Joe> vehicle, use Loran for navigation, buy the mirror from Edmund
Joe> Scientific, and script up the code in Visual Basic. Such a system
Joe> would be trivial to maintain because of the rigid adherence to
Joe> `mainstream' technology. The bulk of the technology would be
Joe> accessible to many with a high-school diploma.
Are you serious? This telescope would be so good, there's probably no
point in launching it at all---a ground based telescope is probably
much better.
Ray