Before its launch on 24 April 1990, officials and scientists from the US space agency, Nasa, and the European Space Agency (ESA) trumpeted Hubble as an instrument that would unlock the secrets of the universe. It would be able to examine the solar system in incredible detail and 'see' back in time to an era before the birth of planet Earth.
Although its 94in (2.4m) diameter main mirror was considerably smaller than many ground-based telescopes, Hubble would be able to take advantage of perfect viewing conditions above the turbulent atmosphere. Its five instruments would observe the heavens with 10 times better resolution and 50 times the light-collecting power of the best terrestrial telescopes.
There was stunned disbelief when the first tests showed that Hubble's vision was blurred. Subsequent investigations showed that the curved primary mirror was the wrong shape - it was too flat near the edge. This flaw had not been detected prior to launch because Nasa had tried to save money by not testing the compatibility of Hubble's primary and secondary mirrors. The error was minute, but the result was disastrous.
The most important effect was that starlight was spread in a fuzzy halo, instead of being focused into a sharp point. ESA's Hubble project scientist Peter Jakobsen said: 'It was like a child playing with a magnifying glass so that the lens became smudgy and images became hazy. Only 15 per cent of the light was useful, so there was a severe loss in sensitivity.'
The US trade journal, Aviation Week & Space Technology, called it 'one of the great technological bloopers of recent years'. No one was prepared to admit responsibility, but after a long legal wrangle, Hughes Aircraft Co (new owners of the division of Perkin Elmer that ground the primary mirror) agreed to pay Nasa a compensation settlement of dollars 25m for defective workmanship.
But, to the dismay of astronomers, Hubble was not only short-sighted, it suffered from the shakes. HST's power source, its pair of 39ft (12m) solar panels, was unable to cope with a temperature change of 200C as the observatory moved in and out of Earth's shadow. The 'wings' shuddered as they expanded and contracted during each 90-minute orbit.
Despite a steadily mounting list of equipment failure, however, HST controllers and scientists have been able to recover a significant part of its original capability.
'Hubble has been at least 50 per cent successful,' said Dr Jakobsen. Computer processing of Hubble's blurred images restores some detail and new software loaded into the onboard computer has succeeded in getting the telescope to damp down some of its vibrations. Even so, HST is in need of an overhaul.
The observatory was designed for in-orbit servicing every three years. Around the exterior of the telescope are 225ft (68m) of handrails and 31 footholds to aid the crews, and more than 80 tools have been designed for HST servicing.
Endeavour should lift off from Florida on 1 December. Three days later, the European astronaut Claude Nicollier will grapple the 43ft (13m) Hubble and gently deposit it in the cargo bay. Five sorties, each lasting six or seven hours, will be needed.
The main tasks are to cure Hubble's blurred vision and ensure it can remain in operation until the next visit in 1997. The optics in three of Hubble's main instruments will be improved by the installation of the Corrective Optics Space Telescope Axial Replacement (Costar). This phone-booth- sized adaptor will act as HST's 'spectacles' by re-focusing the light bounced off the main mirror.
Ten aspherical mirrors, each the size of a fingernail and polished to a precision of one millionth of a millimetre, will be deployed by moveable fingers so that they modify the light beams entering the instruments. In order to find room inside the telescope, Hubble's high-speed photometer, the least used of its instruments, will be replaced by the identically sized Costar.
A similar operation will be required to replace the even larger wide field/planetary camera (WF/PC) with a more advanced version. The wedge-shaped WF/PC I, weighing 619lb, will be retrieved through a servicing door in the side of Hubble, and the substitute slid into position.
Other high priority tasks involve replacement of the solar array drive electronics, the arrays themselves and some of the gyroscopes. ESA planned to replace the solar panels in 1996, but the operation has been brought forward.
Three of the six gyroscopes, which control the telescope's attitude, have ceased to operate - leaving Hubble without a spare. One more failure would halt science operations.
So what are the chances of success for this incredibly complex mission? 'It will be a minimal success if we get one of the new scientific instruments in and working correctly, and replace the gyros,' says Derek Eaton, HST project manager for ESA. 'But if we don't get most of these primary tasks done, there will be agitation for a flight next year.' No more servicing missions for Hubble are scheduled until 1996-97.
The first images from the revived observatory will not be available for several months but astronomers are certain the results will be well worth the wait. 'If all goes well, its quality will be 95 per cent of what it should have been,' says Dr Jakobsen. Hubble should be able to see more faint detail in nearby objects, and be able to probe quasars and galaxies on the edge of the universe.
However, he adds: 'It's really like ordering spectacles by mail order. We have built to a particular prescription, but there could be a mismatch.'
It is vital for Nasa that all goes without a hitch next week. After a series of debacles in recent years, the beleaguered agency's credibility and public support are on the line. This may be its final opportunity to show the world that it still has the right stuff to pull off the most complex and ambitious series of spacewalks in its 35-year history.