Roberto Bagnara bagnara@cs.unipr.it wrote: you are also missing a few things. I think you have misunderstood the direction of my argument. My point is that *ANY* set of *fixed* locations (as opposed to *default* locations, which is what you get with the usual 'configure' setup) is a bad idea, and that changing to another fixed location is not an improvement. If a change is to be made, let it be in a more generally useful way. For example, let the installation be relative to the current directory.
The third thing is that my proposal is simply to improve portability by matching what Jan's makefile already does on all Unix/GNU platforms: No, the standard setup for SWI on UNIX platforms installs quite happily wherever I tell it to. It has *default* locations, not *fixed* locations.
I really wonder how can people in CS departments accept impositions such as the one of not being allowed to know the root password of their own machines.
As long as I don't want them connected to the University's network, I can do what I like with them. But if I want them connected to the University's network, I have to abide by the terms and conditions. As for the University's machines (which are bigger and faster than my old machines), they are the University's property and they have the right to set any reasonable conditions they want to.
I'm old enough to remember the old co-operative days of the ARPAnet, when you automatically trusted the people at the other end of the wire. I'm also very glad NOT to be running a Windows box; from memory this University receives about 3 times as many bytes of Windows viruses as it does of wanted e-mail. It's not that long since one of our labs was shut down for the best part of a day by students plugging a laptop they "knew" was clean into the network. And it was only yesterday that the University network was shut down by what I believe was an e-attack. If I have to wait a while for things to be done because the sysadmins are busy keeping anti-virus software and firewalls up to date and installing all the latest security patches, that's a price I'm willing to pay.
Now if you wanted to start an argument about Sun's FLEXlm-governed C compiler, which costs me serious lost time whenever there's a network outage, and FLEXlm's underlying philosophy (we'll assume you are a lying thieving weasel unless and until you prove otherwise), .... (:-).