> Some claim that if AGPL is very toxic, and that even viewing AGPL code might taint any project you work on later.
Oh please, that's nothing but FUD. If this were true, then working on closed-source proprietary code for your employer would taint any project (open source or proprietary) that you work on later. Clearly that's not the case.
> It's not dissimilar to a certain argument of patents.
It's quite dissimilar. Patents and copyright are two very different legal regimes.
Back in the early 90s, of course the Unix wars were raging hotly. It was posited as a legal strategy by SCO attorneys, that if some developer for BSD had read Unix source code at any point, they would thereby become "Mentally Contaminated" and unable to perform clean-room development of the freely-licensed BSD systems.
So some of us thought this was a hilarious concept, and some joker printed up tons of mugs, tee shirts, and baseball caps which read MENTALLY CONTAMINATED, and they sold them at the Usenix conferences for a while. It was good for a laugh.
There were more factors than that. One was the burgeoning hegemony of the Wintel architecture, standardized with the PCI bus, cheap disks and fast interfaces, etc. The ability of BSD and Linux to leverage this hardware figured mightily into toppling the proprietary giants of the Unix world. Some of these vendors even realized it, such as Sun Microsystems, who introduced PCI-based systems late in the game.
I'd say it was a perfect storm in this regard. Sure, the lawsuit gave proprietary Unix a bad name. But that, in itself, wouldn't have been a death knell, if it weren't for these other practical factors.
Oh please, that's nothing but FUD. If this were true, then working on closed-source proprietary code for your employer would taint any project (open source or proprietary) that you work on later. Clearly that's not the case.
> It's not dissimilar to a certain argument of patents.
It's quite dissimilar. Patents and copyright are two very different legal regimes.