I'll try to cover the then called ATI side of old common cards that might still be in use, with only 2.0 shader support.
http://en.wikipedia.org/wiki/Radeon_8000_Series (1.4)
http://en.wikipedia.org/wiki/Radeon_9000_Series (2.0)
http://en.wikipedia.org/wiki/Radeon_X800_Series (2.0b)
http://en.wikipedia.org/wiki/Radeon_X1000_Series (3.0)
(quick relevant info in the top right box with handy links to preceding and succeeding card series)
And some more condensed info from unofficial source:
http://forums.na.leagueoflegen ... d/showthread.php?t=143647
I did own a card from 1.4 line and some years later 3.0. The thing is the 1.4 still works, and i played LFS on it up until 5 years ago
. It was a GREAT commonly used card, with excellent cost/performance, but I ran win98 on that PC - card is Radeon 8500. In terms of performance I could run first Far Cry and Morrowind on that. I am not sure if I could run win XP though and the card only had 64 MB of VRAM. Probably safe to assume the 1.4 line no one with gaming in mind uses today, maybe only as an emulator machine with the card being so endurable.
I haven't owned a card from the 2.0 lines but I remember in those times new released cards having extremely high cost/performance ratio (high cost, average performance), which only financially backed enthusiasts would be willing to afford right after release. Only after the next series would be released the price of the previous one would drop considerably, becoming affordable to wider rage of people. That means the release date of the last series to support 2.0 (2004-2005 for the X800) could be more accurately represented by adding a year or two (and more if you consider the classifieds).
Looks impossible to estimate the number of LFS players using the 2.0 line but they undoubtedly exist if the 2.0 durability is anything like 1.4.
You could try and gather card usage info by including the card from the unofficial link in some kind of poll and try to estimate by the number of hits and time it took. More precise possibility would be to release a temporary 3.0 official test patch via the auto-updater and examine the number of complaints of LFS not working in the first day or more. This test patch would include only the 3.0 change if possible to minimize false errors. Preceding that would be some kind of announcement so at least some people would have an idea what to expect. And have a revertible patch ready in case the 3.0 doesn't satisfy! But you know how people can be... you give them something, then take it away and... yes
.
Tough shoes to be in. Decision seems easy if no one uses 2.0 in LFS. But that would we gain with 3.0 at this point? More FPS because of increased efficiency? Is it worth the risk of breaking the game for some people? Would it not be better to wait until 3.0 features come to LFS to put them in? I don't know, just asking rhetorical questions
Is anyone familiar with how did
World of Warcraft gathered user card info and made the transition? I reckon they started with 2.0 model and are now on newer verion.