3 ms·
The costs are exaggerated. For example, it is not possible to replicably develop that software infrastructure for $50k. It is possible as a one-off if you're an
by blagie 5y ago
The costs are exaggerated. For example, it is not possible to replicably develop that software infrastructure for $50k. It is possible as a one-off if you're an engineer working on your own project, but for example, a biology journal couldn't drop $50k on that infrastructure. A lot of these things rely on individual heroics. I appreciate heroics, but they don't scale to an industry.
I'm glad for the good work they're doing.
- ta988 5y agoPublication and diffusion software can be Free there are a lot of options. Often university libraries even provide them with support for you.
- blagie 5y agoThey're free only if sysadmin time is free. It's the same rabbit hole. In practice, maintaining, supporting, and running a server, especially with archival-grade backups and retention -- is gonna require more than $50k (let alone $5k) of human time each year. You can't pretend there is no cost or that the cost is an order-of-magnitude lower than it is. At that point, you end up with solutions which can't work. Many people would gladly donate a megabuck a year if a significant portion of science went open as a result. People won't do that if you pretend the annual costs are $5k. Likewise, university libraries fielding that cost is a good solution -- and far cheaper than journals -- but it won't work if you pretend the cost figure is lower than it is.
- ta988 5y agoI said Free as in Free software. Of course there is a cost. But this is nothing compared to the tens of millions of dollars a state university has to pay each publisher each year.