Is it really that hard to switch to [google|libre|apache|free|etc.|etc.]? It seems like at the university level the ideas are the important part, and the need to write/spreadsheet at the bleeding edge of functionality much less so?
Short answer: to Google it's not so bad but it's not like the legal risks are any different from Microsoft. And to the rest -- yes it is very hard.
Universities need cloud storage with online collaboration and a fully functioning office suite.
LibreOffice doesn't work because it's desktop-only and has no collaboration. However, there's an online-collaboration fork called Collabora Online, and you can use something like Nextcloud to provide your own privately hosted cloud backend. But obviously this is a gigantic effort for the university's IT department to provide and maintain with reliable redundancies and backups.
Also, LibreOffice/Collabora is pretty good if you stick to its native formats, but its interoperability with MS Office files has a lot of bugs.
In the end, it's just cheaper and more reliable to use MS or Google like everyone else. Students, professors and administrators wind up having basically the same needs around office software as businesses do.
At 4 european universities i studied/taught this has never been the case. Most universities are used to run their infra, they ran their email servers way before google existed and they run big fleets of servers for thin clients. Afaik they still kept their own internal messaging as backup but it was still email servers hidden behind web gui.
What happened was that the big tech came in and made everything for them free. It is really hard to compete with free. They get windows for free, they get gmail for free at some point even unlimited google drive for free.
Now the situation is changing as the corps are tightening. I've seen 40k student university switch from gmail to office360 in two months because google suddenly wanted money and microsoft didn't. Now Microsoft also wants money. And it's not small money. So the school is doing cost assesment - you can give it to european third party provider that will be way cheaper tham microsoft. Or you go back to your own infra.
Turns out that what to be really expensive when google was giving people 30gb of free space to everyone in 2012 now is actually not that bad and you own your future. My guess is they will pay Microsoft for a year while they transition their email to their infra. The other parts gonna come later. But the students are required to use libre office (or latex) for writing their thesis so i don't think they see google docs as big blocker.
There's a huge difference between running an email server and some additional servers for thin clients -- all traditional stuff -- versus running an entire private cloud that redundantly stores the many many petabytes for your 40,000-person university, and all the web servers for the office software. Keeping it secure, keeping it updated, and having a live failover site if there's a fire or flood in your main data center that takes it out for weeks or months.
If it were that easy and cost-effective to do, large corporations would be doing it too. But there's a reason they're not.
Unis are more suited to do this than most corporations. For example at school i am at
1. They already self host many apps in production. Including their own complex homegrown app that manages the school. All of the students have 30 gigs of “network drive” for their virtual computer. That means lot of the infra is already there - including unified oauth/ldap.
2. There are already experts there teaching programming/devops. So these people can both administer and teach. I wouldn't underestimate inhouse capabilities.
3. It is quite easy to get grants for infrastructure modernization. These funds wouldn't be possible to use on paying third party services.
Overall I think its simply matter of costs and once googles/microsoft is not free they might not be the cheapest option anymore.
How much is the typical dutch university paying MS/Google? Maybe 10k students x 200EUR/year = 2 million EUR/year.
Twenty universities come together to move to make Collabora+NextCloud work for them. That's 40 million EUR/year. How much do they need to actually spend on developers + infrastructure to make it happen?
If you look at the numbers that way, open source usually looks like a slam dunk.
The problem is coordination issues: actually getting people and orgs to look at it that way and spend the money that way, rather than just waiting for someone else to fix the problem.
