An eReader can serve the role of a general purpose computing device, but it would not serve that role as well as a laptop.
So if you optimize a rugged poly-carbonate brick you could take through a Monsoon, Typhoon, Sandstorm and/or War Zone for the sole purpose of loading text and simple document files and a low-power display, stuffed with only a battery, enough compute power to accomplish the task it is given and cover any overhead, and enough storage to hold however many books you decide you want it to hold, without any graphics or sound (or perhaps very simple graphics), then you still have something meaningful.
You have a Library of Alexandria that any kid can carry in their arms without any of the ideological attachments that the XO had. It holds information in a human readable format and is capable of displaying that information to the person who holds it so long as they are literate.
Much like a modern day light bulb can be a general-purpose computing device that for some reason has been limited to turning on and off in different colors. It might not seem terribly useful to you, but it thanklessly serves the role it has been tasked with without a source code button, a mesh network, a fancy GUI, or a Squeak environment.
It is a meaningful distinction to make because it will ultimately shape your budget and your ability to actually distribute devices in meaningful quantities to the device's intended users.
I'm all for designing things right, but I think it's not important to get caught up in the naming.
Display. Storage. Battery. Some way to charge it when you're nowhere near reliable electricity. Maybe a keyboard. Maybe the ability to communicate with other ones in a mesh. Maybe they can hook up to the internet if it's available.
And then lots and lots of great content pre-loaded.
For instance, ka-lite, the downloadable Khan Academy:
"The 4781 videos that are available currently have a size of 57.1 Gigabyte."
That's really not that big, any more.
I'd find that useful. I'm sure people/kids in developing countries would too.
An ereader already contains all of the hardware to be an interactive computer (especially if it has a keyboard, like the kindle 1, 2 and 3). It really shouldn't cost any more to deploy an Alan Kay-style dynabook than to deploy ereader appliances.
> the idea of a general purpose computer you could also teach kids to code on seemed really really cool
The idea wasn't that you could teach programming with the aid of the computers. The idea was that you could teach everything with the aid of computers.
[0] https://www.forbes.com/sites/kellyclay/2012/10/12/amazon-con...
[1] https://www.npr.org/2012/09/06/160697501/new-amazon-kindle-w...
Strangely enough the first kindle e-reader was released the same month as the XO laptop:
Cheap 10" tablets are definitely easier to make than cheap 10" E-readers, though.
Interactive learning is critical, and I'm some cases superior to passive learning. An interactive device can become passive, but not so much the contrary.
> but until we know how to properly introduce computers into the classroom as a general purpose educational device
1. We can only know how they can be used in the third world by giving them to the third world. Any lessons learned with first world children are not obviously transferrable. This is a big problem with a lot of research in psych.
2. You're assuming the third world kids getting these, or who would most benefit, are already learning in a classroom. I see no reason to accept that assumption. Seems to me, the ones who benefit even more are those who have little to no access to consistent education, so your objection does not apply.
Granted eReaders were a niche at the time of OLPC, we didn't fully grok where laptops fell on the useful vs distracting scale and the idea of a general purpose computer you could also teach kids to code on seemed really really cool and probably the best way to on-board them onto the internet.
All in all, OLPC's heart was in the right place, but until we know how to properly introduce computers into the classroom as a general purpose educational device, something more like a rugged eReader and open source textbooks feels like it would be more productive in accomplishing at least some of the goals of OLPC.