Hacker News Reader

Comment by sagischwarz | Hacker News Reader

I think the public educational IT system should not be outsourced to the private sector, where schools completely lose control of the data and depend more or less on the good will of the company. University courses should not be organized by Facebook and pupils should not be forced to use IT systems of non-trustworthy companies like Google, which earn money by profiling their customers. Maybe my view is too German, but I think of education as an sovereign function of the state, where everything (theoretically) can be controlled by the citizens (yes, I know that this is some kind of optimistic). We should teach kids to think critical about centralization of data and knowledge and show them how to manage their digital lives with free and open tools which respect the users rights.

View on HN

I work in IT at a major Canadian University. I hear what you are saying, but the solutions provided by universities IT departments are usually sub-par to completely not useful. There have been attempts in the past to partner with other schools and develop solutions jointly but are often cost over run disasters.

It is really hard to attract any top development talent, or even try and convince current employees to stay up to date with current trends. Lots of development teams are stuck in old waterfall mentalities on huge java stacks.

We have laws that protects students data and doesn't allow us to host student or even employee information on U.S. servers. Provincial laws and mandates handed down from deans because of the Homeland Security Act hinder the use of any sort of SaaS's. It makes it really hard and disheartening to provide and develop good solutions.

I can't suggest someone go use certain survey tools, or use mailchimp to handle a small campaign for example. I am not supposed to promote the use of Google Drive or Dropbox for day to day use with other staff because of fear they will start storing data they aren't supposed to on it. We often end up with sub-par self hosted copies of great applications, but no one wants to use them. Google classroom looks really cool but there is no way any schools in this province would be allowed to use it.

> We have laws that protects students data and doesn't allow us to host student or even employee information on U.S. servers. Provincial laws and mandates handed down from deans because of the Homeland Security Act hinder the use of any sort of SaaS's. Well there is openstack which you can roll your own IaaS after a lot of serious effort. Where I studied in the past would use VMWare heavily to deploy hundreds of OS install. The only reason (people I spoke to since I was sort of involved in that) they couldn't do it was they did not have the incentive to do it. Learning a new technology (rolling OpenStack is not that simple, actually) is very very time-consuming and without the welcome from the top, most people won't even bother to try different technology.

Google Classroom runs on a huge Java stack.

I work at a medium US university and I can say that the thick java stacks and waterfall is a huge part of the sources of headaches in my job.

I've got my team quite agile for a Uni development team but we still face lots and lots of problems with FERPA(US student privacy rules), and trying to work with much less agile internal teams. Most of the rest of the local developers are 40-60 and very resistant to any kind of change. (seriously using .net instead of java was hugely controversial)

Could you or the parent express what you dislike about backends written in Java? I think there tends to be a pretty negative stigma associated with it, but I haven't really seen why. Personally, I've worked with both C# .Net and Java backends, and it's really a horse a piece. Is performance the issue? Security? What is the general dislike of Java about?

I was just saying it off the cuff as sort of a comment on giant waterfall departments that never create anything great or effective. I don't really have any hate for Java and there is a lot of great software created with it (J2EE was my intro to IS development), just the kind of culture and staff it attracts in this sector. Everyone is in their cubical doing java never innovating anything with no care in the world to ever break out of the status-quo. When they do hire new staff it is usually someone who has been doing that Java at banks or for the government, for 5-8 years and it just continues. Then these people move up into the more senior positions and help cement the culture. Being on more of the web-side it is just frustrating.

Yeah, you hit the nail on the head with this explanation though.

You both raise fair points. I do think the community is typically construed that way, and it's probably accurate to say that a lot of the older Waterfall devs would be a part of that community (or C/C++).

In my experience .Net has a very similar set of problems and cultural issues, so if those are the two you're comparing, I'm not surprised you don't see much difference...

Depends on the parts of the .net community a lot. There's some really nice parts of the Java/JRE community too so it's not like it a function of the language, its a function of the organizations choosing the languages.

Java is a pretty nice language, it's the Java old-school waterfall java developers that bother me. Our IT development unit is a slow dinosaur waiting to be eaten up but no native competitors.

I've waited months for a bug to be fixed because it "wasn't in the schedule". Theres a kind of "sealed lab" mentality where they hold a bunch of meetings to get stakeholder views, and then disappear for 3-4 months and come back with some 1/2 working software that I could have(and a few times have) knocked out as an alpha in 3-4 weeks.

It's the worst aspects of people being "enterprisey" with people who have no performance or profit pressure so they're able to put out crap and if it's better than nothing the admins accept it.

I think when people pick a bone with Java, they are really talking about the community, not the language. To generalize, the Java community is seen as enterprise-y. They often like process and unnecessary complexity.

> seriously using .net instead of java was hugely controversial

Was this for a server app or the client side? Because for a server app, I can't imagine any circumstance where this would make sense, since the two are so similar anyway.

At my undergrad, the internal IT department was generally folks who couldn't get full time jobs in technology. Many departmental projects were done by students as projects, and were never updated or supported afterward.

Maybe this is asking too much from Google but maybe they could provide some accounts which only back up data to Canadian servers?

You really don't need a lot of horsepower to set up a reasonable cloud solution. In fact these days it is so simple that you can do it in an afternoon with the help of tools like https://yunohost.org/.

Many Universities already use stuff like Blackboard[1] which for the most part is TERRIBLE (at least how I remember it in school.) These pieces of software are already around, and if anything competition in the space will hopefully make things better.

[1] http://en.wikipedia.org/wiki/Blackboard_Inc.

But as far as I know, Blackboard (or Moodle or Sakai or...) data is stored locally at your institution, not housed on a Google server somewhere. I think the data mining possibilities are much lower in that case.

I'm all in favor of disruptive competition in that market, mind you! But I don't think it's easy.

Not true. Blackboard has a cloud offering. Plus, many competing products also rely on centralized servers.

What little information I can find, indicates that Blackboard is self-hostable, and that the "hosted" option they provide is actually an extra service.

The argument is against a solution that has no possibility for self-hosting and maintaining control of data. That they offer a hosted/SaaS option is irrelevant, because they also offer a regular self-hosted solution.

My University uses Moodle, which is also TERRIBLE. I do all my grades and such in Google Docs, since Moodle is a UX nightmare. Something as simple as entering grades in Moodle requires a lot of clicking around for each grade. We have multi-page documents describing how to perform even the most basic tasks, they require constantly clicking and going to new pages of indiscernible icons. Menus are impossible to navigate without help.

Tragically, the larger the organization, the more likely computers are to increase manual labor and repeated data entry, rather than reduce it. In the days before automation, you'd have simply written a letter on a piece of paper, in a second or so per student. I am not surprised to hear that "designed by committee" results in replacing something fast and easy with a long complicated process.

> I do all my grades and such in Google Docs

You upload all your students grade information to an external private company? Which university do you work for where this is allowed?

It's probably not allowed, but I think apps for education actually includes a promise from google to obey FERPA.

http://www.google.com/enterprise/apps/education/

It is totally allowed BY having a separate ID system. If no one can ID the student then no one would gain the information.

Moodle is such a terrible thing to happen to the education ecosystem.

It's free/open sourced, meaning that every education system with a budget finds it appealing. Once implemented though, the organization will need to spend buckets of money to design, roll out, and maintain Moodle.

What you end up with is organizations installing Moodle, rather than some higher quality, equal-cost education software they should have installed from the beginning.

My university used a combination of Blackboard and Moodle, both of which are terrible. Your memory serves you correctly.

Non-trustworthy seems a fairly harsh assessment. I've not seen evidence to ever make me think Google isn't trustworthy. I may not be happy with what they do with my data, but it's not like they've somehow broken their promise to me in any way, we have an understanding. I get use of their services, they get information about me.

Why the distinction is important is because if they stated publicly that they weren't gathering a specific type of data (and I felt they weren't legally compelled to lie by a government), I would trust that statement.

But you are talking about your personal social contract with Google, and looking over this perspective you are right; but in the end you can choose; The problem is when we take that to education, its a collective social contract, intermediated by a third-party, not a personal student choice.

The OP is right to be concerned; once we are ok with intermediation of the social contract; we are no longer in control; we loose power, and even of we dont agree or dont trust the provider of goods or services; we cant do anything to stop them, and thats whats wrong.

Its a small detail, but a very important one!

Dont accept intermediated social contracts anymore, in digital age, we can all have our own private, self-determined social contract, and that can be enforced even on our democracies and fundamental rights

You seem to implying some position of mine towards the larger discussion based on my statement, which wasn't my intent. Whether I trust Google to do what they say and whether I think they are appropriate for the task being discussed, are not necessarily the same thing. There are other factors of importance, as you pointed out.

Do you trust Google to remain just an adware company forever?

Or will they (a) sit on your children's data for all of their life and (b) within a decade or two break out of the adware niche with Google Services for Health and Life Insurance, Google Services for Banking and Credit, Google Services for Human Ressources, and Google Services for Law Enforcement?

Not a lawyer, but the Google privacy policy does seem very nice to the student here: http://www.google.com/policies/privacy/#nosharing

This is to contrast it with the privacy policy of other systems like Canvas: http://www.canvaslms.com/policies/privacy-policy

Selected out of context quote from the Terms of Use: "You warrant that the holder of any worldwide intellectual property right, including MORAL rights, in Your Content, has completely and effectively waived all such rights and validly and irrevocably granted to you the right to grant the license stated above." Caps to add emphasis.

That sounds like a pretty standard and necessary clause; it's not assigning "moral" right to Google, it's asserting that you have the right to be controlling the content's licensing in the first place.

It wasn't to Google, it was to another company: Canvas. Also, what the hell is a moral right anyway? How is it different than any other? It just seems really scammy to me.

Look it up and post back. It's not a weird scam, it is standard copyright law regarding who can claim authorship of content.

The section you quoted just means that the user is promising not to post pirated content to Canvas

> Also, what the hell is a moral right anyway?

In the context at issue (copyright law): http://en.wikipedia.org/wiki/Moral_rights

If I remember correctly, Google does not use its classroom clients data for profiling purposes. Even in that case, I think that education should not be locked down to a particular vendor and that it would be a very good use-case for open source solutions. I fear that it is not realistic in the near future though.

Cannot agree more with this. Open source software is a perfect fit with public institutions. I don't understand why any government uses proprietary platforms. It would be cheaper and far more effective to hire a dozen open source devs to cater a particular platform to the public's needs, and have it be open and free to use forever. Developers aren't that expensive compared to software licenses for thousands of machines.

As I understand, Germany and many European countries are heading in this direction.

While I agree that Open Source can form part of a good solution I think there needs to be a distinction made between proprietary software that the institution controls, and proprietary SaaS solutions like Google offers.

E.g., In the 90s and early 2000's Novell had a big presence in Universities/etc, providing network systems such as eDirectory (their Directory Services stack) Groupwise (email) etc. Yes these were proprietary, but they were also controlled by the customer - if you wanted to switch to OpenLDAP or AD or whatever, you could dump your whole Directory Services tree to LDIF file(s) and import them elsewhere. If you wanted to migrate from Groupwise to something using MBOX or even to Exchange, there were tools to do so, and there was never any need to worry about "well what happens with the data?"

I have no problem with a school or government using proprietary software, so long as the data created remains in the institution's hands, and is in a format that can be easily read by or exported to alternative formats.

I absolutely have a problem when educational institutions say "yeah we're going to use this solution for email because it's free" - the students have no fucking choice, and there is little or no consideration given to how the company uses the data.

I've been working on KnightOS [1], which is like Unix for calculators. I'm hoping to use it as a platform upon which to evangelize open source software with students. Maybe we should have more open-source investment in more parts of the student life? You just have to pick something to improve and get started.

[1] http://www.knightos.org

Coming from recent experience in EdTech, and also as a recent grad, this is a little idealistic.

> It would be cheaper and far more effective to hire a dozen open source devs to cater a particular platform to the public's needs, and have it be open and free to use forever.

Unfortunately, not many people in Higher Ed are technologists. Even hiring devs to work on open source tools isn't a good solution since most are inept at managing technology. The result of this kind of thinking is Moodle and Sakai, LMS's that have faded and failed. They're terrible to use and incredibly outdated. Horses, Camels, Committees, and all that.

On the other hand, Instructure makes a LMS called Canvas, which is open-sourced under a restrictive license. It's a Rails app on github. The tech is as much of a joy to use as anything in EdTech. They also have an impressive salesforce and have grown remarkably fast. They're due for an IPO soon.

Oh man - government agencies, in many cases, the least equipped to "hire a dozen open source devs".

Think about how many local governments and school districts they are - they couldn't possibly hire developers and manage the development and customization of open source platforms when they can barely keep their own IT running correctly.

In fact, that's a huge driver for the adoption of SaaS in local government - IT is such a scare resource. School districts often have only a handful of IT people for thousands or tens of thousands of students - they certainly couldn't afford to hire developers to customize, nor do they have the skills to manage them.

I understand the open-vs-closed concerns, and agree that open source is better, but SaaS is a godsend for local government.

Experience from Denmark: when public institutions make it-systems they almost always fail.

The "exam submission system" for primary schools failed completely the first two times when the pupils was sitting the exam.

The police completely cancelled their new it-system after spending $90 million.

etc.

So when the (Danish) public sector makes stuff that's worse than BLACKBOARD, it might for the time being be the least of the evils to outsource it to the private sector.

I agree too but there are non-technical problems. Educational institutions regularly spend money on software when some CS students could do it better and cheaper over a summer. The people making these decisions are non-technical and risk averse.

This times one billion.

> when some CS students could do it better and cheaper over a summer

And then move on, leaving an unmaintained chunk of code that, while possibly better than the third-party software, instantly becomes a liability of unmaintained code. There's a reason why contractors have a bad reputation in the industry for the code they deliver. It's only developers that know they'll have to live with their work for the foreseeable future that put thought and effort into maintainability.

The only way that your "CS students over the summer" plan could work is if the University put a ton of effort into bootstrapping an open source community. Even then, it's doubtful they could create something that would be appealing enough to work on. Academia has a contractor-like reputation for producing code bases that are mostly only intelligible to the sole developer that worked on it.

Personally, I'd put my money on a non-profit third-party organization like Khan Academy to be able to design something that gets traction. They know the space, have world-class engineers and would be able to generate enough PR to start a community around a tool that they open source.

If Google offers the sort of export tools that they've offered for other services, then I can feel a bit better about their creating critical tools for public education.

I think of Google's place here like I thoughts of Apple's in the world of smartphones -- they are in the position to normalize a new way of learning, which is the largest hurdle that perhaps only a powerful private company can accomplish these days. We can worry about the need to depose them with better tools if/when the need arises :)

Data and interface to that data are two different things.

Requiring that data be open and standardized (as far as import-export is concerned) makes sense, requiring that UI to manipulate that data is built by a government-affiliated entity is questionable.

Even if you love government-designed Web sites, there might be someone somewhere capable of building faster, more intuitive, mobile-friendlier front-end.

Came to say pretty much the same thing. This is extremely creepy given the nature of Google as a data munging operation that converts data to ad revenue. On the other hand mikeleeorg makes a good point. In the past competitors where sued out of the water by Blackboard. If Google puts some pressure on them and makes the data easy enough export then a more agile company can have a shot at improving things. That's a pretty long chain of extremely happy accidents though.

The reality is that this data is already out there. Faculty will us whatever is most convenient, and that's never been BlackBoard or whatever the official IT service is. Student grades will end up in DropBox or in a Google doc, FERPA or not, unless there's some sort of auditing.

In my country (Spain) it is most likely illegal to store the students' assignments on a server which does not belong to the University.

Then again, the Univ. at which I teach outsources the e-mail to Microsoft.

I tend to not understand.

And I agree: students data belongs to the University, not to google.

My university outsources the email to Google apps for Education.

Are you suggesting that schools develop their own solutions, over and over again? Some school districts centralize their IT solutions provision, but that's nothing compared to a private company providing a solution nationwide.

Reply on news.ycombinator.com