I've worked for quite some time in the digital display industry, at a company which has a very feature rich, Mac OS X based digital signage solution [1] with high profile customers such as Mercedes-Benz, or Disney. Since the company is small, it is easy to be part of all steps of the process, including marketing and sales. I learned a lot about B2B buying behaviour while I worked for this company.
We oftentimes ran into the situation where a potential customer really liked our solution, but wanted to run Windows instead of Mac OS X (this was arguably a lot tougher in 2005 when we started, and less so after the iPhone hit). There were a variety of reasons for this:
- Windows was the OS that the company was also using for employee machines, so they knew it.
- The IT department only wanted Windows machines on their network (our grand solution for this was that our machines were sold with direct support and did not need to run on the customers network, which also removed a lot of security pain)
- The customer was afraid of Mac OS X because they thought it would not be stable enough
But most importantly, it came down to simple risk aversion. The employee at a big company that was internally responsible for the 'digital signage' project, would base every decision on which choice would be less risky for his carreer in case the project fails. Lets say they roll out digital signage, and it costs a lot of money, and it doesn't work right. In that case, the employee needs to be able to defend himself against all sorts of questions: "Why was this product chosen", "Why was this vendor chosen", "Why was it implemented in this way", etc. In that case, being able to answer "We choose Windows, because it is the de facto standard" is better for the career than having to say "It seemed to be a stable product". Out of this reason alone, we found companies would choose a solution where they knew that it was far worse (less stable, less flexible, less features) simply because on paper it looked less risky.
I found that there's a great gap or schism between common digital signage projects. On the one hand you have solutions that just want to display 5 pictures or a simple website (current dish of the day, bus tables, train schedules, current news) and on the other hand you have high profile vendors that want (multiple) absolutely stutter free HD content (with huge bitrates because so much is going on in their animations), animate this in 3D, dynamic content (i.e. current weather, time, stocks, etc), and integration with other external technologies / systems (light control, sound control, etc). Thankfully, nowadays most of this can be achieved with a fullscreen chromium or webkit, but you need serious hardware for that.
The problem arises if people buy the lightweight hardware like a Pi and then try to run the complex stuff.
Also, most people seem to ignore just how important the backend / cms system for these installations is. The content that is deployed on systems is already out of sync as soon as you plug the systems in :) I still remember fondly how I watched one of our competitors distribute content (on a digital signage fair) to his systems using WS FTP for Windows. That's not a solution you should sell to a demanding customer, he'll simply call you whenever he wants a new image on there.
We've had someone put a graphically intensive Flash application on Atom processors. Their reasoning was lower heat output and very small footprint. There are newer i5's that are maybe twice as big that absolutely blow these things away.
Customers want cheap and I don't blame them but it took a second iteration of the shitty Atom line, one with a proper GFX card from Nvidia to have this display in any usable capacity. Atom/Netbook is absolutely not the way to go for Flash content. I would wager the Pi would be roughly identical unless you could leverage a beefy GFX card.
If you weren't using Flash, you could practically use anything. HTML5/CSS3 animations have come quite a distance but its definitely not 1:1 still.
We built a custom web page that auto detects and scales to different resolutions. We also made a self service admin interface but it was just text and a few images.
The Pi has a surprisingly powerful GPU but I've not tried HW acceleration from the browser. Something to look into maybe.
I haven't looked into the Pi yet, we used Mac Minis and frequently ran into its limits, so I just thought that it should be even more problematic on the Pi.
I do think that the market for small and simple solutions (webpage with a bit of text, images, etc) is far more interesting and bigger than trying to sell to the Fortune 500 like my former company does, because these have really difficult requirements, and they're prone to simply rolling their own solution. So in short I think doing Digital Signage on the Pi is a great idea :)
I think this is just a kinda microcosm for why Windows is successful in general. Because they have a sales force and support that markets to and is geared around companies. Their developer ecosystem is heavily geared towards corporate systems as well.
It's surprising that a smart HN blogger can't put himself into somebody else's shoes for a second. When you're a manager and your boss tells you to go install some kiosks, chances are good that you're going to go out there and get some proposals. You're gonna pick whichever one of them best represents themselves, has a price that you want and seems like they'll be there when you need their support. You're probably not going to go with whoever is the most "cool" or "bleeding edge." You're almost definitely not going to try to put together an open source solution yourself just to save a few thousand bucks on OS licenses.
It's just as simple as that. Windows kiosks are in use because there are a lot of people out there selling Windows Kiosk solutions. If at some point another system manages to get a decent marketshare and has sales people out there writing proposals - then you'll start to see more of those.
You frame the three points a bit derogatory and seem to have missed the point of your customers.
> Windows was the OS that the company was also using for employee machines, so they knew it.
They understand it, have an idea about the cost and amount of support necessary. "They knew it" seems to convey that they were essentially stupid or not adventurous enough to try something else.
> The IT department only wanted Windows machines on their network
Again: a complete valid reason in my view. Here on this forum you have even people advocating that server and client in web apps should be written in the same language to keep things consistent.
> The customer was afraid of Mac OS X because they thought it would not be stable enough
You make that sound like there teeth were rattling in horror, but apparently failed to convince them that OSX would be stable enough?
I'm not a Windows developer or -user, indications that win 98 runs on my ATM also make me concerned.
But "muhaha, they are just stupid bricks for using windows" (no, you did not say that, but it sums the attitude) shows a lack of understanding and, as a consequence, does not advance the spread of alternatives into these environments IMO.
Sorry, that wasn't my attitude nor was I intending to sound that way. I fully understand why they'd rather use a Windows solution, I just wanted to explain in a simple way what I thought their motivations were.
> "They knew it" seems to convey that they were essentially stupid or not adventurous enough to try something else.
I did not intend to convey that they were stupid or not adventurous enough, I simply thought "knew it" would be a sufficient enough description. I did not intend "knew it" in a pejorative way at all.
> You make that sound like there teeth were rattling in horror
No, not really, actually I remember discussions with big companies where those were the exact words ("Our IT department is concerned with the stability of this Mac OS X")
> "muhaha, they are just stupid bricks for using windows
Again, no, I just tried to explain the seemingly irrational behaviour (for tech people) of choosing a technologically worse solution even though better solutions exists. I wasn't even talking about Windows actually, but about digital signage solutions based on Windows. I'd rather not go and open the box of pandora that is Windows <> Linux <> Mac OS X. Most of our competitors that ran on Windows (situation may have changed, I left that company more than 2 years ago) were on the feature side and in terms of possibilities worse than our solution (not only because it was easy to pull of a lot of things with Core Graphics really easy).
I really can't see where you read the "those people were stupid" attitude, but it may be a cultural difference, I'm not a native english speaker. Nevertheless, even if I sounded that way, it totally wasn't intended that way. If I were in a big (non-tech) company in such a situation, I might even decide in a similar way out of the above reasons.
So, from my 6yrs in SBC/AT&T, the OP is completely correct and I don't see anything insulting in his comment. That's exactly how they made decisions when I was there. It doesn't make them wrong per se, but that's how they functioned. In fact, I think OP was being nice. A lot nicer than the description I'd give.
I could type up 3 pages of my experience there, but I'll just say that because it was my first job: It took me 4 years to confidently identify the 3 points OP listed(that it wasn't me that had a problem), one more year of research for me to learn that living in the bay area put me in the middle of the tech-start-up scene that had a mindset like my own, and another year to replace all the expensive proprietary software tools on my resume with trendy open-source alternatives.... then I left.
I did not read that attitude from terhechte at all; I did not find the points "derogatory." To me, it seemed like terhechte has a good understanding of why his customers reacted how they did, and believed those reactions were reasonable.
Where I work they have a few gratuitous large-screen displays that in reality just suck up electricity while presenting slides about how "green" our IT organization is.
They use Mac Minis, with full-screen Chrome to run the slideshow, and I often see error messages on these (though typically from having dropped the wifi connection and not from an OS panic or other real error).
I can tell from experience that having a really solid error free digital signage solution takes up a lot of time, because you really want to care about every tiny detail that could go wrong. One example is sudden lack of network connectivity while displaying a website:
- One solution is to just display an error page (bad!)
- Another solution is to display a cached version of the page. Sounds great until you have to display stocks.
- The proper solution is to display a cached version, except for time critical data where you either go to the next piece of content, and if that's impossible, display the customer logo or some default fallback content.
This and thousands of other small configurations are what make up a good digital signage solution. The default web browser, being made for user interaction, doesn't cover it, obviously.
I had a similar experience in my career, well over a decade ago. Granted, my anecdote doesn't prove anything, especially because I could have been wrong at the time, but it reflects the mindset.
I had a choice of implementing a machine controller with GUI in MacOS (System 7.5), some flavor of Windows, or even MS-DOS! Two factors influenced my choice:
1. Familiarity with the innards of a PC assured me that I could make it work all by myself, no matter what, even if it turned out to be ugly on the inside.
2. Looking ahead, and realizing that a bunch of MacOS development experience on my resume would elicit a yawn from most recruiters.
As for the display controller, today I'd do it with something like a Raspberry Pi.
Depends on the solution that you buy. Some digital signage vendors sell complete black boxes where you get a hardware that you connect, and it runs just fine, others sell licenses that you have to install on your own hardware.
The Videro solution is a full packaged system that you connect to a display and to a network (if you're connecting to wireless, there's a special remote UI that pops up after first start and lets you enter wireless info, much like on Apple TV systems) just runs. It usually connects to a cloud system, and then you can distribute content to it. It is really hassle free, which is also important for large scale deployments. (If one machine breaks, you can just pop it out, and put a new machine in, one click and it will receive all the content from the previous system and continue the job).
The solution is written as a native Mac OS X Objective-C app, and can render all kinds of content (images, videos, html, flash, quartz composer, texts, tickers, etc). These can be mixed and combined and layered in any possible way. It is a pretty powerful solution.
Here's a really, really, old demo of the software:
We oftentimes ran into the situation where a potential customer really liked our solution, but wanted to run Windows instead of Mac OS X (this was arguably a lot tougher in 2005 when we started, and less so after the iPhone hit). There were a variety of reasons for this:
- Windows was the OS that the company was also using for employee machines, so they knew it.
- The IT department only wanted Windows machines on their network (our grand solution for this was that our machines were sold with direct support and did not need to run on the customers network, which also removed a lot of security pain)
- The customer was afraid of Mac OS X because they thought it would not be stable enough
But most importantly, it came down to simple risk aversion. The employee at a big company that was internally responsible for the 'digital signage' project, would base every decision on which choice would be less risky for his carreer in case the project fails. Lets say they roll out digital signage, and it costs a lot of money, and it doesn't work right. In that case, the employee needs to be able to defend himself against all sorts of questions: "Why was this product chosen", "Why was this vendor chosen", "Why was it implemented in this way", etc. In that case, being able to answer "We choose Windows, because it is the de facto standard" is better for the career than having to say "It seemed to be a stable product". Out of this reason alone, we found companies would choose a solution where they knew that it was far worse (less stable, less flexible, less features) simply because on paper it looked less risky.
[1] http://www.videro.com/