Ask AnandTech: Tablets at Work, How Important is Backwards Compatibility?

38 Comments

View All Comments

• trane - Tuesday, June 18, 2013 - link

  It depends on your occupation. Some can get by with productivity suites just fine, while others need solutions like Autodesk Revit or Adobe Premiere Pro. For the latter, the reality of the situation is that hardcore productivity apps are only available on the Desktop. Hence x86 and Windows are a must today for engineers, artists, architects, accountants and a pretty vast base of professionals.
  
  iOS or Android - even if we were to remove the Desktop these would still too simplistic compared to Windows 8.1 Metro for productivity. I am sure a lot of people can get by, particularly writers and managers, but again, Windows 8.1 has the advantage. The biggest deal in 8.1 as far as I am concerned it finally brings viable and flexible multi-window multi-tasking to a bonafide touch interface.
  
  Looking further, I see no reason as Metro matures to see the likes of Office first, then Photoshop, followed by the likes of Solidworks or Maya looking at the long term, heading to WinRT. Looking at a similarly long term, I would expect ARM to bridge the performance gap. So there will come a time where Windows RT would require some serious consideration. In this future, I wonder if Android or iOS can catch up to Windows as a powerful and fully featured OS?
  
  But today, there's no doubt in my mind. For my work, give me a Haswell Y series tablet running Windows 8.1.

• Da W - Tuesday, June 18, 2013 - link

  I could not do office work with anything less than Microsoft Office - full microsoft office. Right now it means x86 compatibility, with enough power to be enjoyable (baring atom from the list) and a fair price that isn't highway robery (looking at you samsung), it feels only the Surface Pro is good enough for what i do.
  
  However, i could see a thin and light ARM powered tablet (windows RT or Android) being used with an heavy x86 docking station (think quad core haswell + GTX780M) and you can replace anyone independantly since they remain compatible form a generation to the next. The docking station is desktop replacement oriented and hooked to 3X24" monitors, external speakers and wireless keyboard+mouse. For travel a simple surface-type keyboard cover does the job. If any OEM makes one of those i'm a buyer.

• Flunk - Tuesday, June 18, 2013 - link

  For me, it's not at all important. The older systems that we use all have interfaces that are totally unusable on a tablet, everything is far too small to tap and they try to jam too much info on the screen.
  
  Web applications, however are totally usable so it doesn't really matter what kind of tablet it is, we can support it. So from my perspective as long as the HTML5 support is good enough it everything else doesn't matter.

• keithzg - Tuesday, June 18, 2013 - link

  For me, backwards compatibility with x86 applications is semi-important. That is to say, backwards compatibility with x86 *nix applications. It's a bit annoying that Android doesn't support a normal Linux userland, and I've yet to find an X11 server that isn't junk for it. So that's been annoying (ironically often making my cellphones, which are Linux+glibc-based Nokias, more practical on paper---what I wouldn't give for a Harmattan tablet), but not overly so, as many commandline programs at least will compile and run on Android or have been ported, and although no terminal emulators are as good as the Nokia ones there are some fairly decent ones. So for my work---for which I'm the entire IT department for a small company that develops Windows software, so we're rather strictly bifurcated with Linux on the servers and Windows on each developers' desktop---I end up getting *more* backwards compatibility that actually matters to me on Android. I desperately hope that the Vivaldi tablet comes out sometime soon, however.
  
  It's not like my job doesn't involve Windows; I'll sometimes be at someone's desk and they'll be complaining about weirdness in Microsoft Access, and I'll have my tablet with a few sessions open connected to our MySQL server running on Linux. But what my tablet needs to interface with and play nice with is x86 Linux servers, not x86 Windows desktops.
  
  I know my use-case isn't necessarily a *standard* business use-case, and I'm stretching the description a bit here, but I felt it was worthwhile to give a different perspective.

• yougotkicked - Tuesday, June 18, 2013 - link

  As a computer science student and beginner developer, I honestly feel that backwards compatibility impedes progress, especially in new platforms like android. There is no shortage of developers in the world, and forcing them to make new apps rather than let old ones dominate the market through popularity alone is strictly good for the market.
  
  As a user, I can understand some people's desire to keep the same work environment. But the way I see it if you can get used to using a tablet after a lifetime of using the mouse and keyboard, you can get used to using new software interfaces as well.

• MrSpadge - Tuesday, June 18, 2013 - link

  Sure you can, but is it worth the time and effort? On developer and user side? That answer is surely not a resounding, general "yes".

• yougotkicked - Tuesday, June 18, 2013 - link

  I'd say it depends on what the changes are. If it's a microsoft office style update where they change things just to sell a new looking product, it's not very worthwhile. If it actually improves the user experience or makes the software easier to maintain, then it certainly is.
  
  I'm not saying things need to change just for the sake of change. But new and better things shouldn't be blocked just because some people don't like learning new things.

• A5 - Tuesday, June 18, 2013 - link

  Dumping backwards compatibility is a good way to get your IT guys to murder you in your sleep. Not to mention a huge waste of resources if your old software already works.
  
  You'll learn that once you leave school.

• Sushisamurai - Tuesday, June 18, 2013 - link

  I have to agree, loss of backwards compatibility would waste time and money (training staff, buggy software/technical support, new license purchases, software/developer adoption to the new standard, IT training/knowledge to adapt to new software/hardware), not to mention a layer of redundancy lost in the transition as you already had a prior, working solution that everyone was comfortable to and trained for

• yougotkicked - Tuesday, June 18, 2013 - link

  I see my comment was a tad over simplified. I'm not suggesting that every program available to a platform needs to be unique to that generation of that platform. But if google wants to change some data structures in one of the android Java libraries, they shouldn't bloat the new library with legacy support so some apps will be compatible with the new release. Those apps would still work on the devices they are being used on, and the developers can restructure parts of the app to work with the new version.
  
  In short, I don't think backwards compatibility is universally bad, far from it. I just think current trends put too much value on supporting old software, and forcing users and developers to update more often could do the industry some good.