Reliance On Web Applications

01 Apr 2015

Every day, features of traditional websites approach those of stand-alone desktop programs. Strictly informational websites with simple indexes and content still have their place on the web and they probably will continue for a long time, but this post isn’t about that.

Commonly called web-apps, they often offer only a few actual “pages”(typically there are no true page changes once you’re logged in, the page is simply redrawn to make it appear as though there is page navigation). These web-apps maintain state, redraw their user-interfaces (sometimes with animations), have multiple threads running in parallel and behave like stand-alone desktop applications in just about as many ways as you can think of.

We see these web-apps vying with their native application competitors for market share, performance, ease-of-use, features and reliability among many other things. An example is the comparison of the Google Drive suite with the Microsoft Office Suite or Libre Office. In fact, Google’s offering of their Chromebook laptop whose operating system’s window manager is quite literally a web browser is the most telling example of exactly how far this trend can be taken. On those devices quite literally “everything” is done and runs within in the browser. The main reasons why some choose to utilize these web applications over a native application include the following:

  1. Ease of access. Everyone has a browser, everyone knows how to type a URL into their address bar.
  2. Perceived security. Visiting a website ought to be safer than downloading a full-fledged application which will be installed into the operating system. Presumably the browser can safely “sandbox” the code and minimize any damage that it might try to wreck upon my computer or other browser tabs.
  3. Convenience. All activity seems to be synced quite well into the cloud and are available wherever I can access my user account on a browser.
  4. Exclusivity. The software isn’t available as a stand-alone application to be installed into the operating system, thus the user has no choice if they wish to utilize the app’s features.

As an extra aside, a software developer might choose to develop their next great software idea as a web-application rather than a standalone desktop application simply because it’s typically much less work to do so and the resultant program will be in essence functionally equivalent.

I like to use myself as the first example whenever possible. So I’ll begin by listing the web-apps I find to be most useful to my own workflows.

  1. Trello. It’s a free online whiteboard for stickies and post-its- on drugs. To-date I haven’t found an organizational took like it. Whenever I have a new idea I’m often overwhelmed about where to begin. Enter Trello. I create a new board with a title of the idea “Solve World Hunger” and begin creating lists and cards to organize my plan of assault. It’s also trivial to add collaborators into the mix so that I can organize with my colleagues or partners.
  2. Gmail / Google Inbox. This is much more than just an e-mail access portal. I believe Gmail’s browser offering fits the profile of a full-fledged email client.
  3. Github. It’s the best and easiest graphical Git client/visualization tool that I know of. There are others that come close. (See Atlassian Bitbucket, I use this also because it’s cheaper to host private repositories).
  4. NewRelic Server and Application monitoring.
  5. Google Drive & Office Suite
  6. Soundcloud

The only true initial upfront cost of these applications is bandwidth, at least at face value. However I think it’s worth noting a few considerable drawbacks when it comes to using a web-application.

The first is security. In a nutshell when you use a web-app, your browser requests a page from a computer somewhere else on the planet that is also connected to the internet. Let us use the example of Gmail. You begin by typing https://gmail.com into your browser. Your browser fires a request off to Google’s servers for the web-page and code that make up that webapp. Presumably Google receives the request and sends you back the data you requested. The assumption here is that your request did genuinely make it to Google’s servers and the response has not been tampered with in any way. Anything short of this, and you could actually be receiving a modified version of gmail’s web-app that likes to eavesdrop on all of your emails, or record your password or worse. There are techniques for mitigating the risk of such attacks (and this is just one simple vulnerability, there are numerous) but there aren’t very many guarantees for security with this model.

Another bottom line is that all your precious private data lives in the cloud and you’re just going to have to trust that it’s being stored in the most secure ways possible and that even [the threat of a disgruntled employee wrecking havoc on the inside of the provider’s organization or stealing data] (http://www.businessinsider.com/morgan-stanley-employee-steals-client-data-2015-1) is non-existent.

The next major drawback I see with web-apps has to do with data retention. If you use these services as often as I do, you begin to accumulate a lot of stored data. Emails, documents, bank statements, organization lists, digital media, contracts and everything in between. You then begin to wonder what measures the service provider has taken to reduce the possibility of precious data loss. Then, assuming they’ve left no stone unturned to ensure that every single byte is accounted for, what if [they go out of business or otherwise fall off the face of the planet] (http://www.posterous.com/)? There seems to be no standardized way to keep a backup of the data stored therein and it seems like that’s just too much trust to place in the hands of these entities.

Published on 01 Apr 2015

Next