Web Interfaces/Web Apps - opinions

This topic was created by Greg D .

  1. Greg D

    Web Interfaces/Web Apps - opinions

    So a discussion came up at work, with a friend of mine talking about building an app for his services, but making it a web app to save costs.

    Personally, I hate them. They are the devils own idea. They suck harder than a hard sucking thing. The main reason for my diatribe against web apps/web interfaces is web browsers themselves, as well as the uselessly written and un-maintained plugins used by big-name vendors to jerry-rig them to work with their hardware. Browsers suck down memory at a silly rate, they crash often, and most of these web "apps" use poorly coded plugins to make them work properly.

    What makes it worse, is the big-name vendors jumping on the bandwagon. For e.g. recently we upgraded our vSphere environment to ESXi 5.5, which forces you to use the web app to manage the fucking thing. To say its horrid, and unfit for purpose is being nice. It actually relies entirely on Flash! What annoys me is that they had a perfectly good, respected, and fully functioning native app, which is STILL miles better.

    This story rings true for many others - Solarwinds is another big one - we went from an awesome native app, to a horrible mess of a web version, and they completely pulled support for the functioning and nice to use native jobby.

    The thing is everyone and his dog is doing it (switching to web browsers for their apps). What am I missing? When did native apps become so unloved by vendors? Is the cost saving THAT big? If so, WHY!?

    Where do you lot sit on this one?

    1. This post has been deleted by its author

      1. Greg D

        Re: Web Interfaces/Web Apps - opinions

        "If somebody chooses to use Mozilla or some other hideous browser to run it, that's their problem, you didn't inflict it on them."

        Well therein lies my beef with the browser. As a developer, you are basically pushing the problem on to the customer (over which application to use etc) and removing their personal choice. Once the collection of web apps, designed to run in their supported browser, builds up, you are left with a customer machine running 4 different browsers, which is a nightmare in an Enterprise environment (like mine).

        My ethos has always been, you are providing a service to a customer, therefore your service should not have to rely on the customer using a specific browser, which is a personal choice and there are many. If there was a single browser for the entire internet, and it was really good, stable and well maintained, there wouldn't be a problem.

        To me the main issue with web interface is it just seems outright lazy, and a quick way to save a few quid. I get it can be a pain to patch (a fat client), and get working cross-platform, but thats the industry you're working in, so deal with it :) You [insert software/hardware company here] knew the risks, dont start pushing the problems on to the customer to make life easier for you.

  2. Anonymous Coward
    Anonymous Coward

    A well written 'thick client' is pretty much always going to provide a better user experience than a browser based thin client. That said thick clients suffer from:

    Being a pain to upgrade. You might have hundreds of users all requiring a visit to their desktop by support. The client may have to support running on multiple versions of OS and support 32 and 64 bit. That problem doesn't exist for an app running on one web server.

    Being a pain to support. What version of the app is the user running? If it's a web app there is only one version. Does the app rely on various drivers or OS components that could be of different versions on different PCs.

    Dealing with large quantities of data. A web app could be on a server located physically close to the database server, with a gigabit network connection inside a server room. The data only has to move across the network once to the web server. With a native app you might have to move loads of data across the network or maybe even across the world. This could take a long time and consume masses of memory on the end users devices.

    In my opinion the move to web apps is driven by IT departments, for who they remove a lot of headaches. In the same way that IT departments put horrible little boxes on people's desks that have no hard drive or decent video card and just connect to a users desktop on a remote virtual machine. Both things reduce user satisfaction but make the IT department's jobs easier.

    To me the VMware client is the perfect app to be native. There are normally very few users, who are technical so can handle upgrading it themselves. Those users are likely physical located near the VMware server (mostly). The app doesn't need to move huge amounts of data to the client.

    1. Greg D

      Good points, well made.

      I guess I'm coming at this from the Enterprise Sysadmin side of things, as opposed to the IT department forcing this down. Big name vendors, not outsourced IT depts (companies that do that deserve what they get).

      Why vmWare thought a web app would be better than native, I do not know. I'm still using native, but they removed support for VM hardware above version 8 on the native client. The web interface is HORRID. As you say, its a perfect app to be native. Clearly they don't think so!!

      In my experience, I've noticed that web is always slower, no matter where the database is stored. E.g. our Solarwinds platform is 2500miles away from us, but used to be much faster through the native app we used. Now its on web, the web server is basically slow, and dont forget the vendor is effectively forcing the customer to install MORE hardware (i.e. a web server), to run the damn thing, and the spec of that creeps up and up.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2020