Slashdot Mirror


Windows Software Ugly, Boring & Uninspired

An anonymous reader writes "CPU magazine has written a very straight-to-the-point editorial on the lack of quality and innovation in software for the mainstream OS. They compare it to the Mac, which is found in a much different light. Where has all the innovation gone?" From the article: "There's too much coal and not enough diamonds within the sphere of downloads. The greatest pieces of software are plagued by unintelligent design, and very few rise to the level of ubiquity. Windows users don't have a strong sense of belonging; there's no user community rallying around the platform. We use the computer, certainly, or is the computer using us?"

11 of 924 comments (clear)

  1. Re:Just an idea, but by packetbasher · · Score: 5, Interesting

    Actually when I think of the hardcore mac user I think of people attending XML conferences, Next hackers, people at the MIT doing OS research, etc.

    A friend of my once said that OSX is the 21st century Sun workstation.

    Maybe I just think that because I dig having a unix box that can also run microsoft word at the same time.

  2. Marketshare, Quality, and Economic Viability by G4from128k · · Score: 4, Interesting

    I sure some of this is due to market-share issues. A developer targeting Windows knows there's 200 million new PC shipped each year (and probably a billion PCs installed). They figure that their software only needs to be good enough to snag only 1% of users to sell 2,000,000 copies a year and gain a 5 million user install base. In contrast, the Mac developer looks at Apple's 3% market-share (say 6 million Macs/year) and thinks that they need to attract 33% of the user base to reach the same target sales figure.

    The result is that only the most dedicated and talented Mac developers survive whereas any idiot with a C-compiler can create a PC software title and be assured of some sales (just convince 1-in-10,000 PC users to spend $29 and you gross $600k per year). Given the huge market-share disparity, Mac software must be 30X as good as PC software to survive in its small marketplace. (OK, its a bit more complicated due to dilution by competing vendors, but I'm sure its much harder on the Mac side to attract an economically viable user-base for software package.)

    --
    Two wrongs don't make a right, but three lefts do.
  3. Re:Just an idea, but by Seumas · · Score: 5, Interesting

    Think about it, who do you think of when you think of a mac user? Granted, there are many out there, but when I think of a hardcore mac user I think of somebody who is into designing music, movies, graphics editing, etc. They are designed to cater to a group of people who are more creative and right brained.

    Really? I work for a huge company known for its big iron and most popular unix operating system and a silly coffee-related programming language and a CEO that has been ranked at the bottom of several CEO lists in terms of performance the last few years.

    And do you know what most of the developers and engineers I know around here have with them? Their PowerBook.

  4. Re:Garbage by pherthyl · · Score: 4, Interesting

    Why don't we take this to extremes?

    Want to work on your report? Why wait for your word processor to load when you can just press F12 and it's RIGHT THERE!

    What if you want to watch a movie? Just press F12 and there's your movie player! Wow!

    Dashboard is only a way to keep applications loaded in memory and display a certain subset of them at a keypress, this is absolutely nothing new. So I want to do a quick calculation, I hit the shortcut key I bound to my calculator and there it is. When I'm done with it I close it and it doesn't suck up memory. I see absolutely no value in keeping these applications running all the time when you're barely ever using them and could just pull them up on demand anyway.

    The original author of this article seems bored by his functional applications. That's ok, some people like flash over functionality.
    I've used OS X a fair bit and didn't see anything that I was particularly impressed by. It sure looks nice, but I'm not more productive or happy with it than any other platform.

  5. Re:Windows... by JonTurner · · Score: 4, Interesting

    >>Am I the only one who is completely unclear on what was intended by this comparison?

    No, you're exactly right. The functionality of windows has been essentially static since Win95 and ugly, grey, square windows look equally bad no matter what numbers the "About Windows..." box contains.

    Now, the problem with looking at the changes between NT->2k->XP is that, well, for the most part you can't look at the changes. Other than a green "start" button, what's the difference in terms of *user experience*? Where's the innovation? I can't find it.

    Spotlight, Automator, Rendevous, (and yes, even Widgets) IMO all work to make the user more productive. Apple changes their OS every year. Sometimes for the better, occasionally for the worse ("two steps forward, one step back") but at least they're making progress and trying new ideas.

    Microsoft is simply hung up on locking people into their technology and making it too expensive/difficult to transition away. Proof? How 'bout .Net, just for starters.

    Anyone still doubt? Well, then, did you hear about that beautiful, innovative new technology in Microsoft's latest OS release that makes users much more productive? Yeah, neither did I. The big stories out of Redmond mostly concern what *isn't* going to be in Longhorn.

    Sorry, fanboys, but Windows innovation isn't.

    Disagree? Feel free to list MSFT's post Win95 innovations that improve the user experience right here ___________________________________.

  6. Re:Just wait, it'll come to Linux too. by gregmac · · Score: 4, Interesting

    With rapid development environments like Visual Basic around for the Windows OS, it's not surprising that there is a lot more crap out there for Windows, verses other OS that don't have these easy to pick up IDEs. It simply takes a more developed skill set to write apps for MAC and *nix.

    While this contributes to the problem, there are a ton of of ugly apps for *nix (can't speak for Mac since I don't own one). There are a lot of apps that don't even have GUIs, and are also very hard to use on the command line (cdrecord, for example). These apps are still very useful and work very well, they're just ugly in the sense that you can't "just use" them. You need to specify tons of switches, spending time reading the man page, or they require a front-end application that builds the switches for you.

    You imply that a skilled developer == someone who is good at developing interfaces, while really, it's a totally different skill set. You can tell when programmers design web pages, and think that because they know HTML, CSS, javascript and photoshop very well, that they're incredibly talented graphic designers.

    I think that when (not if) a high quality and easy to learn development platform for Linux comes along, we'll start to see mountains of shit for it, too.

    I think you're right here too. Making it easier to develop apps will mean that more developers will come in, and they probably will also lack basic design skills, which means you get more ugly AND poorly-written code. Just don't confuse the issue and think that it's only unskilled developers that write ugly interfaces.

    --
    Speak before you think
  7. Re:It's a tool, not a piece of art by daviddennis · · Score: 4, Interesting

    If inspiration and brilliance of design were not in some way important to people, we would all wear Mao suits or prison garb. Much more practical, you know. Easily interchangeable. No reason to have different designs just because there are two sexes.

    When I flew to Cuba, I rode on a Soviet jet, something called a Yakolev YAK-42D. It felt like something from the 1950s. I later learned it was a 1950s design that they only got around to making circa 1981.

    A Soviet product is just what you want. If a Soviet plane takes off, flies for a time and lands successfully, it has done its job. There's no need to make the flying experience pleasant. Flying is for those evil bourgeois chaps who can afford to fly anyway, and there's no reason in the world to coddle them.

    On-seat power outlets for your laptop? Forget it.

    Seatback TV screens? Not even close.

    Comfy leather seats? Those are decadant luxuries of the West, don't you know.

    Well, I'm sorry.

    I'm a decadant, luxury-loving product of the West. I like my Mercedes-Benz automobile, because it was carefully and thoughtfully designed. And I love my PowerMac G5 and PoweBook for the same reason. Carefully and thoughtfully and elegantly designed products are a good in and of themselves; millions of iPod users sense this even if they don't quite realize why.

    Maybe a factory punch press isn't something you can design this way, although perhaps that's because nobody's even tried. In any event, we are not working in a factory, and when we work on computers all day, our comfort is essential. If the more creative software vendors realize this is most true on the Mac, and cater to it, it simply means I've chosen the right platform.

    The one designed for people like me.

    You can have your gloomy gus Windows 2000 interface, as long as you don't make me use it.

    D

  8. Most people barely use the shell by EMIce · · Score: 5, Interesting

    Most users I've noticed are perplexed with explorer and its interface. They know specific hierarchies like My Documents and Program Files, but as soon as you drop them into an unfamiliar shell hierarchy, they aren't sure "what to click on" or what in general is possible.

    A new interface based in windows shell may be organized the same as others but is functionally different, and people end up looking for things that they are "allowed" to click, like they might an exe in Program Files, or a doc in My Documents. It is far from intuitive, as these custom hierarchies don't necessarily order things intuitively and even when they do, functionality varies from object to object whether you click, double click, or drag and drop.

    Functionality of different actions should be implicit in the design, so they can be inferred by those unfamiliar with what actions are possible in a particular application context. Now if windows made it standard that right clicking on an object should not only bring up object-specific options, but also describe simply what drag and click operations are available with respect to that object, then these interfaces might not be such a mystery.

    People aren't that dumb, they'll learn given context sensitive documentation like this. Finding their way to documentation is otherwise too frustrating, as it is often mired in a web of unfamiliar material. The frustration the average joe faces at a PC is enough to make him learn, if given a more accessible way to find the immediately relevant sources. He doesn't need to understand why the whole damn system works to find one particular solution, he'll generalize that with enough access to particular solutions.

  9. Yes, and here's what MS did wrong... by tkrotchko · · Score: 5, Interesting

    1) There was no benefit to making the registry a non-text file, except that MS wanted to make it more difficult for end-users to poke around and understand more clearly what's going on

    2) Applications do have to use the OS to read/write/update (so far so good), but the OS *never tracks what the application puts there*. As a result, every developer puts their copy protection in obscure keys in the registry. Even worse, and unforgiveable, are applications that leave crap behind.

    3) Keeping it all in one place (i.e. registry) sounds like a great idea... until you realize you can't readily *do* anything with it from a user's perspective because guess what... the OS won't let you do a simple "c:>copy registry to registry.backup".

    This could be solved easily:

    1) Make it impossible for an application to write to c:\windows or c:\windows\system32 or... you get the idea

    2) Registry files should be stored locally in the directory the application was stored in, or better yet in "My Directory". The system would have its own registry stored in the system directory.

    3) They should be text files that can be copied by the user easily using standard tools.

    4) When a program is uninstalled, the OS would ensure all traces of the registry entry are deleted (this is easy because of #2)

    5) The only thing allowed to alter a program's registry entry is that program. And every time its altered, a new version is kept. This would allow users to go back to old version if required.

    6) A user could tell the OS to lock a registry so that nothing can alter it

    7) The system registry could never be altered by any application. Requests to modify would require the root password entered by the user. Every time.

    This is easy. But MS makes it hard and in the process makes registry damage fatal to the system. With no way to properly back it up. So they have goofy "restore points" that you can't explain readily what it does. So then they'll add more utilities instead of following the KISS principle.

    I sometimes feel over at MS they have a bunch of brilliant programmers who have never set foot outside of Microsoft and don't understand the issues with their own product.

    --
    You were mistaken. Which is odd, since memory shouldn't be a problem for you
    1. Re:Yes, and here's what MS did wrong... by drsmithy · · Score: 5, Interesting
      1. There was no benefit to making the registry a non-text file, [...]

      There are a lot of good reasons why the registry is better than a text file. Performance and fine-grained permissions are two.

      [...] except that MS wanted to make it more difficult for end-users to poke around and understand more clearly what's going on

      Yes, because a system encouraging manual configuration no input validation is such a better alternative.

      Users _shouldn't_ be directly editing the registry. Ideally, users _shouldn't_ be directly editing text files in /etc, either. This is not to hide anything from them, it's so they don't break the system by making a typo.

      Manual editing of text files is an incredibly bad way to configure a system by just about every measure thinkable. That there are few _better_ methods does not change this.

      Applications do have to use the OS to read/write/update (so far so good), but the OS *never tracks what the application puts there*. As a result, every developer puts their copy protection in obscure keys in the registry. Even worse, and unforgiveable, are applications that leave crap behind.

      Neither does any other OS I can think of - so what's your point ?

      Make it impossible for an application to write to c:\windows or c:\windows\system32 or... you get the idea

      They can't unless they're running as a user with sufficient privileges - just like every other multiuser OS.

      Registry files should be stored locally in the directory the application was stored in, or better yet in "My Directory". The system would have its own registry stored in the system directory.

      The user's registry hive is stored in their user profile. The system registry hive is stored in the system directory. Ie: it's already the way you want it.

      They should be text files that can be copied by the user easily using standard tools.

      How are you planning on implementing per-user, per-value ACLs on lines of text in a file ? How about making sure modifications don't end up half finished ? Are you aware parsing text is an incredibly inefficient operation ?

      When a program is uninstalled, the OS would ensure all traces of the registry entry are deleted (this is easy because of #2)

      But how to deal with poorly written applications that don't tell the OS everything they do ?

      The only thing allowed to alter a program's registry entry is that program.

      Funny, I would have thought you'd want to allow the user to manually manipulate arbitrary registry settings.

      And every time its altered, a new version is kept. This would allow users to go back to old version if required.

      This is about the only decent idea you've managed to come up with. Mind you, similar functionality is already available via System Restore points - but I imagine people like you automatically turn them off because you "don't like stuff going on behind your back".

      A user could tell the OS to lock a registry so that nothing can alter it

      Like they could now with ACLs, you mean ?

      The system registry could never be altered by any application.

      Regedit ? Control Panel ? How about applications that want to make system level changes for legitimate reasons ?

      Requests to modify would require the root password entered by the user. Every time.

      Because I'm sure the user will understand the implications of modifying arbitrary registry keys and will give nearly two full seconds' careful and considered thought before typing in their password.

  10. After using OSX, GNU/Linux, and Windows... by DaCool42 · · Score: 4, Interesting

    I've learned that there really is something powerful about combining several generalized apps (like in bash with pipes and such). Apple seems to be catching on to this idea more with their "Automator" in the newest OSX. Windows still doesn't let you combine things in this way, so the solution to most problems is to download (or write) another specialized program.

    --

    ----
    All of whose base are belong to the what-now?