5 ms·

Ahh, Apple. Do you realize that your only value for most of my friends is that your OS is actually a NIX beast? I wish someone had enough money to come up with

by maurycy 17y ago

Ahh, Apple. Do you realize that your only value for most of my friends is that your OS is actually a NIX beast? I wish someone had enough money to come up with decent UI on Linux (pardon me but GNOME looks and works horrible), better typography and write some general, and widely working, framework for supporting Windows drivers. It should not take more than $20-30m I think. Ubuntu, unfortunately, is not aggressive enough in those, critical, areas.

wtallis 17y ago
Did you miss the reports about the estimated value of the Linux kernel being about $1.4 billion? What makes you think that completely overhauling GUIs on Linux to create a new desktop environment and adding binary compatibility with a completely different (closed-source!) OS would cost less than 2% of what it would take to write just the kernel? Developing a UI framework with comparable technical capabilities to what OS X provides might be doable starting from the best of what's available at the moment given your budget of $20-30m, but building a cohesive desktop environment on top of that would be an extremely huge undertaking.
- xtho 17y ago
  It will be 2 billions in ten years when linux + new gui has conquered the desktop (or what is left of it) and mobile devices. The difference between 2 billions and $20-30m (or whatever imaginary sum you suggest) is called profit (or rather some share of it).
- maurycy 17y ago
  It does not require major overhaul of the kernel. I agree that Windows' drivers compatibility is a hard task but I imagine there might be some hack. For instance, you could give up some speed, using virtualization. If such hack is not available (patents?), it is always possible to come up with a set of rock solid supported devices, might be with some cooperation from hardware vendors. It is exactly how Apple solved this problem, and it seems that regular people are happy to accept such restrictions, as long as everything works great. Right now, the problem is that if you install Linux on your computer, the official story is that "Linux supports more hardware than any other OS". Last time I checked (6 months ago), though, hibernation is still problematic, and you have to Google to ensure that your printer will actually work. As for the UI. To be honest, it is not a technical problem. There will be some technical work involved but it does not matter that much. Linux development focuses too much on technicalities. (I write it as software developer) The biggest issue are fonts. Every time I launch Linux, my eyes hurt. Few hours of work are required to get fonts right and they are still not that good as on OS X or Windows. There is a lot of typography know how available right now, and I bet you'd easily find a top notch typography agency or folks to solve this problem. The second issue is design and basic usability. You do not have to rewrite everything, initally. The things I mean is fixing many small issues, such as mismatched padding everywhere, and UI inconsistencies, as well as running serious usability studies to find how regular people cope with the basic tasks. I think it would be the best to pick XFCE for such improvements. Exactly like big corporations picked WebKit instead of Mozilla's engine. XFCE is modular, still relatively small, not that bloated as GNOME, and you could fork it without too much hassle. Also, it helps a lot that most of the world switches to the web. You longer need to write all possible tools, except good integration with the web. I dream about something like transparent Google Gears integrated into the WM. Paradoxically, I'd start work with porting TextMate to Linux. VIM or emacs are a bit too hardcore for regular developers, and there is no good editor on Linux for dynamic languages, which is light weight, nice and works perfectly within a minute, or two. Developers are crucial part of the infrastructure and Linux does not attract developers with eye for usability. I agree that I might underestimate the costs. I don't think, though, that it is going to take more than $50m, and money is not that big problem, actually. The biggest problem is to realize that current Linux experience sucks a lot for people who do not want to spend their valuable time tweaking settings. The improvement is enormous but it is merely a start. As for multi billion corporations that started and failed, it is not true. Right now, Linux's development is focused on getting servers right. The corporations succeed in this area. RHEL is de facto corporate standard. No one relatively big, except medium-sized Canonical, cares about your mother's experience. BTW. To be honest, I was very surprised it is only $1.4 billion when I read it. Given the scope of the project, I expected few times more. But this is me.
protomyth 17y ago
It seems like that would have been the killer Linux differentiator for some distribution. Design a decent UI that isn't Windows, OS X, etc. UI's are hard, but it seems like there are more ways to do things out there that aren't being explored.
- stcredzero 17y ago
  A small and light form factor tablet with a touch interface and a way to hide away a EVDO USB modem would probably sell very well.
  - protomyth 17y ago
    I would say skip the big stuff and just do a netbook or netop that is rock solid and requires no user expert knowledge.
    - stcredzero 17y ago
      Don't see you can't do this completely until you go to a completely touch interface and put up an App Store.
potatolicious 17y ago
> "I wish someone had enough money" Multi-billion dollar corporations have tried and failed, I honestly don't think this is a problem of money.
freetard 17y ago
I use Linux as my main desktop and I'm a professional worker. Works just fine, I do hate the Windows and Mac UIs though, so I can understand people have different tastes. I try not to make general statements about other UIs, just different tastes and opinions. Glad you're happy with Mac or whatever it is you're using.

by maurycy 17y ago

wtallis 17y ago

Did you miss the reports about the estimated value of the Linux kernel being about $1.4 billion? What makes you think that completely overhauling GUIs on Linux to create a new desktop environment and adding binary compatibility with a completely different (closed-source!) OS would cost less than 2% of what it would take to write just the kernel? Developing a UI framework with comparable technical capabilities to what OS X provides might be doable starting from the best of what's available at the moment given your budget of $20-30m, but building a cohesive desktop environment on top of that would be an extremely huge undertaking.

xtho 17y ago
It will be 2 billions in ten years when linux + new gui has conquered the desktop (or what is left of it) and mobile devices. The difference between 2 billions and $20-30m (or whatever imaginary sum you suggest) is called profit (or rather some share of it).
maurycy 17y ago
It does not require major overhaul of the kernel. I agree that Windows' drivers compatibility is a hard task but I imagine there might be some hack. For instance, you could give up some speed, using virtualization. If such hack is not available (patents?), it is always possible to come up with a set of rock solid supported devices, might be with some cooperation from hardware vendors. It is exactly how Apple solved this problem, and it seems that regular people are happy to accept such restrictions, as long as everything works great. Right now, the problem is that if you install Linux on your computer, the official story is that "Linux supports more hardware than any other OS". Last time I checked (6 months ago), though, hibernation is still problematic, and you have to Google to ensure that your printer will actually work. As for the UI. To be honest, it is not a technical problem. There will be some technical work involved but it does not matter that much. Linux development focuses too much on technicalities. (I write it as software developer) The biggest issue are fonts. Every time I launch Linux, my eyes hurt. Few hours of work are required to get fonts right and they are still not that good as on OS X or Windows. There is a lot of typography know how available right now, and I bet you'd easily find a top notch typography agency or folks to solve this problem. The second issue is design and basic usability. You do not have to rewrite everything, initally. The things I mean is fixing many small issues, such as mismatched padding everywhere, and UI inconsistencies, as well as running serious usability studies to find how regular people cope with the basic tasks. I think it would be the best to pick XFCE for such improvements. Exactly like big corporations picked WebKit instead of Mozilla's engine. XFCE is modular, still relatively small, not that bloated as GNOME, and you could fork it without too much hassle. Also, it helps a lot that most of the world switches to the web. You longer need to write all possible tools, except good integration with the web. I dream about something like transparent Google Gears integrated into the WM. Paradoxically, I'd start work with porting TextMate to Linux. VIM or emacs are a bit too hardcore for regular developers, and there is no good editor on Linux for dynamic languages, which is light weight, nice and works perfectly within a minute, or two. Developers are crucial part of the infrastructure and Linux does not attract developers with eye for usability. I agree that I might underestimate the costs. I don't think, though, that it is going to take more than $50m, and money is not that big problem, actually. The biggest problem is to realize that current Linux experience sucks a lot for people who do not want to spend their valuable time tweaking settings. The improvement is enormous but it is merely a start. As for multi billion corporations that started and failed, it is not true. Right now, Linux's development is focused on getting servers right. The corporations succeed in this area. RHEL is de facto corporate standard. No one relatively big, except medium-sized Canonical, cares about your mother's experience. BTW. To be honest, I was very surprised it is only $1.4 billion when I read it. Given the scope of the project, I expected few times more. But this is me.

protomyth 17y ago

It seems like that would have been the killer Linux differentiator for some distribution. Design a decent UI that isn't Windows, OS X, etc. UI's are hard, but it seems like there are more ways to do things out there that aren't being explored.

stcredzero 17y ago
A small and light form factor tablet with a touch interface and a way to hide away a EVDO USB modem would probably sell very well.
- protomyth 17y ago
  I would say skip the big stuff and just do a netbook or netop that is rock solid and requires no user expert knowledge.
  - stcredzero 17y ago
    Don't see you can't do this completely until you go to a completely touch interface and put up an App Store.

potatolicious 17y ago

> "I wish someone had enough money" Multi-billion dollar corporations have tried and failed, I honestly don't think this is a problem of money.

freetard 17y ago

I use Linux as my main desktop and I'm a professional worker. Works just fine, I do hate the Windows and Mac UIs though, so I can understand people have different tastes. I try not to make general statements about other UIs, just different tastes and opinions. Glad you're happy with Mac or whatever it is you're using.