Which is much needed… as there are very few real world applications on Haiku. So props for actually building something on Haiku!
While people often disagree you have the high ground here and in the end since you are actually building something you’ll either prove men dead wrong or come around to my point of view so its all good! In any case have fun!
To attract users and developers, think “platform” - not just “OS”
If you look at the popularity of Windows and Linux, it has occurred not only because they made it easy for the public to use, but also because they created platforms for doing other things with the OS - which attracted users and developers. For example, in order to get wide usage of its platform, Linux ppl utilised the following features:
In comparison, hobby OS’s just promote their own unique “internal” features:
Unfortunately, the computer-using public isn’t interested in “internal” features. They want a PLATFORM that they can use to do “other” things with, and to have some sort of cross-compatability-functionality with other OS’s / apps / filesystems.
So when you look at Haiku as a platform for doing other things, it only has one main feature:
Other useful basics it has are QT, Webkit, C++, GCC, bash, FAT, etc. But it doesn’t have the things necessary to make it a useful platform, like programming languages, virtual machines, office software, skill teaching, etc.
So it seems that if a hobby OS wants to progress by attracting users and developers, it needs to forget about its internal features, and concentrate on developing the OS as a generic platform. Otherwise, it will always remain a hobby OS with quaint, niche features that doesn’t attract many users. A suitable development mantra for an emerging OS, like Haiku for example, could be, “Ask not what your OS can do for you, but what you can do WITH your OS”.
I don’t agree here. Wether the features are “internal” or “external” is just a question of how the project is organized. Haiku being a rather small to medium sized project, it makes sense that a lot of features get into the OS itself, rather than external apps.
To quote some of our own examples, what’s the difference between Wine in Linux and the ability to run windows apps in ReactOS? From the “platform” point of view, none at all. It just happens that ReactOS has this at the core of what they build.
So let’s see the features in the Haiku platform, first from your Linux list:
Let’s add some of the internal stuff that makes the platform useful:
I’m sure other people will add a lot of what they do with Haiku here (or in that other “how do you use haiku” topic).
Then, above this is the way things are marketted to our potential audience. Which is mostly as you quoted it and has seen little changes since the BeOS days: “BeOS compatible OS, database-like filesystem, multitasking, stack and tile windows, 32 desktops, replicants”
As you noticed this is a very technical approach to the thing. It does not talk in terms os things useful to users, but useful to developers. The reason for this is that Haiku is still in beta and I wouldn’t recommend it to general users just yet (for example, I wouldn’t set it up on my parent’s computer, or the free access computer in the music school I help running).
When I try to get people to set up Haiku, I’m often targetting developers or other people who have some technical abilities, and who could help Haiku in its current state. When we get beta 1 out, and on the way to R1, we will need to switch our marketing to target end-users. But for this, we will need the help of application developers to make cool things with Haiku, and then we can show off everything you can do with the OS and the awesome apps.
When I look at the description you made for other “hobby” systems, I see that their target is even more restricted. They want people interested in hacking into the OS itself. They advertise internal kernel design features or the like. To me it would mean that these project are not even trying to get application developers in as Haiku is. They are just trying to get people to develop the OS itself. Which seems about right for most of them as I know them. ReactOS is a special case as their goal is to just reuse an existing application ecosystem. Hence, indeed all the work is on the OS side.
None of these feature requests really matter if there isn’t an alpha release produced at least every 6 months to great fanfare.
To most people, they would look at the web page and interpret it as no work having been done for years. They would not see the busy little bees behind the scenes porting software and adding features.
RELEASE AN ALPHA NOW.
Or hide last alpha release as old alpha, and show instead rolling alpha of nightly/development build in Get Haiku! | Haiku Project
I wonder which part of the home page makes you think that. The Get Haiku page has a big red box telling people to get a nightly build.
We are on the way to a beta (the time for alpha is passed already!). We are getting close to it but there are always a few last bugs to iron out and things to sort out.
And, we are not in an hurry anyway. In the current state, it sounds fine (to me) that Haiku only attract people who are interested enough to get past the website homepage get to try Haiku. As Be once said, “underpromise and overdeliver”.
We could raise the hype about a new release and everything, but if all we get in our user’s hands is a boot prompt where the “boot to desktop” button isn’t even visible (check it in French locale if you don’t believe me), and if they get past that, a crashy web browser, is it worth it?
I remember times of Windows 3.1 and 95 or 98 when "crashy” was not obstacle for system popularity and use.
Haiku is good enough to start probing wider user community.
yes… totally agree… an OS will never be finished and ready…
the decision to have package management for the Haiku Beta was 2010 a big, big, mistake…
anyway Haiku is on a very good way…
…Windows 3.1 and 95 or 98 when "crashy” was not obstacle for system popularity and use.
Yeah, i’m Argentinian and, my pc is an All In One with an Intel atom and 3GB of ram, i can barely run Haiku, and, I’m using FreeBSD only because of how light it is
Have you tried a nightly build? It might run better due to the continued work put into Haiku.