In contrary to other OS’s Haiku’s team is open to the questions from the users (hence the many responces here and guidence given, pointers on where to report issues …), their focus may be in feilds they have their own interest in (and expected as volenteers), but in the end someone could step up (old or new user) to fix things reported by the users (we all). 
I think that ‘More users brings more developers’ has to be taken carefully when it comes to Haiku. What are we talking about? People who will make a piece of software that run on Haiku or people who will help on Haiku development? You don’t code for an OS the same way that you code for an app.
If you develop an app, all you need to know are few rules (like 1 app = 1 file or 1 folder for settings) and how behave package management. You can write in the language you want with your own coding style. Once your thing’s done, you can go elsewhere for another adventure. Majority may not even know about Haiku and let others do the porting. So, at the end, what’s the gain? More project’s owners accepting patches because they heard about Haiku?
When coding for the OS itself, you need to understand the behaviour at another level, to use a specific language and to respect guidelines, to document your code so it can be maintained later but, more important, you need to respect the philosophy of the OS. It takes involvement and way more time to get there. If you have already some skills, you can provide patches for few things meanwhile but, without efforts there’s no way to do more. It makes that kind of developer a bit harder to find and Haiku is lucky to have few of them.
4 Likes
My experience writing apps for Haiku is that I always end p fixing or improving things in the OS. I guess this is what makes this a bit less frustrating than other OS, where, when I find a limitation of the system, I can do nothing about it.
I think several other people started with writing an app and ended up similarly fixing things in Haiku.
8 Likes
By writing applications, you also get new requirements and new errors or better possible additions become visible.
Unfortunately, more is written than natively, which means that the whole thing doesn’t run so well. The test of the ports is done by the user, which I think the developers or special test users have done on the original systems. As a result, many errors only come to light later.
1 Like
The user groups used to have lists of people from the forums, so you could see who came from where
Also webpositive identifies as MacOSX according to user agent sites and html5test.com (obviously done for compatibility reasons), thereby skewing the data.
Perhaps a good metric is number of people updating a specific package.
1 Like
User engagement will increase naturally as the OS improves and people find it more useful.
For example, Linux used to be my first choice when copying lots of data off of any filesystem, whether it was ntfs, fat, or extfs. Even Windows (Vista and newer) with teracopy or fastcopy32 doesn’t have the speed of Linux.
When I was doing PC repairs for customers, I used ultimate boot CD for Windows to copy and recover data and it was fast and lightweight. However, because it’s based on Windows XP, it got harder to add drivers to it that supports newer chipsets. Now my first choice is to boot up haiku. If it doesn’t boot, then I will boot a Linux rescue disc.
2 Likes
kernel+webposite
this is the first important thing of all.
after dd current iso into one usb driver, haiku can use the free space of the usb as extra partition with trying mode.
it is beta stage now.so, haiku should have the complete usb-live-mode. any people can try haiku. if feel good, they also install haiku into their computer with the usb-driver.（special the UEFI thing）
virtual machine， it is first important with alpha stage , but not so important with beta stage.
add a ranking part（the application name which is most popular in haikudepot） at haiku homepage.
it is a glory for application programmer.
am i right?
I’ve brought up making an anonymous hardware + crash info collection system before… but it tends to be an unpopular idea. People who use Haiku seem to lean more privacy focused.
With that said, HaikuDepot does seem to collect user stats. I’ve exported them for your viewing:
Desktop Versions Traffic
0.0.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
0.0.6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11615
API Generation
v1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10
v2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12851
Multi-/Single-page
Single . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3127
Multi . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92631
Bulk Downloads
/__reference/all-** . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
/__repository/all-** . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .362
/__repository/all-** . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .362
/__pkg/all-** . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
/__pkgicon/all-** . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0
Screenshots
/__pkgscreenshot/** . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3805
I have no idea how to interpret these, or what the timeframe is.
Slightly more interesting though… there are 2,768 user accounts in HaikuDepot. Those 2,768 users took the time to register (you can use HaikuDepot without an account)
2 Likes
I’d bet it 10% of users who have a account, see how many distinct ip addresses exist. ie regional ip adress blocks, prob be best way to estimate
Being spied on is of course a very serious issue.
But having at least a rough estimate of the number of serious users (people who regularly update their system) helps developers of Haiku how well the OS is doing.
1 Like
Just counting the number of users is a very long way from spying. And we ought to have a rough idea.
Do we know how many people downloaded R1B4?
If we knew the number of users, even approximately, it would be interesting to see how the number rose or fell over time, and especially as the system becomes more and more usable. By the time R1 comes out I would hope (and expect) to see a large increase in the number of users.
Bear in mind that a larger number of users could be expected to contribute more funds, and widen the developer pool. Having more users isn’t just for boasting about, it will have tangible benefits.
1 Like