Haiku Beta Discussion

So long as you can run BeOS apps without issue on Haiku (regardless of 32-bit or 64-bit), which is the biggest reason for Haiku’s mere existence. I have to assume that a good number of folks in the Haiku community are here because of BeOS and the assumption that Haiku will carry its legacy forward concerning applications, etc…

If the original intended mission of the project has changed, then that needs to be well communicated so folks (like myself) can decide if it’s worth sticking around or not.

Haiku x86_64 can’t build itself nowadays. An old revision can. Devs arguing over what direction Haiku should go in. R1 possibly never being reached admitted to. The ship is slowly tearing itself apart, because there is no captain to helm it in the storm.

Under Steve Jobs, Apple was a very focused machine (if not always 100% winning on everything). Without him (then and now), fracturing occurred.

We need one man to lead. To say, THIS is how we’re going to do things. Right or wrong, at least there is focus and purpose. Of course, I still believe, as I did many years ago, when Phipps was OFL (Our Fearless Leader) that the cart was before the horse, which is WHY we are where we’re at, today, but it is what it is.

I propose a way out of the mire. I propose extreme focus and a razor-sharp goal. But it comes at an extreme cost that few, if any, are willing to consider. The status quo will only cause further degradation. 20 yrs. between releases? How about NEVER? The goal of “BeOS R5” was a nice idea, but its growing more and more untenable, since it was passed, so long ago, nobody can even remember when it occurred. 32-bit apps in a 64-bit world? Pointless.

But open-source, for all its glory of available man-power, also has a glaring weakness. Lack of leadership, because everyone works for free and therefore say “I wanna do whatever I wanna do”. Not what NEEDS to be done, for the good of the project, but whatever makes the individual dev happy. “Cart before the horse”, as mentioned above. And now the cart and horse are so horribly tangled, it’s almost time to shoot the horse and dismantle the cart for firewood.

Was this end worth the journey to get here? Really?

1 Like

I like this proposed path!

It is likely that, within less than five years, 32-bit systems will belong to the realm of vintage computers.

On the other hand, some 32-bit BeOS compatible support will need to be provided for the BeOS/32-bit applications which are much beloved and will never transit to 64-bit because the source code has been lost. Having the compatibility support via a Haiku specific BeOS compatible virtual machine (BeVM?) rather than a bridging layer may probability be advantageous. In this fashion, Haiku could evolve as needed with much less constraints from the legacy APIs.

The legacy and ideas should be carryed on. BeOS was a fast, modern, had a good object orientated API that developers like, had little legacy feature bloat from previous OSes.

That’s what was attractive about BeOS and what I hope Haiku aspires to be.

Being binary compatible BeOS is history, not legacy. It is technical debt.

I disagree that being able to run BeOS applications is the reason Haiku exists. Haiku was created to continue what Be had started with BeOS. One of the goals was binary compatibility to aid in the transition from BeOS to Haiku. Running BeOS programs was an important steping stone, but not the final goal.

My hope is that Haiku should not be a tomb to run dusty old BeOS R5 programs. That’s what virtual machines are for. Haiku should strive to be an outstanding and innovative operating system, based on ideas of BeOS.

3 Likes

It may NEVER reach R1… and that’s ok?!? Have you lost all focus, man? I believe the longer this goes on, the more devs will start caring less and less about the end goal and just what tickles their fancy at the moment. Of course, that’s been the problem all along, hasn’t it? Cart before the horse. It was never about what the project really NEEDED, to progress correctly, but what each individual WANTED to do for the project, regardless of progress.

But it’s too late to change course, isn’t it? It’s plow forward or give up. At least on this track.

But another track lies just off to the side… so radical and different… no one DARES to consider it.

2 Likes

why do we actually talk in two forum posts about the same topic at the same time?

1 Like

It is a sort of haikuiš supertreading.

3 Likes

Like the old BeOS phrase: “Because one thread is not enough” :slight_smile:

2 Likes

Because I try to “spark” the conversation, when I see points that converge. When I see discontent… spark! When I see issues mentioned I believe should not exist… spark! When I see someone mention that if R1 was never achieved, that would be ok… spark!

Eventually (I hope) the fuse will be lit and (eventually) a massive BOOM! will be heard across the Haiku-sphere, as everyone stares agape, and says, “How’d he do that?!?” You know… like Steve Jobs and JLG did years ago…

Many will see these “sparks” as an annoying flash of light and noise. You know, like an empty Bic lighter. But it only takes a well placed spark… to light a fuse… to ignite a… explosive result, to the joy of all Haikudom!

If it can be done. But, if it can, will we ever know, if no one dares to try?

2 Likes

AHAHAHAHA! Good one!

I know that Adrien (Pulkomandy) is a Haiku developer, but the English word “we” is normally used to mean a group of people, my question is not about who Adrien is, but which group of people (if any) he’s claiming to represent with these views and why he thinks he’s right to do so.

1 Like

Contrary to many forum posters that insist to know what “we” have to do to succeed, Adrien is one of the most competent people to talk about that. And one who actually has been working hard for many years and is involved enough that he can actually speak for the project.
All this has been discussed plenty of times on the dev mailing list. I think he presents the opinion of the majority of the Devs from those discussions.

5 Likes

I agree with that consensus. I think it was obvious, when I read his post. “We” is all developers who are working on Haiku, period. Any of us (myself incl.) who are not part of that ilk (devs), have no room to question the use of “we”. We (end users) have our own club to belong to. That gives us the right to protest decisions made, but without “them”, where would “we” be? :smiley:

1 Like

It isnt true.

If that’s not the case, I hope the other devs would say so and tell their own view on things.

But as it was mentionned, being a member of the Haiku project only happens if other people in the project think you understand the “Haiku Way”. There is no leadership and project members just happen to share a more or less common vision about the project.

As a result, I can use “we” and speak for the project when I am writing things that I feel are part of the “Haiku Way” and shared with other devs. I may sometimes put too much of my own vision in it, but I know there are other devs watching and they will correct me if needed. And this is how the “Haiku Way” becomes more precise and refined over time, and how it also sometimes evolves (we will need that, because currently it stops abruptly at the R1 release).

2 Likes

If there’s a lot written people, then I’ll write if you do not mind.
Still, compatibility, nuts are necessary for me to return for patches, full compatibility with BeOS for the implementation of R1, or Beta version and there already move forward. The project is more than 17 years old, he must have been of age :slight_smile:

Keeping binary compatibility for SoundPlay? Really? Okay, let’s not improve MediaPlayer and introduce addon support or something like that to bring its functionality closer to better parts of SoundPlay. Let’s not do that, it’s better to remain in the abandonware nostalgia bubble for what’s left of the decade. That will surely bring more developers, donations and user interest to Haiku…

5 Likes

Adding a tidbit related to 32-bits and latest generations of CPUs.

Neverware,which has been putting a distribution of ChromeOS called CloudReady in 32 and 64 bit variants, has encountered issues with its most recent 32-bit beta release on Braswell and Kaby Lake based systems ( https://www.neverware.com/blogcontent/2018/2/27/update-cloudready-home-edition-v631-available-on-dev-and-beta-channels ). Essentially nothing is displayed on the screen!

Just a heads-up about potential issues with maintaining a 32-bit environment with latest generations CPUs.

In the other hand i suggest haiku should start to receive monero donations and bitcoins, is q good amount of easy donations

Not enough people know about APlayer, it seems. It already has add-ons support and a lot of other nice features, and it is open source. However there is no development effort besides the little bugfixing I could spare the time to make. Help very welcome!

1 Like