View Single Post
Old 04 April 2015, 15:22   #12
NetBSD developer
Join Date: May 2012
Location: Warsaw, Poland
Posts: 353
I'm the owner of Sakura, some time ago I expressed my opinion about openness in the thread about Sonnet on AmigaWorld. It pretty much sums up my attitude.

When I produced the PCMCIA SRAM boards, did the fact that project is open impact sales anyhow? I highly doubt that.

Proponents of open source need to realise one important thing. Just opening everything does not magically solve the problems of our community. Sure, it often makes things easier, but the most important thing is getting people involved.

Unfortunately, there are a lot of people that are only good with writing stuff on forums, bitching and pretty much nothing else. But I'm not making the projects open to appeal these kind of people. What I care about is other developers, who (if the project was closed) in the future might have to reinvent the wheel. The most scarce resource in any computer community is the developer. Don't let this resource go to waste with constant need for reimplementation of existing things (or just figuring out how these things work)! It is also important for educational reasons, if we want our platform to survive, the new developers they need some examples.

In various Amiga-related projects I already had to reverse engineer, reimplement and rewrite existing things. I consider it a waste of time to write again something that already is out there.

Some hardware vendors are even refusing to release programming documentation. This is pathological situation. Fortunately a lot of new products are coming from Jens and he is doing a good job supporting independent developers with necessary information. I wrote NetBSD drivers for ACA500, X-Surf 100 and a couple of other products. It was a relatively painless process, compared to writing Mediator drivers where reverse engineering was necessary.

In the Amiga community there is some weird assumption that open source means the project must be non-commercial. At dayjob I work very closely with a major open source company. The amount of money they make is insane, and honestly I think it's more than all the Amiga companies make together... Could these methods be applied in Amiga community? I don't know, Sakura is a kind of experiment that will show it.

We need a commercial model to accelerate development and close the gap with mainstream platforms.
I think, what the closed model is currently doing, it's not accelerating but slowing down the development. We need to get out of this 90s mentality. Some project leaders need to realise that this closed-secretive approach makes it very difficult for other people to participate in development of existing projects.

Or even makes it difficult to create new things. Let's imagine I wanted to create a new graphics card for classic Amiga. Fine, I can make the hardware. But then, it needs to be programmed. The only viable OS3 RTG systems are Picasso96 and CGX. Can I just develop the drivers? No. Cause siikretz and driver development kit is not publicly available. Also see what Ratte wrote here. Crap, I don't have the time to reinvent the RTG wheel now, guess that graphics card project will be postponed.

PCI programming "standards" on OS3 are similar mess. Amiga developers managed to create a dozen of incompatible APIs through the years. All of them closed source (excluding the AROS one, but well, it needs AROS). OpenPCI is so "open" that it provides development kit publicly. At least that, but it's not nearly enough for more advanced projects anyway, like Sonnet. It's possible we'll have to reimplement PCI wheel for Sonnet (more about that here).

Concluding, I believe Amiga projects could benefit a lot from more open approach.
strim is offline  
Page generated in 0.04687 seconds with 9 queries