Haiku spins could help app development

Supplying a few Haiku spins with a collection of dedicated apps at release time could help encourage app development. For example, a multimedia spin, a graphic arts spin, a repair/diagnostic spin, etc. This would encourage people to try out Haiku for the apps that are included without having to worry about installing the OS and downloading apps from repositories.

If you look at Linux, it began to take off after the release of the Knoppix live CD. Because it provided many useful apps for administering computers. Since then Linux has successfully plugged the specialist LiveCD niche. Haiku could learn from this.

I used to think that was the way to go, to promote Haiku. But now package management makes it even easier. If we send out a vanilla version of Haiku, and a script to access and install apps, utilities, content from HaikuDepot, that’s all you need.

Andrew

This is the exact reason that many people in the conventional markets won’t use operating systems like Haiku… as soon as they hear the words ’ configure yourself ’ or ‘package manager’ they are like “What?” I know i was when i first saw linux.

I have to agree with syd.

Spins are the way to go.

Loads of people , some who don’t even like computers, have tried or even heard of Ubuntu. (even though it’s a pile of pants) people know it. - This is because it was a flavour of linux that was specifically designed to focus on User Friendliness.

As soon as you start talking to normal people about ‘vanilla’ and ‘scripts’ they’re like ‘What?’

If Haiku is going to gain any traction whatsoever in it’s current state,

it needs to focus on a market, namely ultrabooks/netbooks/low end hardware revival, and be open and easy enough for people outside of the current standard demographic to like and want.

The main thing is that it sounds easy and fun to try out. Whether it’s good or not is mostly irrelavent at this stage. We , as Haiku supporters, already know that Haiku is good.

We just need to make it easy to access.

Focussing on a specific Hardware Platform, or small set of hardware platforms as well as modern standards will be the way to go i think. Though the logistics of that will be for someone else to solve.

Also, Compared to other OSs that are become ever more contemporary and user focused, Haiku looks like an awkward blast, from a redundant past.

We need a new easy to use software manager, Clear and Focused Brand Strategy, Platform Promotion amongst new developers and enthusiasts; and a hell of a lot of funding from philanthropists.

In Promotion,

My suggestion would be to push:

  • it’s not Dos, Linux or Unix Based: In fact it’s got a specialised customised Modular Kernel…
  • That runs lightning fast - boots in only 5 seconds (on old hardware!), and runs crisp and lagless, unlike most linux distros you’ve tried
  • It’s got a revolutionary Graphical User Interface/Windows Manager (Though it still has some bugs to iron out.)

That sort of thing…

and then the rest would just have to be working on compatibility and stability.

Obviously

The Real Major Thing would be to create a central, Official, standardised repositiory or hub where both software designers, developers, and other people interested in Haiku could congregate and make some meaningful contribution to the project.

Just tell them Haiku Depot is like the play store

Or change the Haiku Depot, to The Play Store LOL

I read a comment here abou making Haiku more accessible, and proposing a standard reference platform for it. I think these are quite opposite things. Currently Haiku has low requirements. You need a PC with an MMX compatible CPU (anything that can run Windows 98 or later should be good to go) and a reasonable amount of RAM (192MB).

If we started advertising one single specific platform instead, people would be like "this OS looks great, but they don't support my hardware and I'm not going to buy their specific computer to try it".

So yes, just explain to them it comes with a "depot", which is just like the play store/app store, except everything is free and has no ads. I think everyone can understand this? :)

[quote=PulkoMandy]I read a comment here abou making Haiku more accessible, and proposing a standard reference platform for it. I think these are quite opposite things. Currently Haiku has low requirements. You need a PC with an MMX compatible CPU (anything that can run Windows 98 or later should be good to go) and a reasonable amount of RAM (192MB).

If we started advertising one single specific platform instead, people would be like "this OS looks great, but they don't support my hardware and I'm not going to buy their specific computer to try it".

So yes, just explain to them it comes with a "depot", which is just like the play store/app store, except everything is free and has no ads. I think everyone can understand this? :)[/quote]

I concede to your point, I was more making a generic refrence that Haiku Depot is kind of like play store in that is has apps in it, thus the users not be burdened with terminal and package management intricacy like some linux distros.

I would have to disagree with the ‘package manager’ part of this. I think people nowadays generally understand what packages are in 2016; this isn’t 1997. Android uses APKs alongside the Play Store, and the Anniversary Update for Windows 10 will include an Ubuntu core image courtesy of Canonical - and with it, preliminary package support that the CIT/IS sort who will use said features will be familiar with. OS X (macOS now) has the Mac App Store, but developers are divided on whether to use it (http://www.leafandcore.com/?p=16292), mainly because the Mac has used stand-alone packages since 10.0 publicly. Rather, it is the App Store on iOS that has exploded and flourished, (mainly because iOS is a tightly integrated platform locked to Apple mobile hardware.) OpenSolaris and OpenIndiana use packages, as does Minix and some flavors of BSD. Several Gnu/Linux-based operating systems have excellent package managers (rpm/yum, and now dnf, is one), and with more flavors focused around app titles (think Gnome Software), it’s clear the direction that operating systems are taking and it is currently toward a centralized package manager or store; (although, imho, I truly believe containers or even a modular system might just be the next phase of things. I could be wrong.) Even though I’m uncomfortable about the Depot idea myself, I do see that it makes sense, and is necessary for Haiku to move forward to beta 1 and reach a larger audience.

As for ‘easy to access’ or ease of use, though I discovered Haiku back at alpha 2, have a copy of BeOS R5, and run ‘Dano’ on an old HP - I will humbly admit my personal favorite is the Mac. In fact, it was the UI’s closeness to Platinum and its simplicity that made me fall in love with it. Under that gorgeous simplicity, though, was power, and I would pine for that to be pulled away simply because 90% of ‘users’ don’t care to understand how to ‘configure’ something. That isn’t to say anyone’s copy of publicly available software should be deliberately exclusive to developers, either. Good software should be universal, and Free Software (as in free speech) is about empowering people to do more… but this idea means the newcomers will need to put forth at least a little effort to learn how it works and better themselves. When an interface marries the concept of beauty and power together, something like the earlier Mac OS X versions (10.1-10.6) is very possible. Today, this idea of a balance between power and simplicity is going away. In Gnu/Linux right now, a power struggle for systemd is taking place, Windows locked from inside out long ago, and now the Mac has partially locked itself already with SIP/rootless, and most of the user-base doesn’t even ‘get’ what the long-term impact of it is (btw, if you have a Mac, restart into the recovery HD and turn it off). That said, I fully agree with ease of access, as long as Haiku doesn’t go the path of Redmond.

This is a compliment, actually, as the mission of Haiku is to preserve not only the functionality, but also the look and feel of the classic BeOS and take it forward. It isn’t redundant or awkward; it is keeping history alive with each time it boots up to that look and feel. That said, I know the Be-like UI isn’t Aqua, but this is part of what Haiku is about - and there are decorators one can compile and use with Appearance preferences, or even modify a theme to their taste. I think for those that want Haiku to look like Google Aura, Gnome 3, Plasma 5, etc. it is totally understandable, but I would ask them to remember why Haiku has kept to the Be look… or if all else fails, to make your own fork or a new spin, which brings me to the very last point…

Sigh. As for an official re-spin, like how Mint and Fedora have spins with different DEs, scientific tools, etc. - sure. That’d be fairly simple. But, taking a copy of an OS project and trying to maintain it as one person is not simple as it may appear. I had this itch to fork Haiku and optimize it for netbooks since 2010 (around when netbooks were still booming), and finally tried last year. It was called Poem; it ended in me splitting it in two editions, and I finally gave up and quit the project. Whether I’ll ever start back up again, I don’t know; I actually loved Poem and it was hard to quit on it.

All this said, I wrote this lengthy essay of a post not to be annoying, but because I noticed that you and I have similar ideas, such as a Deskbar alternative and the idea Haiku should focus toward smaller machines. But I also wanted to mention all this to say that Haiku is a great system, and there is a positive side to all the suggestions mentioned so far. :slight_smile:

One of Haiku’s features is a “Database-like file system (BFS) with support for indexed metadata”

Databases have their own security measures, including firewalls, activity monitoring, auditing, two-factor authentication, privilege system, intrusion detection systems, Virtual private databases, disaster recovery programs, and forensics. Can any of this be applied to Haiku’s “database-like filesystem” to give it a unique approach to OS security? Would it be worthwhile recruiting database programmers? So that besides a media OS, a security OS could be another avenue of development considering that security is a hot topic today.

it only makes sense really. if Haiku is supposed to be an OS that focusses on ‘Personal Computing’ I want to be sure that it is only sharing data with entities that I choose it to… and my data and communication methods remain transparently ammendable and easily so…

You have to nowadays. - look at android . - perhaps instead of just setting whether an app has ‘access to features’ there could be more settings like (PUBLIC/PRIVATE/GLOBAL/LOCAL/PERSONAL) for each data item/ and application…

I’d expect it to be coded in such a way, even if it doesn’t support multi user access. Most apps have a habit of going online nowadays. It’s not so much that the OS isn’t multi-user - but it should be able to exist robustly and effectively within a multi user environment.

in my opinion…

1 Like