Haiku Alpha now very close

As you might have read on OSNews, the alpha version of Haiku is closer than ever. Only two bugs remain to be fixed and we will have the first stable version of our beloved operating system.

I can't begin to tell you how happy I am. Although I have not been really active in the community for more than 3 years, I have always kept in touch with the latest developments. And this one gives me hope that soon, I can get back on the BeOS/Haiku wagon.

The Haiku community has my congratulations and my gratitude for their work.

Some news

I am now preparing 2 systems to continue development of BeOS Max on them. Work will restart this week. Main focus will be to add some new drivers and test them and also finish the Max Updater so I can update and install new software with it.

As you may know the CUPS port has been restarted by other developers, so I am not going to go on with it (I got to a dead end anyways). Hopefully my few notes on that have helped them.

I know I am going to get bashed, but..

I know it has been four months since I had anything to report, and I still don't have anything to report. I haven't had time to check even one line of code for CUPS or an update for Max since my real life (tm) work is keeping me busy with quite a few 12 hour work days.

Thinks aren't looking good either. I am going to try to find some time to get the work I have done on the Internet, so anyone interested can go on from there, since I don't know if I am going to have time in a month or in a year from now.

I am not giving up, I just don't know when I can get back to everything.

I will try to keep you updated. Thanks for being patient.

no progress yet

Tomorrow is my last day at work for this month, so I have a lot to do and I couldn't get any work done on BeOS. However, this weekend will be spend mostly working on CUPS, so I expect to have something to report on Monday. If not, then that means I have a new girlfriend... (working on it) and threading priority has changed!!!

libxml2

Yesterday I ported the latest libxml2, version 2.6.27. This should help in other projects that require it. I have mostly been studying the inner workings of CUPS and will do so during the weekend. I will blog any progress made.

and in comes the usual... "I wouldn't do it like this. It is the wrong way."

Well, I am trying to do it like this just to see if it can be done. I don't even know if I can do it. So please, either help with real suggestions or work, or get out of the way...

This is a general response for the people that would like it done in another way. Well, everyone is entitled to their opinion, but I would prefer if they stated it as their opinion and not as the teachings of the holy bible...!!!

CUPS requirements so far...

So far I have found the following requirements for the various parts of CUPS:

Perl 5.8.0
file 3.31
automake 1.8
autoconf 2.58
libm 5.4
libxml 2.6.2

Now:
perl 5.8.0 exists.
file 3.31 seems to be ported with no problems.
autoconf 2.58 I had to do a dirty port as BeOS R5 doesn't support flock
automake 1.8 can be ported after autoconf 2.58 (done)
libm 5.4 needs to be ported. I haven't yet managed to create the static library, however using trig.h PI definitions from Haiku makes it compile. Any help?
libxml 2.6.2 was a surprise as it seems to be ported with no problems. At the time of this post my dev system is running the libxml tests.

Update: Found out that ginstall is also needed. No time to check on that now. Be back tomorrow.

Now this IS cross platform!

BeDMDiceRoller running on Suse Linux with Flyab, without any change. Great work Team Maui

More on CUPS

I managed to compile the cups 1.1.23 source tree, but no drivers yet and it gives a Missing library error on startup. Probably I need to compile the rest of the cups tree as it contains the drivers and a database driven system for intergrating the drivers, called Foomatic.

To build Foomatic, I need the unix utility called File. I downloaded it and started porting that too. The newest version however did not compile due to the usage of mmap. So I downloaded an older version and it compiled after tampering a little with its code.

Tomorrow I will look into compiling the rest of the tree. But first I need to read a bit more on the architecture of CUPS.

CUPS

Today I restarted work on BeOS Max V4 and I had an epiphany...while talking to my BeOS friends on IRC who suggested it :-). Why not try and see what would be involved in porting CUPS to BeOS R5. It has already been ported to Zeta with success, so I downloaded the source code and started playing with it.

3 hours into fiddling with the source code, I have managed to compile more than 70% of CUPS. A few functions are missing from BeOS, but I believe I can have it fully compiled. Having it work though is another issue!

Max update

It has been almost 3 weeks since my last post on the blog. I am currently under a lot of pressure from work, so it is going slowly. I won't be able to have a final release next month, so I will be giving a second Beta.

This beta will include the Haiku usb stack as an option. If it is detected that your system isn't supported by the R5 USB stack, the Haiku stack will be automatically installed, otherwise you will have to select it.

I will be keeping you up to date on progress. Thanks.

Disappointed with Bernd

Although I consider Bernd Korz a friend since I have met him personally at Cebit a few years ago and I liked him, I am disappointed by his actions. Without offering any apology or explanation for the legal status of Zeta he closed his blog and he is not to be seen in the Zeta/BeOS (now only BeOS) community. He should at least do that.

Don't get me wrong, this is not about the legal case, for which if ACCESS didn't care why should I, but it is about consistency. I have made some mistakes too, but I like to admit them so I can move on, or if I am misunderstood I like to give explanations.

Well, Bernd, if you are reading this, we are waiting. Please respond either way.

Haiku Distroes

I put a piece of news yesterday on BeOSMax.org, that work on WalterOS, a haiku distro is starting. This seems to have caused some debate on OSNews and on the OpenBeOS mailing list.

So to explain:

I personally believe that there should be rules when making a distro outside the rules of MIT, GPL and usage of brand names like the Haiku name. Those rules should be set by the team that controls the main system repository, and this is Haiku. All others should follow those rules.

I have no intention what so ever to cause any trouble for Haiku or to make an obscure claim to fame. Haiku has already said that they wouldn't want a distro before R1 is out and I agree with them completely. Also I don't have to be the one leading the distro that comes out if it is one, or even one of the distroes that come out. I just want to be a part of it. This is not about leaders it is about working to create.

Back in the early days of BeUnited, in September 2002, I had an offer to work with Simon Gauvin on a distro of Haiku for BeUnited and I accepted it. BeUnited however dissolved so I guess we can't do that. But I can surely work with others.

I can even say I had 2 offers of people wanting to work with me (in just under 24 hours!!). One of them was planning his own distro but thought it would be better if more people united.

So, this stage is just planning. It is not implementing anything, and all plans will be discussed with the community.

The only thing I hold for my self is the joy of participating in the effort to bring down Microsoft...

If you have any questions feel free to ask them. I will try to answer them all.

going on....

For those who haven't followed what has been happening the last few days you can read this: Zeta Officially a Dead Fish

and this: ACCESS Confirms Zeta BeOS Version Infringing

I have talked with David Schlesinger, the Director of Open Source Technologies at ACCESS on the matter of BeOS Max, and he told me that he sees no reason to interfere with the project. This is not official support, but it is the best we can do for now.

A change is comming...

You can actually feel it in the wind. A sudden change in the Force as a Star Wars fun might say, or the wake of a time-space distortion if you prefer Star Trek.

The demise of Zeta is a sad day for the community. I personally wish it was able to find new funding, but it seems 3 in a row is too much. So what change am I talking about?

Well, I think Haiku will and should try to get the rights to the code written for Zeta, specifically drivers. Imagine R1 getting out with SATA support. Perhaps it is too early yet, but I think an exploratory contact should be made with Bernd.

Yabbing on...

I have worked a bit with Yab, and I found it extremely easy to program with and quite powerful. Now, of course you won't create a Diablo killer with it, or an office suite, but certain software and utilities can easily be created with yab.

A few features that should be added for a version 2 would be SQL support (SQLite and Postgres for starters), attribute and query support (after all we are talking about BeOS) and perhaps basic network support. What do you think?

stop the presses

I here by stop every activity related to BeOS. My reasons are explained here. Thank you.

Haiku or what?

On the previous post a discussion has started that seems interesting. Instead of answering in comments I thought it would be an interesting article. So here we go:

Emil, posted a comment that proposes a few things and a short reply by two more bloggers followed (read them here). So here is my response:

Development of Haiku is now in a very good stage, and being a clean room implementation of BeOS R5 has advantages that I am afraid buying the code from Access won't have.

To elaborate, even if Access sold the rights to the code, they wouldn't allow certain patents to go with it. So open sourcing the code would be illegal even if someone owned it. Only Access can open source it and I don't think it will or will sell it for any amount under the 11 million US$ it cost Palm.

Even with 1/100th of that amount, haiku can go far and finish at least R1. So perhaps calling for a Haiku donation round would be better. We should address that question to Michael Phipps.

To start, I have already given about 200$ to the Haiku project in one way or another (mostly with direct donations) but I will be willing to put out another 200$ if an organized donation round starts.

Continuity of magnussoft Zeta

It seems that Magnussoft was expecting Zeta to bring profit very quickly. I don't know how much they have invested, but the short period they had financed Zeta for wasn't enough to bring more sales. Features like Multiuser and a couple of games are fine, but it needs more to get out there.

Fortunately they appear to be listening to the community as the post on the Zeta-OS site says. We will wait and see. Perhaps the horse isn't dead yet. Perhaps it is just a tired little pony :-)

BeOS Max gets Wireless

Today I added some wireless drivers and tested them with real hardware I happened to have. Guess what (it's easy). They work great. So for now, Intel PRO Wireless 2100 and 2200BG are supported. The second seems to be found on a number of laptops, and mine has it too.

Intel PRO Wireless 2200BG work is closed source, so I have sent an email asking for permission from Patrick Lafarguette to use it in Max. Intel PRO Wireless 2100 is open sourced and it is the work of Michael Lotz.