There are a lot of stories going round about yellowTAB and the Zeta kernel. To put an end to this, ZetaNews decided to have a closer look at the latest kernel included in the “Venture” beta of the Zeta OS. We have also tested the kernel on a wide range of new processors.
A close look at the Zeta ‘Venture’ kernel
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
I didn’t see much about the kernel, only about the tests. Did I miss something?
The issue about the kernel is that many people don’t believe yellowTab has access to the BeOS kernel source and that they’re just using patches posted on BeBits and just changing the kernel date. This report puts that issue to rest.
it seems like they have finaly modified the kernel, but why didn’t they just say “Yes we have the source code” they could have really took off when they first announced you could buy Zeta if they would have just said they had the source rather than be shady the whole friggin’ time. now every ‘geek’ assumes the company is just being shady and geeks never dig that. but bad management is what BeOS Is all about! Long live the beloved OS!
This is good.
The reason they cant just say they have the source code is that they shouldnt have the source code. Its the property of Palm and they never released it.
And how did you come by this information? So many people are saying things but I have yet to hear anyone prove themselves. It would be great if you could tell us whereby you came up with this information.
Because Palm bought Be, and never released, or even licensed the source (though, they did consider it). It’s nothing that needs to be proven, it’s what happened.
Again, this has been said over and over again but no one can point to anything saying this. If Be, Inc. (and subsequently Palm) and yellowTAB had an NDA then how are we to know otherwise? It feels so much like FUD when people say, “It’s nothing that need to be proven, it’s what happened.” I have yet to see anything which would prove yellowTAB doesn’t have their own agreement with Palm. I’d love to be proved wrong but too many people say “it just is” without any justification.
“…the reason they cant just say they have the source code is that they shouldnt have [it]…”
They obviously DO have the source code. And it is legal. Unlike PhOS, they are a formally established company, so they would have a lot to lose if they released anything illegal, anything without the proper permissions.
In the early stages of Zeta, Yellowtab probably did not have all the required permissions. However, admitting this would have been very much damaging, since it is preferable to have your customers with a reasonable doubt that something could be right or not, than just plain admitting there’s something utterly wrong.
Though it has taken a long time, Yellowtab has been working behind the scenes to get the required legal “pass”.
Yellowtab probably will not put out a press release saying “We Now Have the Legal Code”, since that would automatically confirm that they really didn’t have it before.
Of course a lot of you are gonna be repelled by this behaviour of YT.
While I don’t justify (nor attack) Yellowtab’s move on the press issue, one thing is for sure: lots of people can stop holding their breath now, and follow through together with the BeOS story.
Many readers here will disapprove YT’s manner of acting, or even say they are disgusted. However, then again, many zealots here WILL ALWAYS COMPLAIN.
People will always complain about Windows, about Mac OS X, about BeOS, and yes, even about bloody Linux.
You can never have everybody happy.
But I, for one, am glad we can finally leave this chunk of rant that dragged YT’s image down behind, and hopefully have the small BeOS community united again and get on moving forward.
Wow, I can now see where the 40% speed improvement claim came from. For AMD machines, Zeta is finally applying the correct patch files, therefore a few memset and memcopy commands are being accelerated (using SSE or whatever). That’s great. It would be interesting to know whether Zeta (kernel and all kits) have been recompiled with the newer version of the GCC compiler available in the BeOS world (Oliver Tappe’s version). This compiler gives probably the best code in the BeOS world, which should suffice until we move to GCC 4.
Wow. My old OS is hanging in there. It may be rehashed and renamed but it’s hanging in there.
I have a theory that would probably create doubt as to them having the kernel source code, but won’t share it, because all that matters is the following:
(1) YellowTab should focus on greater driver support (& support newer features & improve performance where possible).
(2) More & better applications need to be developed for BeOS & current ones be updated (for stability & enhancements).
The above 2 points are what really count. Doesn’t matter if YT has the source code or not. I’ll tell you why.
BeOS 5 & Dano where pretty stable & good performers as they were programmed. BeOS 5 had a kernel of around May 2000 & Dano around May(or Sept) 2001. (For comparison, Windows 2000 came out a little sooner & Windows XP later if I’m correct).
Yellowtab has done some improvements to the original BeOS increasing hardware compatibility mainly & creating nicer menus. YT will incorporate more & more Haiku code as it becomes available over time. Eventually (my prediction is in 2 years time) the next Zeta OS (v2) will be mostly based off of Haiku & YellowTab code.
Zeta OS is also very good as it is now (since based off of Dano).
BeOS is a very usable system and great performer (responsive). (Actually reminds me somewhat of Mac OS 8.x because of the look).
Only by creating more drivers and applications for BeOS can it become mainstream. So top 3 OSes will be (1) Windows (2) Linux (3) BeOS.
We need developers to start making or improving apps. Then Zeta/BeOS will start to take off once again. This is the only thing that will hold it back.
Most current BeOS programs range from satisfactory to very good, but I would like to see newer & improved versions come out just the same.
PS
It has all the basic applications (email, web browser, office suite) to satisfy the average computer user & be used as the main OS (for those who are tired of Windows or Linux). 🙂
Sh@t! I really dont care if they have the source or not. I admire what they are doing and think they are proving themselves to the community (or me at least)
— and I like to focus on the positives. It’s also reassuring to know that after installing an OS that it can handle my hardware without all the RAM/CPU bandaids etc..
Anyhow good job Bernd and gang. bla blablabalbalabl
i would have prefered to have YT to do that! but i beleive Frans so it does not matter.
They also talked about new “VM”. Did you test if beos still crash when not using VM at all and running out of ram? (when let say you boot with the partition so full it can create any swap).
Other interesting test would be networking speed (as it’s a different bone than 7a if i remember). Both usuall and wifi. BeOS R5 BMessage speed comparaison (as i think this is changed). Of course this would be better done at official release, but i think those thing are not gonna change.
> There are a lot of stories going round about yellowTAB
> and the Zeta kernel. To put an end to this, we decided
> to have a closer look at the latest kernel included in
> the `Venture` beta of the Zeta OS.
The question was NOT whether the Zeta kernel performed better than the unpatched BeOS R5.03 kernel.
The question was whether or not YT has access to the BeOS source code.
UNANSWERED
The secondary question was that if YT does not have access to the BeOS source code, does their patched kernel perform differently than what can be achieved by applying publicly available patches to the BeOS R5.03 kernel.
UNANSWERED
The only thing that is answered is whether Zeta Venture performs better than Zeta Neo (Venture performs significantly better).
Do you honestly think the guys at YellowTab are stupid enough to burry themselves and their endeavours in a legal quagmire?
Get over it. They are a business and one with a future and they have shown that they are making key changes to the core of BeOS to bring it upto date. For once stop bitching and just be greatful that we have Zeta and Haiku keeping the only decent Desktop OS experience alive. Windows, OS-X, and Linux still suck by comparison to BeOS 5.03 on the desktop. I hope this thing bolts from the stables and provides us with a great platform for computing and creating.
The BeOS community sure has drawn a lot of attention lately, mostly because of yellowTAB but also Haiku has made some attention due to it’s progress. Let’s hope this can help the BeOS community to grow a bit. I sure hope it will.. BeOS will never die if we dont let it, and it will gain superpowers if we realy want to.. hehe..
http://bitsofnews.com
Is it just me, or does anyone else think it’s strange that the 2Ghz Centrino outperforms (on avg) the 3Ghz P4, XP 3000, Athalon 64 3000 and all other processors presented in these benchmarks?
>The question was whether or not YT has access to the BeOS source code.
Answer=Yes. Answered by a former-well known (in beos community) yellowTab employee. They had the code but were only looking at it and not touching it. Until now that is. Either they have permission from Palm to mod the code or their lawyers are telling them it is abandonware because Palm has not protected their copyrights.
Also look here:
http://www.iltuosistema.it/news2/screencebit.jpg
>does their patched kernel perform differently than what can be achieved by applying publicly available patches to the BeOS R5.03 kernel.
Answer=Yes. That’s the answer I got from reading the article.
ha! I must be tired;
should read: well-known former yt employee
So, BeOS is alive and kicking, supported even with the Athlon 64. WOW!!
I wonder if it would work with the VIA EPIA architecture. Did anyone try? It would be sweet to run BeOS with one of them fanless machines.
You are asking for the wrong thing to be proven.
We have the generally accepted facts of what transpired a few years ago between Be and Palm, anything that disputes those facts needs to be proven, not the other way around.
The first: Yellowtab has licensed the source code and is allowed to make, distribute and sell Zeta. This is obviously clear because it is not a group of anonymous hackers, but a registered company.
The second: Zeta or BeOS or Haiku will never ever(!) become “mainstream” or “the 4th top OS”. Why? This is simple: Zeta can per se only be used in private households because it is a single-user OS. Zeta will never become a multi-user OS because nobody is working on it and nobody will ever be. Using a single-user OS in a corporate environment is unacceptable, and then there is the lack of quality(!) software.
Private households are not profitable at all. There is a monopoly, Microsoft. Microsoft forces every private customer to buy Windows with every new PC. Then there is Apple. Apple had a single-user OS and almost died because it was ****. Then Apple made a multi-user OS and now everything is fine because there is enough software, even Microsoft Office. And then there is Linux which is free both as in speech and as in beer and very much hyped by geeks and IBM and some others, but although it is hyped by geeks, it can be used by anyone and has a reasonable amount of quality desktop software.
Conclusion: In my opinion, there is no room for Zeta. Companies and universities cannot use it because it is not good enough and not secure enough because every Zeta user per definitionem has administrator rights. Private households could theoretically use it, but they will not because there is a monopoly and they want games and 3d and 100% hardware support and a pirate copy of Microsoft Office, whether they actually need it or not.
Therefore Zeta or BeOS or Haiku will never ever(!) become “mainstream” or “the 4th top OS”. By the way, someone said that a combination of Zeta and Haiku might be ready for wide usage in two years. Today, it is 2005. In two years, it will be 2007. GoBe2 was written in 2000 and it will never be updated because its source code is not available. Do you really think that anyone will use a 7 years old and since then unmaintained productivity suite in 2007? You must be a real BeOS fan(atic)!
Well this is kind of reassuring. Too bad they have such a dorky name. Zeta! I wonder how long did it take them to come up with that?
>>Is it just me, or does anyone else think it’s strange
>>that the 2Ghz Centrino outperforms (on avg) the 3Ghz P4,
>>XP 3000, Athalon 64 3000 and all other processors
>>presented in these benchmarks?
That is because the Centrino’s have 2Mb of cache. Therefore it is hard to compare the real clockspeed of these CPU’s with others who have much less.
Re: blob
Is it just me, or does anyone else think it’s strange that the 2Ghz Centrino outperforms (on avg) the 3Ghz P4, XP 3000, Athalon 64 3000 and all other processors presented in these benchmarks?
As Frans said, it could be either due to the 2Mb cache, or due to memory timings/speed. The difference is small enough to be more memory timing CAS/RAS/DELAY etc than anything else.
Re: Zeta scepticist
Zeta or BeOS or Haiku will never ever(!) become “mainstream” or “the 4th top OS”
Windows 95/98 and MacOS Classic were not multiuser, and they sure had a large market share. Zeta/Haiku only needs 1% market share to be considered sustainable (both for drivers and usable apps). Zeta/Haiku dont have to achieve it with R1 – Haiku will pursue a multi user system for R2, and you can be assured that Zeta will follow.
GoBe2 was written in 2000
The source code is still available, who’s to say that Gobe wont resurrect? An alternative is also possible – the Java port for BeOS is waiting Sun certification (beaurocracies take time), but once that’s done, you can expect a OO.org porting group to show up and have something usable in 2007. Case in point – Mozilla/Firefox is very decent under Zeta these days.
my comments
Sceptics dont understand that if yellowTab manage to sell a decent number of copies, they will get the finances needed to further expand the OS, and add everything thats needed for a 21st century OS. They do have the original source code after all, and there are quite a number of qualified Haiku engineers to be poached.
it must support some as it work on the Nimble V5.
http://www.yellowtab.com/support/faqs/show.php?id=6
I heard that Zeta is using some illegal code. Is this true?
No. yellowTAB does not use illegal or leaked software.
but bad management is what BeOS Is all about!
Right, but this wasn’t exactly invented by Be Inc. They just cloned it. Amiga already had it in the 80ies you know
Apart from that: Nice to know that YT Zeta is still alive & kicking.
I can not wait, this should be interesting to try out. I’ve tested BeOS 5.x on a variety of hardware, and it has booted on less than half of those machines, be they late 90’s hardware, or more up to date stuff. Regardless, I’ve never given up on this OS, because simply put, it is the BEST thing that ever happened to me since Amiga OS.
You have to recall that some of us were multi-tasking with a multi-colored GUI based OS with only a smidget of RAM, while in the mean time, EVERYONE else was either busy with command prompt, or some slow, singletasking, unresponsive B&W GUI absed OS, that locked up when it ran out of RAM.
Needless to say, when Amiga went bottom up for the 5th millionth time, the transistion to winblows 3.11 and later 95 on up, was not at all fun. It was more like taking 10 HUGE steps backwards, and learning how to walk all over again, not that winblows as hard to learn, it was just that it was and is so behind the times, and demanding and abusive to system resources, unlike Linux, and Mac OS.
BeOS on the other hand required a decent chunk of RAM to run at all (unlike Amiga), but once you had that RAM, it treated your resources VERY nicely, multi-tasking was smoothe as could be, you could actually move a file while it was in use which is something I’d NEVER seen before in my life. The configuration was as easy if not easier to learn and set up as Windows’, I could go on. For me, it is about perfect, it just needs better software, and of course better hardware support, and the completion of Dano, or what ever they are going at, the Netcode sux, and we desparatly need some Hardware Accel for games.
Me personally, I could care less about multi-user account, I never use them anyhow, leave that to system admins, which I am not, and find to utterly boring to partake in anymore.
I ment, “not unlike” OS X or Linux, they are all abusive.
Still sounds like a lot of patching to me – not that it actually matters. I just hope my bank account is considerably healthier than it is now come 1.0 release time
Don’t be fooled by the “patching” issue. It is done this way, that’s normal.
“This is obviously clear because it is not a group of anonymous hackers, but a registered company.”
Ah yes, the “No company can break the law” theory
Companies break the law all the time. In fact since the company cannot effectively be jailed, it’s much safer to form a company, use that to break the law and then if you get caught the company maybe receives a fine, or whatever and you have no real consequences.
I think what you want to say, but can’t quite bring yourself to say is…
“Breaking the law is OK if I get my new BeOS”
“Breaking the law is OK if I get my new BeOS”
In this particular case…damn right!
Mat:
Zeta is a dorky name, but do you know what it is? Zeta is the 6th letter of the greek alphabet. They are shooting for an “R6”, hence, Zeta. I wonder how many others haven’t figrued this out all these years….
Zeta scepticist:
As for multiuser, it isn’t a defacto requirement in businesses. In fact, it’s more of a buzzword for people to throw around saying something isn’t good enough because it doesn’t have it. True, multiuser exists, but not on the same level of many Unices. Most corporate desktops user Windows. Period. Sure, you can log in as another user, but you can see files and folders, whether private or not. OS/2 is still used at a lot of businesses, and it’s not inherently multi-user, although filesystem addons can make it that way, much like Unices. Your using this as a fact or requirement of businesses or for *anything* at all clearly shows your lack of being in touch with businesses and their needs. It is rare to see a full on Unix shop, not using a single windows machine. I guess you could count the Macs in this case (using OSX), but they usually give their users admin rights anyway in my experience. BTW, PhOS has multiuser. Now don’t get me wrong, I’m not rooting for Zeta. Just pointing out your theory using multi-user is flawed and unproven.
I’d love to play with it, but until they get the license cost down to something reasonable…
Like $39.95US or $49.95US, I’m not going to invest over $100US for an OS to “play” with.
I didn’t pay $100 for BeOS, I’m not going to pay that for Zeta.
You are asking for the wrong thing to be proven.
We have the generally accepted facts of what transpired a few years ago between Be and Palm, anything that disputes those facts needs to be proven, not the other way around.
No he’s not. Nobody is disputing the deal between Be and Palm. They are disputing that claim that YT does not have legal access to the sources – a claim which has never been demonstrated with anything even close to real evidence.
YT originally got the sources from Be inc. under license – they said this at BeGeistert some years ago.
They may not have got the kernel sources from Be but it’s pretty clear they got them as well at a later date (over a year ago now), they wouldn’t be making changes and putting their copyright notice on it if they didn’t.
Using leaked code and applying patches to it would be not only illegal but downright suicidal commercially.
I got the distinct impression there was a deal in place when I spoke to YT last year – they’ wern’t saying much but there’s definately an NDA of some form as they were being very cagey about what they would say.
Just because a deal hasn’t been publicised doesn’t mean it doesn’t exist, in fact most deals will *not* be publicised – that’s how business works.
people bother with the trolls. Nothing will ever convince them they are wrong. That is one of the perks of being a Zelot. No fact can make you stray from your unreasonable conclusions. Must be fun to live in a world of your own creation and rules.
For those of us with an open mind, no loger need an official word. It would be plain insane for a company as public as YT to being using illeagle code. The Palm lawers would have been all over them before RC1 even hit the street.
But logic has no room in the mind of a Zelot, of any kind.
I don’t know about Europe but in the US can you imagine a Palm Lawer in court over YT. ‘Yes your Honor. We have been living on Mars for last few years and had no idea that YT had stole our code. Then improved it and sold a product to thousands of people and we never heard of it.”
Thanks, Nicholas, for saying more elloquently what I’ve been trying to say. This fud from people like CPUGuy is getting old. This “generally held” knowlege about what deal was made with Be, Palm and yellowTAB isn’t “generally held” knowlege at all, just conjecture.
I originally thought this would put all the issues to rest with yellowTAB but unfortunately some people just can’t let it go. For me, I’ve moved on to the Mac but I still have a PC sitting here for playing with Zeta. It’s been pretty cool seeing the transition from the first beta of Zeta to where we’re at now. For the most part it’s gotten better and better, and it looks like the best is yet to come.
Also don’t forget that the centrino is based on PIII technology not P4.
On the same frequency/speed P4 is way slower than PIII but can run on higher frequencies because of longer pipelines.
BUT:
Longer pipelines = less execution per frequency.
…but on DANO. :/ So at least some of the changes mentioned were already present, and the kernel comparisons should have been against Dano kernel… or am I missing something else?
:/
I am surprise that no-one has suggested the other way that Zeta could have gotten legal Kernel source and it explains the long delay in doing up dates.
Reverse-engineering.
If Zeta disassembled the original Kernel they would end up with legal source code but no useful labeling of routines/variables. I remember running across a dis-assembler for Beos a year or two ago, so it can all be done inside BeOS itself.
Once they have the basic code it would take time to figure out the routines and variables, about now would be the right timing for being able to produce useful modifications.
the other alternative which is to buy a new kernel from a software company, and have it rewired to suit BeOS. For a couple of engineers full time, modest investment, little problem really, and no one need ever know.
Doubt it though.
I guess we should wait for the kernel to be disassembled and analysed at instruction level before the real ammount of changes are known.
lol
“Using leaked code and applying patches to it would be not only illegal but downright suicidal commercially.”
Two years of selling people this 3rd rate junk doesn’t seem like suicide to me. Bernd has money in his pocket, and PalmSource can only sue the empty shell of the company at most, with no benefit to them at all.
Of course for the users stuck with this junk it’s a bad deal, and you can see plenty of them complaining, but BeOS always was about unfinished code and “we promise we’ll fix it later, now where’s our $100?”
@ Anonymous
> PalmSource can only sue the empty shell of the company at
> most
yT has been doing business for quite some time now. IMO if there are any important legality issues involved it’s up to PalmSource or other parties themselves to come with a public statement and take legal action. So far they haven’t even made a public statement and legally they would be required to protect their own assets.
So IMO if you like Zeta, support the effort and let the companies themselves worry about any alleged legality issues.
I’ve never gotten deep into the politics of these two, but being an open source supporter, I’ve resided to wait for OpenBeOS to complete. After all, they seem to be trying to fix things from the ground up.
Anyway, whatever the issues (and I would like them to be CLEAR) competition can only be a good thing. Imagine what Windows would be like if OS/2 were binary compatible with it, if there was a force keeping them convergent.
> Imagine what Windows would be like if OS/2 were binary
>compatible with it
Uhm, it was, at least for awhile… 16 bit (Windows 3.1) anyways.
I’ll guess that the reason people get so excited about the legality issue is that most of us would not use software of questionable legality, except possibly for their own use. The idea of a company founded and making $ from illegally obtained software is repugnant, especially if you have a fondness for BeOS. As far as proof, YTab doesn’t have to go into gory details about any “deal” with Palm; all they have to do is include a simple statement that their product includes code licensed from Palm. Do they have such a statement?
As for the price, since most of us get the operating system included with a system (or from work), $100 for a “hobby” system seems pretty steep.
Never mind what lies under the hood….
For me, the only important thing about Zeta, is the progress that Yellow tab is making….
Anyway..
If someone wants proof of what is what, and who owns what, under the hood. Then why not find a way to “backward compile”; Beos-5-PRO, Zeta 1.0 and BeOS-5-PE, then compare the source code’s, and the “battle” for what is what and who owns what will be decided.
Zeta is too expensive ($100)
There is a home version you know, which costs 50 Euro. This is the average price of a computer game today.
I wont bother with Zeta, I’ll wait for OpenBeOS
If you’re not a coder, the best way to help Haiku is to get Zeta. By expanding market share, you get more software, which again helps expand market share, which increases Haiku awareness, etc. If you cannot see that connection, then I can only deduct that you have the intelligence of a teenage kid. Learn from your adult friends for a change.
If it’s not free, I dont want it
Have you ever been to a restaurant in your life, or do you eat from garbage bins?
yellowTab is using illegal code
Now where did this stupid idea come from. Goebels once said – repeat a lie often enough, and eventually it will become true. Though to be honest, it appears that it took some time for YellowTab to get the licence, since YellowTab from 2 years ago is different from YellowTab today. I think that YellowTab have finished making their baby steps, and have finally learned to walk, and are becoming a decent geek friendly company. Yeah.
Disclaimer – I purchased Zeta RC1 the very day it was available, cannot wait to get my Zeta FINAL
Reverse-compiling or whatever you wish to call it between any variant of BeOS proves NOTHING one way or the other, except for what’s different under the hood. It doesn’t prove anything for having source code, legally or otherwise. All it can do is give a hint that the optimized code output is identical or not, without giving full source code as output that was written by a human. A decent compiler when optimizing changes the code significantly enough that it won’t look like the input source code if you were to disassemble it. How do I know this? From experience in stepping through release-optimized code with debugging symbols, that’s how!
So your suggested method of trying to prove anything proves absolutely nothing more than I stated above: what the compiler with that set of optimizations for whatever the input code was, ended up generating, and whether or not the code is byte-for-byte identical with something else you are comparing against. At that rate, you might as well go along and do a binary diff, for all the useful data it will give you, because it won’t give anything more or less towards your stated purpose than decompiling the code completely.
This is great news! I bought Zeta RC1, and i’m gonna invest in R1.0. Thats for sure.
Great work YT.
PS: Kudos to the Haiku team aswell. They are showing great progress. Cant wait to dualboot Zeta and Haiku.
I am not saying that YT doesn’t have the source, I’m just trying to get you to understand something other than your own viewpoint.
I could personally care less whether or not they have the source, as OpenBeOS will be the dominant BeOS when it is done.
I’ll fork over the $100. I’ve had no bigger disappointment in computers than BeOS tanking. I can’t believe those management idiots stopped developing the OS and focused on imbedded garbage. I’ll bet they kill Palm too.
This Zeta news is fantastic, I’m in, ASAP. I’ll always use linux but windoze and mac suck, so Zeta is my only hope for a pure desktop os.
These ‘idiots’ were having trouble getting investors, and entering into the embedded market was the only way to for them to get any.