[openbeos] Re: [6?] Re: OpenBeOS in competition with BeOS
- From: revol@xxxxxxx
- To: j-schwerer@xxxxxxx
- Date: Sat, 25 Aug 2001 20:14:07 +0200 (MEST)
Well, anyone wants to take over next time ? (kidding) ;-)
En réponse à j-schwerer@xxxxxxx:
> First things first: I speak only for myself, not for Open BeOS in
> general=
I too, for this time.
> >Going quickly this time too, have a lot to do today.
>
> Mmmh. I assume you shouldn't worry about giving slow answers. There are
> t=
> hings
> that I believe need time to be discussed ^_^
I didn't say I wanted to get rid of the question. I just said that I had a high
load and that I put my ideas without taking the time to clean-up some mistakes.
I agree that it's a key-point that must be disscussed.
>
> A quick explanation so you can understand my point:
> [strip]
On my side I know BeOS for a bit long (since R4 I think), but I really begun to
use it on a daily basis 2 months ago :)
I use Linux also and I must admit I have a Windows partition :P
> BeOS. This is one of the reasons I dived into the Open-BeOS Project,
> beca=
> use I
> want to make sure of a way I can stick to BeOS as long as possible ^_^
Same goal, but maybe not the same approach.
> Right now I have no idea if this project will ever succeed or not. It
> mig=
> ht
> completely succeed, partially succeed (achieve only a few components),
> or=
>
> completely fail. I don't know.
Neither do I
> I see another reason for doing it: I like knowing what my computer can
> do and
> try pushing it further, I love networking and I'm glad I'm going to go
> deep
> down into the networking stack. There's a lot to learn in there and
> I'm impatient to learn that :^)
>
> But as you can see none of this is incompatible with a commercial
> project=
> =2E
>
> As far as I'm concerned I'd love the BeOS to have open components, such
> a=
> s the
> network, the kernel, the display and printing systems.
> But I would perfectly understand if something like the media kit was to
> remain
> closed-source, as I see it as an area of innovative advance NO other
> Operating
> System I know ever got close to and represents a lot of added-value the
> BeOS
> can take advantage of.
Hmmm... I think that this part is _the_one_ that needs, not to be open-sourced,
but at least given access to for external contributors. the media-kit codec
headers are the perfect example of what not to do if you want to keep BeOS
up-to-date: they were kept under NDA for +than 1 year telling "We will release
them one day when they are finished..."
Where are we now ? yes, we have a Divx codec, but look what HACK it is !
same for OpenDivx, there is two implementations known to me, one using the
libraries (and so is in law) but is old and very buggy, and the other has been
comiled without thinking about the NDA issue, making it forbidden to release
because it is based on both GPL code and the NDA-needing headers.
>
> Enough with the preliminaries. I'll explain more as we go :^)
hehe...
> >The maintainance _is_ the primary reason why we don't want to rely on
> Palm
> >(I mean that we want to get it Free): Because even if Palm agrees to
> help
> >develop it today, nobody knows what they can decide in the future, and
> also
> >nobody knows what will happen to them. We _don't_ want another Amiga
> story
> >with IP passing from hands to hands, not letting USERS have control
> over it.
>
> But if you get Palm blessing for stripping out licenced code and adding
> into
> current server opensource code then you would be a fool not to rely on
> Palm
> for as long as they want to help you, be it only 6 months if they decide
> not
Sure I would. What I am afraid of is Palm saying one day 'well, thanks, but
it's over now, and you won't be able to distribute what you did' or maybe they
could well bankrupt the same way Be did and we would have to get things all
over again. That is what I don't want to happen.
> to help you longer than that. Because every little piece of opensource
> code
> you'll add into BeOS and every bit of licence code you remove from BeOS
> will
> be steps going the same direction as your goal.
> You won't be able to do everything at the same time: maintain the BeOS,
> expand
> it, and recode it. As Peter put it in an earlier mail, if you only
> recode it
> you will end up way too far behind other operating systems and HW
> evolution
> when it's done.
I know all that. That's why I really want Palm to answer positively.
> I strongly believe that if Palm were to outsource the code it would be
> the
> best thing for BeOS ever, period.
I too, period.
> Should they outsource the network stack I would:
> - implement IPv6, PPPoA, PPTP and such protocols support in an
> opensource way
> - expand the network kit API, which is something I'm already working on
> in an
> additional shared library way
> - look for licenced code to remove if any
Prems !!!
(french for "I was the first on this :P")
> While it doesn't seem to fit in your view to work with Palm I believe it
Who said that ? not me.
> would also serve you, as it would:
> - help expand the BeOS capabilities
> - make BeOS closer to open-source reality (your goal)
> - let someone market, distribute and support the BeOS, which would bring
> more
> users and developers (and that will help you as well because you'll have
> more
> people to help you with the code and you'll gain more momentum).
Also my goal.
> >> That said, whereas you guys could happily hack the _servers and
> >> improve the Kits after hours, I can tell you right away that
> >> marketing the product in order to muster more support, get more
> >> users, more developers, etc. costs money. This means: BeOS will
> >> cost money. It will NOT be free. There WILL be a free version
> >> like PE, but the real product will be a commercial one. (Free
> >> bankrupt, unless you are willing to pay out of your own pocket,
> >> but this is impossible if you think about the scale!). =
>
> >
> >Do you know what Linux is ? Not talking about RedHat or Caldera or
> so.
> >Talking about the Linux _kernel_. This is Linux. And _this_ has been
> mad=
> e by
>
> >people _on_their_ spare time, _without_ being paid.
>
> I believe this to be a wrong assumption. The Linux kernel might be Linux
> to
> you, but it is not what MAKES Linux. Linux wouldn't be much of a thing
> today
I know it does not, but it is what created all the stuff we see around.
Without Linus playing with a teeminal emulator (yes this is what Linux was
first) during his spare time, there would have been no RedHad, no Caldera.
I don't say they don't contribute to Linux I said they are NOT Linux, only
part of it, the heart remaining the kernel (and all the GNU tools that BeOS
also uses).
> if it hadn't got press coverage, marketing, companies backup. And I feel
nor would it have without the kernel.
> sorry
> but there actually are people being paid for improving code in Linux's
> parts.
> Not the kernel, sure, but there are developers hired by Redhat and
> Caldera
> working on improving some components (I believe someone from RedHat is
> working
> on the glibc) and on adding some distribution-specific tools.
And there are also people that do it without being paid.
That is what makes Linux profitable for both commercials and OSS/FS advocates.
Without the other part, either it would be a Redmond-like software, or an
still unnown AtheOS-like. (not to flame AtheOS, which is great technically)
> >> > There I see you don't understant what we want to do.
> >> > We don't want to _market_ BeOS but _make_it_free_ We won't get
> >> > paid for that (or if you want to pay us, well I'm glad to hear
> >> > from you)
>
> I am not an open source crusader, I am a BeOS crusader. I hope you feel
> the hell of a difference it makes to me.
I'm both, sorry for that :-D
> >I take care on this myself. I know not every BeOS user is a
> programmer
> >(though the proportion is much higher than in Windows' world), but what
> >I like in Linux is that if something soesn't suit me I just have to
> recode/code it and more, I can share with others what I did.
>
> You can do this as long as you have enough time and knowledge. There are
> people like me who hardly get time for everything they need / want / wish=
> /whatever. And I'm afraid I think most of the people are like this.
> If you wish to make an Open BeOS for yourself only, then go with it. But
> don't
> be surprised if you find some people jumping off. I find it a very
> selfish
> point of view to say you take care about reliability and consistency
> yourself,
> because many people just can't do it, and they're expecting people like
> you to
> do it for them, and people like Helmar or Palm to bring what you do to
> them.
> They cannot do anything with what you did because you did it for you, in
> your
> corner, and despite your sharing it, it cannot fit everyone's needs.
> This is
> what I hate in Linux, the "I do it for me, it works for me" attitude.
> Not
> being selfish is not about sharing your results, it is about thinking of
> the
> others when you develop and test, this is entirely different.
This is also what is good is Linux, you take what fits you. Of course,
some won't find something that fit them. It's just another approach.
Linux has always been "selfish" from the beggining, since it was just a
personnal hobby. but the point is many people found that what someone needed
was also what _they_ needed, not what someone wanted them to need (like at
Redmond...).
Of course it's even better if you listen to others, but I don't see why you
couldn't do it while coing what fits you.
Just to take an example, I coded nplay because I wanted to see DivX comfy
under BeOS. That fited _me_. Then someone saw it and told me "hey it works
for me". Then they asked "could you implement this or that ?" and since I'm not
totaly selfish, well I did it also. and I find it great. Of course I still
have the power to say "well this will make it huge and I'd like to keep it
lighweight, because I want it to play on my old comp, but do it if you want".
> >> Nobody -apart from a handful of folks- cares about a free BeOS.
> >Hmmm... I know lots of ppl in the Linux/FS community that are diing
> >to see BeOS become open-sourced :-) They just say :
> >"BeOS ? it's proprietary, it sux"
>
> Your quotation doesn't fit. If they said "I'll develop for BeOS when
> it's open
> source" I would say you have a point. Saying "it's proprietary, it sux"
> shows
> closeness, narrow-minded opinion, and the person who said this is
> definitely
> not someone I'd like to be working with.
> I don't think people saying this will ever bring anything good to the
> BeOS,
> even if the BeOS was entirely open-sourced.
I think you never met RMS :))))
Well both kind of monkeys exist. some will be happy when it comes, and some
will remain angry because it's philosophically incorrect to work on closed
source things.
> >> that they put money on the table, because they/I know that 1) at
> >> date X there will be a new version or an update to the code, 2)
> >> within 5min. I get an answer to my support query, 3) if I go to
> >> Shop A, they know about the system I am using and 4) the software
> >> vendor offers a BeOS version of their product.
> >
> >Others don't have to have a vendor that knows their system, cause they
> know it
> >better than everyone else, because _they_ did it.
>
> Therefore you're saying only the guys developing OpenBeOS will be able
> to use
> it? Sorry for playing the Devil's advocate here but what I undestand
> from what
> you write is that it takes to be a developer to be able to know the
> OpenBeOS,
> which is very selfish. It means my grand mother can't use OpenBeOS? But
> that's
> just what I want! Then she needs a vendor to know the OpenBeOS.
NO, never said that, maybe I should have expressed it an other way, or maybe
you should have read it once more.
I said that though lots want to rely on someone to implement what they need,
because they don't have knowledge, need, time, whatever , _some_ want to do it
themselves, be-it either to keep more control over it, or to keep changes
secrets, or... or because they don't want to rely on someone else for
security reasons (thinking about an organisation I do'nt recall the name that
told they choose Linux because they had the source, and that they could check
_by_themselves_ if there were security issues, not having to trust any monkey
at Redmond).
> >> 2) I want to develop and market it professionally / commercially.
> >OK for develop, market, well, maybe.
>
> I wouldn't say you meant OK. Because the whole sentence says "develop
> professionally AND market professionally" while all your point is
> about
> hobbyist development. I don't mean that you write bad code [ah those
> tricks of
> thinking in your mother tongue instead of trying to think in the
> language the
> sencence was written],
I perfectly understood the sentence, and I don't think using my native language
(tongue is not very apropriate here) (even sometimes reverting to thinking in
french is hard...), and passed the TOEIC with 940 out of 990.
Sorry for this one :P
What I said is I agreed to help on the devel side "professionnally", but I left
the marketting side to others just because I estimated I wasn't competent on
this.
> I mean you write it on your spare time, that's
> hobbyist
> development. Professional development is doing it for a living, which
> you can't do if it's free :P
I know what it is. Simply I cannot do it full-time, because I have other
occupations.
That is also why I said I wasn't reaching for money.
Anyway he still could have some full-time ppl and other additionnal developers,
one doesn't exclude the other.
And btw if I'm right Apple _does_ rely partly on spare-time developpers for its
Darwin thing.
That thing is what is closer to me idea on what we could do.
I mean if Helmar has a deal with Palm, he will have to work for Sony and
others... What I already suggested so many times I could adapt and sum it up
like this (would it be the only thing ppl read from this mail):
****
1) Helmar gets a licence from Palm to
a) Continue BeIA development (maybe) using what we put in BeOS
b) Market BeOS
c) finalize R6 and ship it
Both will need "professionnal" development, and require deadlines.
Think of it as the 'Aqua' part
2) WE :
a) endorse an NDA to get the source code opened bit by bit.
b) recode stuff removed in an open-source fashion.
c) enhance BeOS with cutting-edge open-source technology.
This doesn't require any fix deadline, and so can be made by ppl during
their spare time and so doesn't need money.
Think of it as the 'Darwin' part.
Of course it's over-simplified, but it's the best way IMO.
***
> >> 3) I want to pay the people who work for me.
> >
> >I didn't say I wanted to work for _you_ I want to work for _BeOS_.
>
> Huh huh, I might be wrong but I believe you want to work more for open
> source
> than for BeOS.
I said I was advocating for both, so... think what you want :^)
> >> 4) I don't want anyone tell me "sorry, I can't finish that
> >> crucial bit of the kernel because I need to find work to put
> >> food on my table".
> >
> >The problem is most of us (at least me, I'm still a student), can't
> >stick to delays, just because we have got others things to do.
>
> And therefore you underdeliver with overdelays. What happens? Users
> and
> application developers will just walk away because they don't see
> anything
> coming. When you're done there's nobody but you to use your work.
That's totaly correct, but what can I do on that ?
You want to have my exams for me ? pass them so, or... :-D
> >Open-sourcing (GPL-ing ?) BeOS as I said will cause all the
> open-source
> >community to endorse us (and trust it doesn't reduce itself to a Finish
>
> >beer-drinking man and a heary one that loves animals).
>
> I don't buy this for a second. Come on, you're a dreamer. It's a nice
> dream
> but if you want to achieve your goal you HAVE TO be pessimistic.
> I know someone who used to say: "Always get prepared for the worse if
> you want
> the best to happen". And I'm afraid he's right. If you want the Open
> BeOS
> project to go somewhere you MUST think about this.
> The open source community has not really endorsed AtheOS so far. The
> BSD
> community is very different from the Linux community. Which community do
> you
> expect to join you? Those who say "BeOS is proprietary, it sux"? They
> wil=
> l get
> everything they think is nice from OpenBeOS and put it in Linux, not the
> other way round.
You're just thinking globally...
btw I didn't say I counted on it. I said they won't be indifferent. How
they will react is another matter.
> >> Yep, and this is EXACTLY why I said "let's NOT try to have all
> >> these different initiatives that confuse Palm", but let's stand
> >> there united, but let's make it clear to Palm that our intention
> >> is to replace the licenced code with free code and to create a
> >> collaborative system on commercial basis. =
>
> >
> >I think some won't agree on the commertial thingy, but mainly I
> agree.
>
> As far as I'm concerned I wholeheartadly agree to Helmar's vision. I
> think
> it's the best for BeOS. And I'm interested in the BeOS and in
> networking, not in open source itself.
I'm interested in keeping BeOS alive and the way I think of is open-source.
I never said it was the only way.
>
> Jean
François.
P.S.: I complained on my webmail, but you also have a strange MUA :P
P.P.S.: wow more than 1 hour on this one ! who said I was taking it hastily.
Other related posts:
- » [openbeos] Re: [6?] Re: OpenBeOS in competition with BeOS
