incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jason van Zyl <>
Subject Re: [PROPOSAL] Apache Ace
Date Mon, 06 Apr 2009 15:22:47 GMT

On 6-Apr-09, at 12:33 AM, Richard S. Hall wrote:

> Jason,
> Although, we keep trying to point out that OBR != p2, you seem to  
> keep missing that point.

The argument is not lost on me. That they are not that far apart  
insofar as providing a repository system with an API to retrieve and  
manipulate artifacts is the point you are missing. They are not that  
wildly different and one could either of the easily evolve into the  
other. That's why I keep pointing it out because when anyone said OBR  
at EclipseCon the word provisioning always followed in the next  
sentence. Followed by a comparison with p2. I think both technologies  
are relevant in any discussion about provisioning OSGi.

> OBR is a simple repository model and API for accessing it, that's  
> all it is not a provisioning system. As such, OBR has been  
> "done" for a long time. All other functionality should be hopefully  
> be buildable as layers on top, such as what Luminis has done with  
> their provisioning work.
> You act like there is some "gotcha" that OBR is not an OSGi spec,  
> but OBR has always existed outside of the OSGi specs, so who cares?  
> The proposal literally only mentions the letters "OBR" once as a  
> dependency and nothing more. It is hardly the main selling point.
> No one was shouting about OBR or p2, that was only you.

In a year from now when anyone is talking about provisioning OSGi what  
do you think the main underlying technologies bases will be? They will  
be OBR and p2. Either one of them will be expanded and changed and  
they will be the basis of most if not all provisioning technologies. I  
think you know that as well as I do.

> Also, the notion that we should just lay down because we can't  
> compete with some big company and all these man years they have  
> invested is somewhat ridiculous. If we all bought into that, then  
> none of us would be here.

As I also stated there are lots of small companies involved as well.  
I'm just saying pick your battles. Do you think a business manager is  
going to say "Hmm, this system has 5 man years of work in it and is  
used by a lot of people ... Well let's not consider that because  
that's ridiculous." Anyone trying to use the technology will not use  
that argument as a reason not to use it. I'm just saying it's a  
possible reason for not wanting to develop something else. If this was  
a proprietary solution not accessible, and not extensible then an open  
solution would be great, but that's not the case with p2. I would  
argue it's more of a proprietary case for OBR given the constraints to  
participate in the forming and implementation of the specification.

> If you just wanted to point out that p2 should be mentioned as a  
> competing technology in the proposal, I think you could have  
> accomplished that in a more reasonable manner.

Maybe. I'm not a dancer.

> Lastly, it is somewhat difficult for me to take community building  
> lessons from someone who claims to have had an OSGi awakening and is  
> willing to cull all of their own personal projects as a result, yet  
> I can count on probably a couple fingers how many discussions you've  
> instigated (or even responded to) regarding OSGi, OBR, or any topic  
> in the Felix community in all the years it has existed.

Heh. I _never_ claimed to be a an example of a good community builder.  
I wouldn't take any community building lessons from me. I write stuff,  
if you want to use it great. If you don't it's no skin off my back.  
That's the extent of my community building skill.

On the OSGi front I probably wouldn't be involved in many discussion  
on the Felix list because I use Equinox. So I don't think that's  
overly odd.

> -> richard
> On 4/5/09 2:11 PM, Jason van Zyl wrote:
>> I'm suggesting that  you two groups figure out how to work together  
>> on a very hard problem.
>> I'm also saying that you are unlikely to out do the 5 man years in  
>> p2 already.
>> As I said in the previous email if you want to make a competing  
>> system that's fine. But don't couch the proposal as something  
>> that's new and hasn't been addressed elsewhere because it has.
>> You might want to be more clear in the proposal about p2 being a  
>> competitor, also make it clear that OBR has gone back to  
>> specification, and what it is you're actually working from. So when  
>> a user or potential developer looks at this and says what  
>> specification are you working from they can see "there isn't one  
>> yet", and if they ask "what about p2?", then it's clear you decided  
>> not to collaborate with them. I think you can even point out that  
>> they didn't collaborate with you either. Give people all the  
>> information.
>> When I walked into the OSGi BOF at Eclipse I was dumbfounded. The  
>> same dose of sniping and grin fucking as other groups I've worked  
>> with which was disappointing but I guess I'm not surprised. There  
>> were attacks abound at EclipseCon. The way p2 came into existence  
>> probably could have been handled better, no doubt. But I don't find  
>> guys like Hal very compelling with his melodrama (

>> ).
>> Make it clear to people looking at the proposal that provisioning  
>> is a hard problem. These arguments about the "Eclipse way" of p2  
>> and non-focus on server side or other types of systems is nonsense.  
>> If you actually  have a pointer to p2 in your proposal -- which is  
>> conspicuously absent -- siting them as a direct competitor users  
>> will have a clear point of reference. If people had the background  
>> story they will probably go WTF just like I did.
>> Both sides of the p2/OBR seem to be equally obstinate and non- 
>> collaborative. I used p2 because from a technical level as an end  
>> user because it worked. There are nightly builds, lots of  
>> documentation and at least 5 people working on it full-time at any  
>> given point in time. If you look at the p2 code and the OBR spec  
>> they are 90% the same thing and any differences are easily  
>> compensated for with a little effort.
>> Competition is fine, I would just be more open about that aspect of  
>> it in the proposal.
>> On 5-Apr-09, at 8:47 AM, Karl Pauls wrote:
>>> On Sun, Apr 5, 2009 at 5:00 PM, Jason van Zyl  
>>> <> wrote:
>>>> On 5-Apr-09, at 2:46 AM, Marcel Offermans wrote:
>>>>> Hello Jason,
>>>>> On Apr 5, 2009, at 1:09 , Jason van Zyl wrote:
>>>>>>> Equinox p2 was designed to replace the aging Update Manager in
>>>>>>> Eclipse. It focusses on installing Eclipse-based applications
>>>>>>> from
>>>>>>> scratch and updating them and can be extended to manage other
>>>>>>> types
>>>>>>> of artifacts. If you look at the "agent" part, it is geared 

>>>>>>> towards
>>>>>>> desktop environments
>>>>>> Not true.
>>>>>> Jeff McAffer's demo at EclipseCon is a case in point. He  
>>>>>> provisioned
>>>>>> an EC2 node using p2. [...] Jeff is very much focused on server 

>>>>>> side
>>>>>> provisioning as am I.
>>>>> Let me rephrase that, it's geared more towards desktop and server
>>>>> environments, as compared to smaller (embedded, mobile)  
>>>>> environments. That
>>>>> was the point I was trying to make here.
>>>>>>> Note though, I'm no Equinox p2 expert. :)
>>>>>> Then why are you proposing this when you don't even know what  
>>>>>> p2 is
>>>>>> capable of?
>>>>> We started working on this system when p2 did not even exist. I  
>>>>> even
>>>>> remember talking to Jeff in those days about our system, but  
>>>>> they decided to
>>>>> make their own, so you could equally well make this argument the  
>>>>> other way
>>>>> round.
>>>> I'll use the same story I used on Richard. I created a DI and  
>>>> runtime system
>>>> 5 years ago. So what. Guice and Equinox have a massive user  
>>>> community,
>>>> professional support is available for both and so I will cull the
>>>> technologies I developed. I don't think it really matters so much  
>>>> who was
>>>> first but who got to a production system first that is known and  
>>>> support by
>>>> thousands of users.
>>> Are you suggesting that we shouldn't incubate projects that overlap
>>> with an existing production system outside the ASF?
>>>>>> It's just my opinion but anyone doing provisioning with OSGi  
>>>>>> has had
>>>>>> their asses handed to them on a plate by the p2 guys.
>>>>> In my opinion, p2 is fine if you are already doing everything  
>>>>> "the Eclipse
>>>>> way" and are targetting desktops and servers. There are however  
>>>>> other types
>>>>> of systems that need provisioning, and Apache Ace tries to cater  
>>>>> for those
>>>>> too.
>>>> Again you haven't really even looked at p2. What is the "Eclipse  
>>>> way" ?
>>>> You're going to make/keep another system entirely because it's  
>>>> the "Eclipse
>>>> way" ? I've seen JBoss and Tomcat servers provisioned with p2 so  
>>>> I'm not
>>>> sure what the "Eclipse way" means. I'll repeat again that p2 is not
>>>> targeting desktops whatever aspects may appear most visible right  
>>>> now. I
>>>> really don't think there is a system that couldn't be provisioned  
>>>> even with
>>>> p2 in its current state. I have personally not found one yet.
>>> I don't think anyone is attacking p2. If people like and use it:
>>> great. I certainly think the proposed project should be able to
>>> interoperate with p2 repositories seamlessly. It sure would be great
>>> If you could suggest any improvements to the proposal in the area of
>>> interoperability with p2.
>>> With that out of the way, I do think there is room for another
>>> provisioning solution out there. Granted, it might be that it just
>>> doesn't have any added value over p2 and that people are going to
>>> ignore it but I'd say this risk exists for all projects, no?
>>> During the incubation, we will see whether the project is able to
>>> attract enough users and contributors. The initial interest looks  
>>> very
>>> promising IMO.
>>> regards,
>>> Karl
>>>>>> Oleg and I were trying to make something and after looking  
>>>>>> around at
>>>>>> everything -- and we did look at OBR -- we decided that p2 was  
>>>>>> good
>>>>>> enough and we would help improve that.
>>>>> OBR is a repository for components, augmented with metadata that  
>>>>> describes
>>>>> dependencies. As such it's not a provisioning system, so in my  
>>>>> opinion you
>>>>> should not compare it to p2.
>>>>>> There's nothing wrong with competition but I think anyone doing 

>>>>>> OSGi
>>>>>> provisioning is just going to look around in a year and find p2 

>>>>>> has
>>>>>> 95% of the market. It's a complicated problem and I think p2 is a
>>>>>> solid base and be improved and adapted to support  things like  
>>>>>> OBR or
>>>>>> anything else including non-OSGi systems.
>>>>> Nobody can look into the future, and since both p2 and Ace are  
>>>>> indeed
>>>>> software provisioning solutions, there will definitely be  
>>>>> overlap in
>>>>> features. There are also differences though. In the end, the  
>>>>> users will
>>>>> decide what they like best.
>>>> There's no doubt they will.
>>>>> Greetings, Marcel
>>>>> ---------------------------------------------------------------------
>>>>> To unsubscribe, e-mail:
>>>>> For additional commands, e-mail:
>>>> Thanks,
>>>> Jason
>>>> ----------------------------------------------------------
>>>> Jason van Zyl
>>>> Founder,  Apache Maven
>>>> ----------------------------------------------------------
>>>> You are never dedicated to something you have complete confidence  
>>>> in.
>>>> No one is fanatically shouting that the sun is going to rise  
>>>> tomorrow.
>>>> They know it is going to rise tomorrow. When people are fanatically
>>>> dedicated to political or religious faiths or any other kind of
>>>> dogmas or goals, it's always because these dogmas or
>>>> goals are in doubt.
>>>> -- Robert Pirzig, Zen and the Art of Motorcycle Maintenance
>>>> ---------------------------------------------------------------------
>>>> To unsubscribe, e-mail:
>>>> For additional commands, e-mail:
>>> -- 
>>> Karl Pauls
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail:
>>> For additional commands, e-mail:
>> Thanks,
>> Jason
>> ----------------------------------------------------------
>> Jason van Zyl
>> Founder,  Apache Maven
>> ----------------------------------------------------------
>> To do two things at once is to do neither.
>> -—Publilius Syrus, Roman slave, first century B.C.
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail:
>> For additional commands, e-mail:
> ---------------------------------------------------------------------
> To unsubscribe, e-mail:
> For additional commands, e-mail:



Jason van Zyl
Founder,  Apache Maven

In short, man creates for himself a new religion of a rational
and technical order to justify his work and to be justified in it.

   -- Jacques Ellul, The Technological Society

To unsubscribe, e-mail:
For additional commands, e-mail:

View raw message