Re: [RT] Updates to the provisioning model

Hi Carsten, can you provide some insights on timing wrt Felix Snapshot dependencies. When we plugged together the demo we still had a few dependencies on unreleased artifacts for the OSGi R7 work. I would rather see this change happening yesterday than tomorrow to start working on how the model

Re: [RT] Updates to the provisioning model

Hi Bertrand I suppose it would be cat provisioning-model | jsmin | jq . I have not tested this, however. Regards Julian On Thu, Oct 5, 2017 at 3:39 PM, Bertrand Delacretaz wrote: > On Thu, Oct 5, 2017 at 11:39 AM, Julian Sedding wrote: >>

Re: [RT] Updates to the provisioning model

I don't think jq supports comments. But as I stated, all our code writes out perfectly valid JSON without any comments. So you can use this with any tooling that supports JSON. *If* you want to have comments in your source feature, we currently support them as described. Now, looking at the

Re: [RT] Updates to the provisioning model

On Thu, Oct 5, 2017 at 11:39 AM, Julian Sedding wrote: > ...lets stick with JSON + comments... Is that format accepted by JSON tools? Can I for example do cat provisioning-model | jq . ? Otherwise I fear the format combines the worst of both worlds: hard for humans to

Re: [RT] Updates to the provisioning model

Hi Carsten That's fine with me. It just seemed like it might be a good fit because of the comments. Seen that you went down that route already, lets stick with JSON + comments. Regards Julian On Thu, Oct 5, 2017 at 11:23 AM, Carsten Ziegeler wrote: > Hi Julian > > yes,

Re: [RT] Updates to the provisioning model

Hi Julian yes, we considered YAML. First of all, the OSGi Configuration Admin specification will use JSON. That spec started with using YAML, but in the end no one was happy with the format, so we replaced YAML with JSON. While YAML can support JSON, no one really does this - so we end up with

Re: [RT] Updates to the provisioning model

Hi Carsten This looks very interesting! Regarding the format, have you considered YAML? It is a superset of JSON (well, it's designed to also support JSON syntax) and it allows comments out of the box. Furthermore, it's well specified (in contrast to JSON with comments). I assume that using

Re: [RT] Updates to the provisioning model - runtime

Totally agree with what Karl said. One major aspect of the new feature based model is that it describes what a feature is and how features are combined and how the result looks like. Any tooling can then act on this. This will allow us to create Karaf features or an OSGi subsystem or launch it

Re: [RT] Updates to the provisioning model

On Wed, 2017-10-04 at 10:01 +0200, Carsten Ziegeler wrote: > Robert Munteanu wrote> On Tue, 2017-10-03 at 11:52 +0200, Carsten > Ziegeler wrote: > > > > > 2. Reading the section on configuration merging, it is not > > > > clear to > > > > me > > > > if merging is merging of values or overriding

Re: [RT] Updates to the provisioning model

Robert Munteanu wrote> On Tue, 2017-10-03 at 11:52 +0200, Carsten Ziegeler wrote: >>> 2. Reading the section on configuration merging, it is not clear to >>> me >>> if merging is merging of values or overriding of values. >>> >>> Consider >>> >>> >>> feature 1: >>> >>> com.foo.bar.Service >>>

Re: [RT] Updates to the provisioning model

On Tue, 2017-10-03 at 11:52 +0200, Carsten Ziegeler wrote: > > 2. Reading the section on configuration merging, it is not clear to > > me > > if merging is merging of values or overriding of values. > > > > Consider > > > > > > feature 1: > > > > com.foo.bar.Service > >prop1="A" > >

Re: [RT] Updates to the provisioning model - run mode

On Fri, Sep 29, 2017 at 5:15 PM, Carsten Ziegeler wrote: > Stefan Seifert wrote> > >>> * No support for run modes in the model - run modes can be modeled >>> through separate features >> >> is there a plan to generally remove the "run mode" concept from sling - or >> just

Re: [RT] Updates to the provisioning model - runtime

Hi Nicolas, I think the points you mention are definitely worthwhile and for sure, inline with what we have been thinking about. While it is true that for starters, we are only considering the starting of the instance it should be not too difficult to implement the runtime extensions you

[RT] Updates to the provisioning model - runtime

Hi, from what i understood, new provisioning model discussions and thoughts are targeted on starting an instance from a future standardized description. What i would like to understand is if there is any room at this time for thinking about runtime: a way of - describing what is running (feature

Re: [RT] Updates to the provisioning model

I don't want to split hairs here. Whatever is generated by our code is JSON. We are able to process JSON *and* JSON enhanced with comments. Carsten Julian Reschke wrote > On 2017-10-03 11:52, Carsten Ziegeler wrote: >> ... >>> 1. How can we add comments to a feature or application json file?

Re: [RT] Updates to the provisioning model

On 2017-10-03 11:52, Carsten Ziegeler wrote: ... 1. How can we add comments to a feature or application json file? AFAIK JSON does not allow comments Right, strict JSON does not allow it, but we allow comments and the JSON is preprocessed by a JSMin like processor. But after reading all

Re: [RT] Updates to the provisioning model

Robert Munteanu wrote> Hi Carsten> > Thanks for the extensive write-up, it took some time to digest :-) :) > > I have a couple of questions based on who we currenly can (or can't) > use the provisioning model. > > 1. How can we add comments to a feature or application json file? AFAIK > JSON

Re: [RT] Updates to the provisioning model

Hi Carsten, Thanks for the extensive write-up, it took some time to digest :-) I have a couple of questions based on who we currenly can (or can't) use the provisioning model. 1. How can we add comments to a feature or application json file? AFAIK JSON does not allow comments 2. Reading the

Re: [RT] Updates to the provisioning model - run mode

Stefan Seifert wrote> >> * No support for run modes in the model - run modes can be modeled >> through separate features > > is there a plan to generally remove the "run mode" concept from sling - or > just not express it in the feature model? > if it is kept - where come the features and run

RE: [RT] Updates to the provisioning model - maven coordinates

>Carsten Ziegeler wrote> > >> Thanks for the pointer Stefan. We should definitely support this, I'll >> update the code. >> For some reason we have based the provisioning model and therefore the >> feature model on >> https://ops4j1.jira.com/wiki/spaces/paxurl/pages/3833866/Mvn+Protocol >> > >I

RE: [RT] Updates to the provisioning model - run mode

>* No support for run modes in the model - run modes can be modeled >through separate features is there a plan to generally remove the "run mode" concept from sling - or just not express it in the feature model? if it is kept - where come the features and run modes together? stefan

Re: [RT] Updates to the provisioning model - maven coordinates

Carsten Ziegeler wrote> > Thanks for the pointer Stefan. We should definitely support this, I'll > update the code. > For some reason we have based the provisioning model and therefore the > feature model on > https://ops4j1.jira.com/wiki/spaces/paxurl/pages/3833866/Mvn+Protocol > I think the

Re: [RT] Updates to the provisioning model - maven coordinates

Stefan Seifert wrote> >> One minor thing, the current proposal uses the Maven coordinates to >> describe artifacts. It's using the path part of a maven url, so >> basically group id, artifact id, version are separated by a slash. It >> seems that the notation of separating these things by a colon

RE: [RT] Updates to the provisioning model - maven coordinates

>One minor thing, the current proposal uses the Maven coordinates to >describe artifacts. It's using the path part of a maven url, so >basically group id, artifact id, version are separated by a slash. It >seems that the notation of separating these things by a colon is more >common/natural. And

[RT] Updates to the provisioning model

Hi, as some of you might already have seen or heard during adaptTo, the whiteboard contains a new feature based model at [1] which is intended to be a better provisioning model. The repository at [1] contains a readme which details the reasons and is a basic description of the approach. But I