I agree with you and Jason van Zyl about Maven probably doesn't need to support another option. However, it would be nice if the architecture supported it more easily.
This would mean everything is accessed through a clean API and that we could easily inject our own POM parser. If someone wrote a plugin that didn't use the API's correct and bound to the XML, then the person that injected his own POM parser and POM format would be left to deal with that too. He/She would have to decide if it is worth it for them. I imagine some people would still do it and either sacrifice the use of those plugins or they help to contribute to fix them to work the API more appropriately, thus giving back to the community in a constructive manner. The current problem is that some of the maven code is very nasty and some it isn't always easy to inject your own implementations into it. The way I see it, it is more about building a nice injectable architecture with really good clean API's, then power users can basically do whatever they want easily. Therefore, you wouldn't be directly supporting this feature...but by creating a clean injectable architecture you would support without that intent anyway. This way, the maven team isn't supporting it per se, but rather the architecture supports the ability for someone to do it at their own risk. kind regards, Jason ________________________________________ From: Stephen Connolly [stephen.alan.conno...@gmail.com] Sent: Saturday, September 05, 2009 5:45 AM To: Maven Developers List Cc: Maven Developers List Subject: Re: Re : Re : non-xml poms in 3.x personally, given the fun with rewriting XML at the moment, (see versions maven plugin) I would prefer to just have the current XML format. adding more formats makes some of the things that the versions maven current does a little harder to support. Sent from my [rhymes with myPod] ;-) On 5 Sep 2009, at 12:00, Julien HENRY <henr...@yahoo.fr> wrote: > In the very specific case of groupId/artifactId/version pattern > which is currently very verbose I would tend to agree to allow > shorter syntax using attributes instead of elements. > > <dependency groupId="" artifactId="" version="" classifier="" > scope=""/> > > <plugin groupId="" artifactId="" version=""> > <configuration> > ... > </configuration> > </plugin> > > This is not what I consider a "big" change for endusers. > > Still my 2 cents. > > Regards, > > Julien > > > > ----- Message d'origine ---- >> De : Jason Chaffee <jason.chaf...@zilliontv.tv> >> À : Maven Developers List <dev@maven.apache.org> >> Envoyé le : Samedi, 5 Septembre 2009, 1h00mn 02s >> Objet : RE: Re : non-xml poms in 3.x >> >> FYI, I know that in the past Resin supported both Elements and >> attributes in >> it's config XML. It was really neat. If you preferred one over >> the other, you >> could use it. Don't know if it is still that way though. >> >> Jason >> ________________________________________ >> From: Jason Chaffee [jason.chaf...@zilliontv.tv] >> Sent: Friday, September 04, 2009 3:27 PM >> To: Maven Developers List >> Subject: RE: Re : non-xml poms in 3.x >> >> I like the idea of having some things as attributes. >> >> See the following links on information on when to use attributes >> and when to use >> elements. >> >> http://www.ibm.com/developerworks/xml/library/x-eleatt.html >> http://nedbatchelder.com/blog/200412/elements_vs_attributes.html >> http://www.x12.org/x12org/comments/X12Reference_Model_For_XML_Design.pdf >> >> In particular, from the X12 Reference Model for XML Design: >> >> 7.2.5 Elements vs. Attributes >> Description: Often it is possible to model a data item as a child >> element or an >> attribute. >> >> Benefits of Using Elements >> >> -They are more extensible because attributes can later be added to >> them without >> affecting a processing application. >> -They can contain other elements. For example, if you want to >> express a textual >> description using XHTML tags, this is not possible if description >> is an >> attribute. >> -They can be repeated. An element may only appear once now, but >> later you may >> wish to extend it to appear multiple times. >> -You have more control over the rules of their appearance. For >> example, you can >> say that a product can either have a number or a productCode child. >> This is not >> possible for attributes. >> -Their order is significant if specified as part of a sequence, >> while the order >> of attributes is not. Obviously, this is only an advantage if you >> care about the >> order. >> -When the values are lengthy, elements tend to be more readable >> than attributes. >> >> >> Disadvantages of Using Elements >> >> -Elements require start and end tags, so are therefore more >> verbose. (NOTE: not >> all elements require a start and end tag — elements can be declare >> d in a single >> line.) >> >> Benefits of Using Attributes >> >> -They are less verbose. >> -Attributes can be added to the instance by specifying default >> values. Elements >> cannot (they must appear to receive a default value). >> -Attributes are atomic and cannot be extended and its existence >> should serve to >> remove any and all possible ambiguity of the element it describes. >> They are >> “adjectives” to the element “noun”. >> >> Disadvantages of Using Attributes >> >> -Attributes may not be extended by adding children, whereas a >> complex element >> may be extended by adding additional child elements or attributes. >> -If attributes are to be used in addition to elements for conveying >> business >> data, rules are required for specifying when a specific data item >> shall be an >> element or an attribute. >> >> >> Jason >> ________________________________________ >> From: paulus.benedic...@gmail.com [paulus.benedic...@gmail.com] On >> Behalf Of >> Paul Benedict [pbened...@apache.org] >> Sent: Friday, September 04, 2009 3:05 PM >> To: Maven Developers List >> Subject: Re: Re : non-xml poms in 3.x >> >> Yes, the XML is verbose, and tooling helps but I think most people >> write it >> by hand. The only evolutionary change I support is the ability to >> specify >> simple nested elements as attributes. >> >> Paul >> >> On Fri, Sep 4, 2009 at 4:40 PM, Jason Chaffee wrote: >> >>> For what it is worth, I have heard many complaints either about >>> using XML >>> and/or about the current structure of the XML as well. I have >>> heard this >>> from developers I have worked with and I have read some blogs >>> about this >>> too. I can't tell you where those blogs are now because I pretty >>> much >>> dismissed them as I like using XML myself. >>> >>> Jason >>> ________________________________________ >>> From: Christian Edward Gruber [christianedwardgru...@gmail.com] >>> Sent: Friday, September 04, 2009 2:29 PM >>> To: Maven Developers List >>> Subject: Re: Re : non-xml poms in 3.x >>> >>> Who said anything about a reasonable person? :) I don't have such a >>> hatred - I'm quite used to it, but it has come up in nearly every >>> client in the last 3 years - not as a final or deal-breaking barrier >>> to adoption, but a barrier nonetheless. >>> >>> I'm happy to support it - I just need a seam or hook where I can >>> provide something that pre-processes before the maven run to >>> generate >>> the pom.xml. Maven itself doesn't need to support the alternate >>> format at all. If it could be something that was auto-detected as >>> well (or at worst, put into a settings.xml) then that'd be great. >>> Essentially I'm doing that now with the maven-yamlpom-plugin... it's >>> just that I have to do a separate run because it modifies the >>> pom.xml, >>> and so maven fails on the first sync because the pom was modified. >>> >>> In a pinch, this can all be handled with shell scripts wrapping >>> maven, >>> but I'd prefer to have a cleaner place to integrate. >>> >>> Christian. >>> >>> On Sep 4, 2009, at 5:12 PM, Jason van Zyl wrote: >>> >>>> >>>> On 2009-09-04, at 10:59 PM, Christian Edward Gruber wrote: >>>> >>>>> So I agree that it is a valid concern, and there needs to be a >>>>> canonical format (which will probably be XML) which all artifacts >>>>> are saved as - but in my source tree, it should be entirely >>>>> possible to have an alternate way to specify, since often I've >>>>> found that XML-hatred is a barrier to Maven adoption in some >>>>> firms. >>>>> >>>> >>>> I have not found this to be a concern. There's lots of other things >>>> that are barrier but the XML has honestly never come up in any >>>> conversations I've had. >>>> >>>>> You should always be able to get the pom.xml... help:canonical-pom >>>>> would be a decent way to quickly do it, and artifacts should be >>>>> published that way - but a Project is an object, and alternate >>>>> serializations shouldn't be a problem. >>>> >>>> Therein lies the problem, the only real canonical form is the >>>> object >>>> model. With 3 XML formats which one is the canonical format? The >>>> first one? >>>> >>>>> >>>>> I would, also as an evangelist and implementor of build systems in >>>>> companies, encourage that a company standardize on a format, but >>>>> if >>>>> that company wants to use YAML, or some other terser, more human >>>>> readable format for the pom, then I'm good with that. >>>> >>>> I'm not. Again this falls into my category of "if you want it that >>>> way, you support it." A company can standardize on whatever it >>>> wants, but I'm not going to hide the real cost of that. We may >>>> ultimately decide it's not worth it having another XML format. >>>> There >>>> are a lot more things in 3.0 that interest me then another XML >>>> format. >>>> >>>>> >>>>> I used to feel more like you, Brian, but for the sheer intensity >>>>> of >>>>> hatred of XML out there (as a format for human-maintained source). >>>>> >>>> >>>> Again, I've not witnessed any XML hatred or that being a >>>> justification by a reasonable person not to use Maven. >>>> >>>>> The problem you're describing about one project using one and >>>>> another using a different one - that's no different than one >>>>> project deciding to use a different set of integration test >>>>> plugins >>>>> (invoker vs. shitty) and confusing the noobs. The bottom line is >>>>> that you're not going to be able to constraint people from going >>>>> for the "new thing" and messing up the inexperienced, so providing >>>>> sane defaults with lots of room to customize is the best option, >>>>> in >>>>> my view. >>>>> >>>>> Christian. >>>>> >>>>> >>>>> On Sep 4, 2009, at 4:25 PM, Brian Fox wrote: >>>>> >>>>>>> Just my 2 cents as a Maven evangelist in a big private company. >>>>>>> Even if >>>>>>> Maven is around for years now, basic endusers just start to get >>>>>>> accustomed to pom.xml and Maven philosophy (really! people are >>>>>>> far slowest to change than in OpenSource project team). >>>>>>> >>>>>>> Please, please don't mess everything. Small additions are fine, >>>>>>> but I think a new format is a bad idea even if it is optional. >>>>>>> One of advantage of Maven is standardization across all our >>>>>>> projects. If there are several format allowed, some projects >>>>>>> will >>>>>>> start using new one when others are still using the former and >>>>>>> it >>>>>>> will lead to a total mess. >>>>>>> >>>>>> That's my main concern as well to be honest. >>>>>> >>>>>> --- >>>>>> --- >>>>>> --------------------------------------------------------------- >>>>>> To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org >>>>>> For additional commands, e-mail: dev-h...@maven.apache.org >>>>>> >>>>> >>>>> >>>>> --- >>>>> ------------------------------------------------------------------ >>>>> To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org >>>>> For additional commands, e-mail: dev-h...@maven.apache.org >>>>> >>>> >>>> Thanks, >>>> >>>> Jason >>>> >>>> ---------------------------------------------------------- >>>> Jason van Zyl >>>> Founder, Apache Maven >>>> http://twitter.com/jvanzyl >>>> http://twitter.com/SonatypeNexus >>>> http://twitter.com/SonatypeM2E >>>> ---------------------------------------------------------- >>>> >>>> >>>> --- >>>> ------------------------------------------------------------------ >>>> To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org >>>> For additional commands, e-mail: dev-h...@maven.apache.org >>>> >>> >>> >>> --- >>> ------------------------------------------------------------------ >>> To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org >>> For additional commands, e-mail: dev-h...@maven.apache.org >>> >>> >>> --- >>> ------------------------------------------------------------------ >>> To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org >>> For additional commands, e-mail: dev-h...@maven.apache.org >>> >>> >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org >> For additional commands, e-mail: dev-h...@maven.apache.org >> >> >> --------------------------------------------------------------------- >> To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org >> For additional commands, e-mail: dev-h...@maven.apache.org > > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org > For additional commands, e-mail: dev-h...@maven.apache.org > --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org For additional commands, e-mail: dev-h...@maven.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@maven.apache.org For additional commands, e-mail: dev-h...@maven.apache.org
