devicemap-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Werner Keil <>
Subject Re: Let's be clear (was: Should we implement the W3C DDR standard or not?)
Date Tue, 06 Jan 2015 19:44:56 GMT

Many of the earlier discussions should have been marked "internal crap" I
agree or better not even started in the first place;-)

Some terms like "classifier" (occasionally I called this module "parser"
because that's exactly what it does while the "simpleddr" module reads most
XML files in a slightly different way, closer to say how Spring Framework
uses its XML files, there are embedded class references for its DI
mechanism) may not be clear to external bystanders, but in the DeviceMap
codebase that's exactly how this module is named. "DeviceMap Java Client"
is a label but since it's not the only client or library to access the data
it sounds more confusing than "classifier" which is exactly the name of the
Maven module. started to explain the
problem of the "command line console" built into the "classifier" client.
You see in the thread, that Volkan (as a relatively new participant in the
project via JIRA tickets or mailing list) also confused the built in Main
class (command line client) with separate Console examples to use the
classifier client from the command line, too.

Happy to create a separate thread on this list on how to separate them,
because the current duality (the initial example was provided by Reza, too)
makes no real sense.


On Tue, Jan 6, 2015 at 7:04 PM, Bertrand Delacretaz <>

> (I'm sorry for people who are trying to follow this project - maybe we
> should label such discussions as "internal crap" or something ;-)
> Werner,
> I'm sorry to pick on you and I don't usually use top-quoting, but to
> me the below message is almost impossible to parse and understand.
> It mixes at least three different topics, makes vague references to
> things that can be interpreted in many different ways ("the
> classifier", "current data", "admin console", "MIS implementation"
> etc.) and is generally very hard to read and understand for me.
> Communicating on our lists is hard, IMO we need to be clearer and way
> more concise and specific to avoid wasting each other's time. I tried
> to explain that yesterday at [1] - the longer and more complex our
> discussion threads, the lower the chances are that we'll understand
> each other.
> -Bertrand
> [1]
> On Tue, Jan 6, 2015 at 5:29 PM, Werner Keil <> wrote:
> > Nobody says we should stop or discontinue the classifier. It's an
> alternate
> > (though not W3C compliant) way to access the data file in the current
> form.
> > So for the current data I suggest we should release them in parallel. And
> > what you proposed (a bit prematurely) goes along the thread before and
> > matches what I said is possible, so on the long run Classifier should
> > ideally be more modular (also with regards to e.g. the "admin console")
> and
> > a module may fully implement the W3C API and behave as "drop-in
> > replacement" in an app that used another (fully) compliant
> implementation.
> > See
> for
> > the MIS implementation
> >
> >

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message