Ruby client and optimized data

[Werner Keil <we...@gmail.com>]

Guys,

Since I have barely ever touched Ruby code myself (there are so many other
languages not just on the JVM, nobody has to do everything at once, at
least not in a proper team where members try to work together instead of
against each other;-) I asked the author of the Ruby port and that's what
he responded:

Hi Werner

Thank you. This gem contains necessary Device Data files.
https://github.com/soylent/device_map/tree/master/lib/resources

When new version of Device Data is released, I simply update these resource
files, run tests, and release a new gem version. There is no need to
download Device Data files separately because I can just provide them
together with the gem. And its size in this case is only 119 KB.

During gem installation I also parse and generate “optimized” device data
files for later use. This improves performance significantly since we don’t
need to parse the XML files every time during initialization.

---
I know, at least Eberhard and I discussed a similar form of optimization he
tried at least against pre-DeviceMap OpenDDR XML files by putting them into
a NoSQL DB. So far it remained a theory for DeviceMap like many other ideas
or discussions. As you see some folks outside the core project (and ASF) do
these things already. And as a side-effect should "Reza's data" aka 2.x
break their code, the Gem is still valid for the most recent 1.x release
available.

Some commercial vendors also mention products and services where the
repository runs in e.g. SQLite, the same system Android also comes with.

Great to hear. Similar to e.g. Browsermap it runs on its own Travis CI
instance so the green bar on the project page shows, everything is in
order.I'm not sure, if Konstantin saw value in contributing any of that to
DeviceMap. It can't hurt to ask, Again, please speak up if you constantly
code in Ruby, at least a license change would not be necessary as his code
already is under Apache;-)

Werner

Re: Ruby client and optimized data

[Bertrand Delacretaz <bd...@apache.org>]

On Tue, Apr 21, 2015 at 4:30 PM, Werner Keil <we...@gmail.com> wrote:
> ...I asked him to get in touch with Bertrand who also helped with
> this before....

Discussions about code donations are welcome on this list.

-Bertrand

Re: Ruby client and optimized data

[Reza Naghibi <re...@apache.org>]

>> Just stop personal attacks

Fair enough.

On Tue, Apr 21, 2015 at 1:19 PM, Werner Keil <we...@gmail.com> wrote:

> Reza,
>
> Just stop personal attacks like you always do with neither technical or
> procedual experience or anything noteworthy to show.
>
> I am not only in the industry for over 25 years now, so I wrote software
> before you and others were born or still (sorry but you just started;-)
> pooped your pants.
> Experienced in programming languages some of you neither now nor heard of
> (maybe except Bertrand or others of my generation)
>
> From leading or mentoring projects to simply contributing I participate in
> a dozen widely accepted standards at JCP, Oasis, OGC or other standard
> bodies. And at least as many open source projects. Except Konstantin the
> author of the Ruby port every contributor shows more activity and has
> therefore more followers on GitHub.
> Contributions in the last year22 totalApr 21, 2014 – Apr 21, 2015
>
> 22 total per year.Wow, that's impressive. I simply let the numbers speak:
> Contributions in the last year1,669 totalApr 21, 2014 – Apr 21, 2015
> Longest streak22 daysFebruary 3 – February 24
>
> my longest streak was 22 days, not 22 contributions.
>
> If you do this in a silent chamber, then you don't have to work with other
> people. And archives on this mailing list show how "well" your team
> abilities really are.
>
> So please shut up and stop being disrespectful as you have been several
> times before.
>
> Cheers
> Werner
>
>
>
> On Tue, Apr 21, 2015 at 6:24 PM, Reza Naghibi <re...@apache.org> wrote:
>
> > http://s.apache.org/L57
> >
> > Werner,
> >
> > I think you are the last person who should be lecturing this project
> > regarding how versioning and support works. You have shown tremendous
> > incompetence regarding all aspects of software dev, releasing,
> lifecycles,
> > and this project as a whole. How you have made it this far in the
> industry
> > is amazing... When it comes to decisions regarding how we support old
> > versions and how we move onto better and greater things like JSON, I will
> > use my own judgement and listen to the feedback from members of this
> > project and people on the list. I have been writing, releasing, and
> > supporting software through myself and many many companies and
> > organizations for 15 years. I think I know what I am doing here and as an
> > added plus, I know how to listen, accept feedback, and work as a team.
> >
> > If you want to control this project, start contributing in meaningful
> ways.
> > Bossing people around, bickering and arguing with me at every chance you
> > get, and spewing nonsense on this list accomplishes nothing. Otherwise,
> > maybe just sit back and shut up for a few months?
> >
> > On Tue, Apr 21, 2015 at 11:22 AM, Werner Keil <we...@gmail.com>
> > wrote:
> >
> > > Good to know. I don't recall he mentioned being subscribed to this
> > mailing
> > > list, but he would have to subscribe to post, I suppose.
> > >
> > > The current project (whether "device_map" was a legal name on GitHub or
> > > not, I trust Bertrand can advise Konstantin) is based on the 1.0.x
> data,
> > > and according to Travis CI makes a stable impression. With its Ruby
> gems
> > it
> > > is fairly self contained even if DeviceMap was to stop exposing 1.x
> via a
> > > URL (the downloads are always there anyway) so I guess everyone
> involved
> > > can figure out, if a 1.x "clients" contribution for Ruby makes sense or
> > > wait for 2.x.
> > >
> > > Werner
> > >
> > > On Tue, Apr 21, 2015 at 4:53 PM, Reza Naghibi <re...@apache.org>
> wrote:
> > >
> > > > Just for reference:
> > > >
> > > > http://s.apache.org/4ES
> > > >
> > > > Remember, the context for this is 2.0.
> > > >
> > > >
> > > > On Tue, Apr 21, 2015 at 10:41 AM, Reza Naghibi <re...@apache.org>
> > wrote:
> > > >
> > > > > Werner,
> > > > >
> > > > > Konstantin is already listed as a potential Ruby client contributor
> > for
> > > > > 2.0.
> > > > >
> > > > >
> > > > >
> > > >
> > >
> >
> https://svn.apache.org/viewvc/devicemap/branches/2.0/classifiers/README?view=markup
> > > > >
> > > > > This list is non binding, its just being used for organizing 2.0.
> > For a
> > > > > new client to be brought it, it needs to be written, pass the
> needed
> > > > > specification tests, the contributor needs to demonstrate they can
> > > > release
> > > > > and maintain said client, and finally the contributor and client
> > needs
> > > to
> > > > > be voted in by the PMC (standard committer vote process). Im my
> > > opinion,
> > > > > this is how this should work.
> > > > >
> > > >
> > >
> >
>

Re: Ruby client and optimized data

[Werner Keil <we...@gmail.com>]

Reza,

Just stop personal attacks like you always do with neither technical or
procedual experience or anything noteworthy to show.

I am not only in the industry for over 25 years now, so I wrote software
before you and others were born or still (sorry but you just started;-)
pooped your pants.
Experienced in programming languages some of you neither now nor heard of
(maybe except Bertrand or others of my generation)

>From leading or mentoring projects to simply contributing I participate in
a dozen widely accepted standards at JCP, Oasis, OGC or other standard
bodies. And at least as many open source projects. Except Konstantin the
author of the Ruby port every contributor shows more activity and has
therefore more followers on GitHub.
Contributions in the last year22 totalApr 21, 2014 – Apr 21, 2015

22 total per year.Wow, that's impressive. I simply let the numbers speak:
Contributions in the last year1,669 totalApr 21, 2014 – Apr 21, 2015
Longest streak22 daysFebruary 3 – February 24

my longest streak was 22 days, not 22 contributions.

If you do this in a silent chamber, then you don't have to work with other
people. And archives on this mailing list show how "well" your team
abilities really are.

So please shut up and stop being disrespectful as you have been several
times before.

Cheers
Werner



On Tue, Apr 21, 2015 at 6:24 PM, Reza Naghibi <re...@apache.org> wrote:

> http://s.apache.org/L57
>
> Werner,
>
> I think you are the last person who should be lecturing this project
> regarding how versioning and support works. You have shown tremendous
> incompetence regarding all aspects of software dev, releasing, lifecycles,
> and this project as a whole. How you have made it this far in the industry
> is amazing... When it comes to decisions regarding how we support old
> versions and how we move onto better and greater things like JSON, I will
> use my own judgement and listen to the feedback from members of this
> project and people on the list. I have been writing, releasing, and
> supporting software through myself and many many companies and
> organizations for 15 years. I think I know what I am doing here and as an
> added plus, I know how to listen, accept feedback, and work as a team.
>
> If you want to control this project, start contributing in meaningful ways.
> Bossing people around, bickering and arguing with me at every chance you
> get, and spewing nonsense on this list accomplishes nothing. Otherwise,
> maybe just sit back and shut up for a few months?
>
> On Tue, Apr 21, 2015 at 11:22 AM, Werner Keil <we...@gmail.com>
> wrote:
>
> > Good to know. I don't recall he mentioned being subscribed to this
> mailing
> > list, but he would have to subscribe to post, I suppose.
> >
> > The current project (whether "device_map" was a legal name on GitHub or
> > not, I trust Bertrand can advise Konstantin) is based on the 1.0.x data,
> > and according to Travis CI makes a stable impression. With its Ruby gems
> it
> > is fairly self contained even if DeviceMap was to stop exposing 1.x via a
> > URL (the downloads are always there anyway) so I guess everyone involved
> > can figure out, if a 1.x "clients" contribution for Ruby makes sense or
> > wait for 2.x.
> >
> > Werner
> >
> > On Tue, Apr 21, 2015 at 4:53 PM, Reza Naghibi <re...@apache.org> wrote:
> >
> > > Just for reference:
> > >
> > > http://s.apache.org/4ES
> > >
> > > Remember, the context for this is 2.0.
> > >
> > >
> > > On Tue, Apr 21, 2015 at 10:41 AM, Reza Naghibi <re...@apache.org>
> wrote:
> > >
> > > > Werner,
> > > >
> > > > Konstantin is already listed as a potential Ruby client contributor
> for
> > > > 2.0.
> > > >
> > > >
> > > >
> > >
> >
> https://svn.apache.org/viewvc/devicemap/branches/2.0/classifiers/README?view=markup
> > > >
> > > > This list is non binding, its just being used for organizing 2.0.
> For a
> > > > new client to be brought it, it needs to be written, pass the needed
> > > > specification tests, the contributor needs to demonstrate they can
> > > release
> > > > and maintain said client, and finally the contributor and client
> needs
> > to
> > > > be voted in by the PMC (standard committer vote process). Im my
> > opinion,
> > > > this is how this should work.
> > > >
> > >
> >
>

Re: Ruby client and optimized data

[Reza Naghibi <re...@apache.org>]

http://s.apache.org/L57

Werner,

I think you are the last person who should be lecturing this project
regarding how versioning and support works. You have shown tremendous
incompetence regarding all aspects of software dev, releasing, lifecycles,
and this project as a whole. How you have made it this far in the industry
is amazing... When it comes to decisions regarding how we support old
versions and how we move onto better and greater things like JSON, I will
use my own judgement and listen to the feedback from members of this
project and people on the list. I have been writing, releasing, and
supporting software through myself and many many companies and
organizations for 15 years. I think I know what I am doing here and as an
added plus, I know how to listen, accept feedback, and work as a team.

If you want to control this project, start contributing in meaningful ways.
Bossing people around, bickering and arguing with me at every chance you
get, and spewing nonsense on this list accomplishes nothing. Otherwise,
maybe just sit back and shut up for a few months?

On Tue, Apr 21, 2015 at 11:22 AM, Werner Keil <we...@gmail.com> wrote:

> Good to know. I don't recall he mentioned being subscribed to this mailing
> list, but he would have to subscribe to post, I suppose.
>
> The current project (whether "device_map" was a legal name on GitHub or
> not, I trust Bertrand can advise Konstantin) is based on the 1.0.x data,
> and according to Travis CI makes a stable impression. With its Ruby gems it
> is fairly self contained even if DeviceMap was to stop exposing 1.x via a
> URL (the downloads are always there anyway) so I guess everyone involved
> can figure out, if a 1.x "clients" contribution for Ruby makes sense or
> wait for 2.x.
>
> Werner
>
> On Tue, Apr 21, 2015 at 4:53 PM, Reza Naghibi <re...@apache.org> wrote:
>
> > Just for reference:
> >
> > http://s.apache.org/4ES
> >
> > Remember, the context for this is 2.0.
> >
> >
> > On Tue, Apr 21, 2015 at 10:41 AM, Reza Naghibi <re...@apache.org> wrote:
> >
> > > Werner,
> > >
> > > Konstantin is already listed as a potential Ruby client contributor for
> > > 2.0.
> > >
> > >
> > >
> >
> https://svn.apache.org/viewvc/devicemap/branches/2.0/classifiers/README?view=markup
> > >
> > > This list is non binding, its just being used for organizing 2.0. For a
> > > new client to be brought it, it needs to be written, pass the needed
> > > specification tests, the contributor needs to demonstrate they can
> > release
> > > and maintain said client, and finally the contributor and client needs
> to
> > > be voted in by the PMC (standard committer vote process). Im my
> opinion,
> > > this is how this should work.
> > >
> >
>

Re: Ruby client and optimized data

[Werner Keil <we...@gmail.com>]

Good to know. I don't recall he mentioned being subscribed to this mailing
list, but he would have to subscribe to post, I suppose.

The current project (whether "device_map" was a legal name on GitHub or
not, I trust Bertrand can advise Konstantin) is based on the 1.0.x data,
and according to Travis CI makes a stable impression. With its Ruby gems it
is fairly self contained even if DeviceMap was to stop exposing 1.x via a
URL (the downloads are always there anyway) so I guess everyone involved
can figure out, if a 1.x "clients" contribution for Ruby makes sense or
wait for 2.x.

Werner

On Tue, Apr 21, 2015 at 4:53 PM, Reza Naghibi <re...@apache.org> wrote:

> Just for reference:
>
> http://s.apache.org/4ES
>
> Remember, the context for this is 2.0.
>
>
> On Tue, Apr 21, 2015 at 10:41 AM, Reza Naghibi <re...@apache.org> wrote:
>
> > Werner,
> >
> > Konstantin is already listed as a potential Ruby client contributor for
> > 2.0.
> >
> >
> >
> https://svn.apache.org/viewvc/devicemap/branches/2.0/classifiers/README?view=markup
> >
> > This list is non binding, its just being used for organizing 2.0. For a
> > new client to be brought it, it needs to be written, pass the needed
> > specification tests, the contributor needs to demonstrate they can
> release
> > and maintain said client, and finally the contributor and client needs to
> > be voted in by the PMC (standard committer vote process). Im my opinion,
> > this is how this should work.
> >
>

Re: Ruby client and optimized data

[Reza Naghibi <re...@apache.org>]

Just for reference:

http://s.apache.org/4ES

Remember, the context for this is 2.0.


On Tue, Apr 21, 2015 at 10:41 AM, Reza Naghibi <re...@apache.org> wrote:

> Werner,
>
> Konstantin is already listed as a potential Ruby client contributor for
> 2.0.
>
>
> https://svn.apache.org/viewvc/devicemap/branches/2.0/classifiers/README?view=markup
>
> This list is non binding, its just being used for organizing 2.0. For a
> new client to be brought it, it needs to be written, pass the needed
> specification tests, the contributor needs to demonstrate they can release
> and maintain said client, and finally the contributor and client needs to
> be voted in by the PMC (standard committer vote process). Im my opinion,
> this is how this should work.
>

Re: Ruby client and optimized data

[Reza Naghibi <re...@apache.org>]

Werner,

Konstantin is already listed as a potential Ruby client contributor for 2.0.

https://svn.apache.org/viewvc/devicemap/branches/2.0/classifiers/README?view=markup

This list is non binding, its just being used for organizing 2.0. For a new
client to be brought it, it needs to be written, pass the needed
specification tests, the contributor needs to demonstrate they can release
and maintain said client, and finally the contributor and client needs to
be voted in by the PMC (standard committer vote process). Im my opinion,
this is how this should work.

Re: Ruby client and optimized data

[Werner Keil <we...@gmail.com>]

Hi,

Good news. just follow the thread in the issue tracker, Konstantin replied
there, he would be happy to donate his project "device_map" into a matching
place at DeviceMap if his contribution was welcome (and at least the PMC
also voted on this I assume) Since it's been at least 3 years now, that we
did this from the OpenDDR side as a large chunk to the initial codebase,
and I was not the only OpenDDR contributor involved in administrative
matters, I asked him to get in touch with Bertrand who also helped with
this before.

He already licensed everything under Apache 2, that might make the review
or IP clearing process by relevant legal team easier, than e.g. with the
different licenses OpenDDR used.

If succesful, it would not only win another contributor/committer or even
PMC member (depending on what qualifies those who donate for either role)
but add Ruby to the list of supported languages. Especially in the Social
Media Start-Up scene around the world I know first hand, it is still of
great importance. 2006 it even scored "Programming Language of the Year",
see http://www.tiobe.com/index.php/content/paperinfo/tpci/index.html.
It had its ups and downs, but still remains in the Top 20 list, and taking
just "Web or Mobile relevant" languages out of these, I'd say it must be
still in the Top 10.

Cheers,
Werner

On Tue, Apr 21, 2015 at 12:55 AM, Werner Keil <we...@gmail.com> wrote:

> Guys,
>
> Since I have barely ever touched Ruby code myself (there are so many other
> languages not just on the JVM, nobody has to do everything at once, at
> least not in a proper team where members try to work together instead of
> against each other;-) I asked the author of the Ruby port and that's what
> he responded:
>
> Hi Werner
>
> Thank you. This gem contains necessary Device Data files.
> https://github.com/soylent/device_map/tree/master/lib/resources
>
> When new version of Device Data is released, I simply update these
> resource files, run tests, and release a new gem version. There is no need
> to download Device Data files separately because I can just provide them
> together with the gem. And its size in this case is only 119 KB.
>
> During gem installation I also parse and generate “optimized” device data
> files for later use. This improves performance significantly since we don’t
> need to parse the XML files every time during initialization.
>
> ---
> I know, at least Eberhard and I discussed a similar form of optimization
> he tried at least against pre-DeviceMap OpenDDR XML files by putting them
> into a NoSQL DB. So far it remained a theory for DeviceMap like many other
> ideas or discussions. As you see some folks outside the core project (and
> ASF) do these things already. And as a side-effect should "Reza's data" aka
> 2.x break their code, the Gem is still valid for the most recent 1.x
> release available.
>
> Some commercial vendors also mention products and services where the
> repository runs in e.g. SQLite, the same system Android also comes with.
>
> Great to hear. Similar to e.g. Browsermap it runs on its own Travis CI
> instance so the green bar on the project page shows, everything is in
> order.I'm not sure, if Konstantin saw value in contributing any of that to
> DeviceMap. It can't hurt to ask, Again, please speak up if you constantly
> code in Ruby, at least a license change would not be necessary as his code
> already is under Apache;-)
>
> Werner
>