Release announcement draft

[Kostas Tzoumas-2]

Hey folks,

The Flink 0.6 release is out, and it would be great to announce it at the
Flink blog and perhaps at announcements@apache among other places. I
drafted a small text that also explains a bit what is Flink since this is
the first Apache release.

Would be great to get some comments on the text! Also, if you contributed
in the release and your name is missing, please respond.

Best,
Kostas

We are happy to announce the availability of Flink 0.6. This is the
first release of the system inside the Apache Incubator and under the
name Flink. Releases up to 0.5 were under the name Stratosphere, from
the academic and open source project that Flink originates from.

What is Flink?

Apache Flink is a new general-purpose computation engine that can run
on YARN clusters on top of data stored in Hadoop. Flink currently has
programming APIs in Java and Scala. Jobs are executed via Flink's own
runtime engine. Flink features:

- Robust in-memory and out-of-core processing: once read, data stays
  in memory as much as possible, and is gracefully de-staged to disk in
  the presence of memory pressure from limited memory or other
  applications. The runtime is designed to perform very well both in
  setups with abundant memory and in setups where memory is scarce.

- Object-based APIs: when programming, you do not have to pack your
  data into key-value pairs or some other framework-specific data
  model. Rather, you can use arbitrary Java and Scala types to model
  your data.

- A modular system stack: Flink is not a direct implementation of its
  APIs but a layered system. All programming APIs are translated to an
  intermediate program representation that is compiled and optimized
  via a cost-based optimizer. Lower-level layers of Flink also expose
  programming APIs for extending the system.

- Data pipelining/streaming: Flink's runtime is designed as a
  pipelined data processing engine rather than a batch processing
  engine. Operators do not wait for their predecessors to finish in
  order to start processing data. This results to very efficient
  handling of large intermediate data sets.


Release 0.6

Flink 0.6 builds on the latest Stratosphere 0.5 release. It includes
many bug fixes and improvements that make the system more stable and
robust, as well as breaking API changes.
The full release notes are available here:
https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315522&version=12327101

To download the release please visit
http://flink.incubator.apache.org/downloads.html

Contributors

Wilson Cao
Ufuk Celebi
Stephan Ewen
Jonathan Hasenberg
Markus Holzemer
Fabian Hueske
Sebastian Kunert
Vikhyat Korrapati
Aljoscha Krettek
Sebastian Kruse
Raymond Liu
Robert Metzger
Minglianh Qi
Till Rohrmann
Henry Saputra
Chesnay Schepler
Kostas Tzoumas
Robert Waury
Timo Walther
Daniel Warneke
Tobias Wiens

[Robert Metzger]

Well written, thank you!
I think you have a typo in Mingliangs name: "Mingliang Qi", Jonathan's last
name is "Hasenburg"


On Tue, Aug 26, 2014 at 3:42 PM, Kostas Tzoumas <[hidden email]> wrote:

> Hey folks,
>
> The Flink 0.6 release is out, and it would be great to announce it at the
> Flink blog and perhaps at announcements@apache among other places. I
> drafted a small text that also explains a bit what is Flink since this is
> the first Apache release.
>
> Would be great to get some comments on the text! Also, if you contributed
> in the release and your name is missing, please respond.
>
> Best,
> Kostas
>
> We are happy to announce the availability of Flink 0.6. This is the
> first release of the system inside the Apache Incubator and under the
> name Flink. Releases up to 0.5 were under the name Stratosphere, from
> the academic and open source project that Flink originates from.
>
> What is Flink?
>
> Apache Flink is a new general-purpose computation engine that can run
> on YARN clusters on top of data stored in Hadoop. Flink currently has
> programming APIs in Java and Scala. Jobs are executed via Flink's own
> runtime engine. Flink features:
>
> - Robust in-memory and out-of-core processing: once read, data stays
>   in memory as much as possible, and is gracefully de-staged to disk in
>   the presence of memory pressure from limited memory or other
>   applications. The runtime is designed to perform very well both in
>   setups with abundant memory and in setups where memory is scarce.
>
> - Object-based APIs: when programming, you do not have to pack your
>   data into key-value pairs or some other framework-specific data
>   model. Rather, you can use arbitrary Java and Scala types to model
>   your data.
>
> - A modular system stack: Flink is not a direct implementation of its
>   APIs but a layered system. All programming APIs are translated to an
>   intermediate program representation that is compiled and optimized
>   via a cost-based optimizer. Lower-level layers of Flink also expose
>   programming APIs for extending the system.
>
> - Data pipelining/streaming: Flink's runtime is designed as a
>   pipelined data processing engine rather than a batch processing
>   engine. Operators do not wait for their predecessors to finish in
>   order to start processing data. This results to very efficient
>   handling of large intermediate data sets.
>
>
> Release 0.6
>
> Flink 0.6 builds on the latest Stratosphere 0.5 release. It includes
> many bug fixes and improvements that make the system more stable and
> robust, as well as breaking API changes.
> The full release notes are available here:
>
> https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315522&version=12327101
>
> To download the release please visit
> http://flink.incubator.apache.org/downloads.html
>
> Contributors
>
> Wilson Cao
> Ufuk Celebi
> Stephan Ewen
> Jonathan Hasenberg
> Markus Holzemer
> Fabian Hueske
> Sebastian Kunert
> Vikhyat Korrapati
> Aljoscha Krettek
> Sebastian Kruse
> Raymond Liu
> Robert Metzger
> Minglianh Qi
> Till Rohrmann
> Henry Saputra
> Chesnay Schepler
> Kostas Tzoumas
> Robert Waury
> Timo Walther
> Daniel Warneke
> Tobias Wiens
>

[Kostas Tzoumas-2]

Oops, thank you, sorry for the typos.


On Tue, Aug 26, 2014 at 3:57 PM, Robert Metzger <[hidden email]> wrote:

> Well written, thank you!
> I think you have a typo in Mingliangs name: "Mingliang Qi", Jonathan's last
> name is "Hasenburg"
>
>
> On Tue, Aug 26, 2014 at 3:42 PM, Kostas Tzoumas <[hidden email]>
> wrote:
>
> > Hey folks,
> >
> > The Flink 0.6 release is out, and it would be great to announce it at the
> > Flink blog and perhaps at announcements@apache among other places. I
> > drafted a small text that also explains a bit what is Flink since this is
> > the first Apache release.
> >
> > Would be great to get some comments on the text! Also, if you contributed
> > in the release and your name is missing, please respond.
> >
> > Best,
> > Kostas
> >
> > We are happy to announce the availability of Flink 0.6. This is the
> > first release of the system inside the Apache Incubator and under the
> > name Flink. Releases up to 0.5 were under the name Stratosphere, from
> > the academic and open source project that Flink originates from.
> >
> > What is Flink?
> >
> > Apache Flink is a new general-purpose computation engine that can run
> > on YARN clusters on top of data stored in Hadoop. Flink currently has
> > programming APIs in Java and Scala. Jobs are executed via Flink's own
> > runtime engine. Flink features:
> >
> > - Robust in-memory and out-of-core processing: once read, data stays
> >   in memory as much as possible, and is gracefully de-staged to disk in
> >   the presence of memory pressure from limited memory or other
> >   applications. The runtime is designed to perform very well both in
> >   setups with abundant memory and in setups where memory is scarce.
> >
> > - Object-based APIs: when programming, you do not have to pack your
> >   data into key-value pairs or some other framework-specific data
> >   model. Rather, you can use arbitrary Java and Scala types to model
> >   your data.
> >
> > - A modular system stack: Flink is not a direct implementation of its
> >   APIs but a layered system. All programming APIs are translated to an
> >   intermediate program representation that is compiled and optimized
> >   via a cost-based optimizer. Lower-level layers of Flink also expose
> >   programming APIs for extending the system.
> >
> > - Data pipelining/streaming: Flink's runtime is designed as a
> >   pipelined data processing engine rather than a batch processing
> >   engine. Operators do not wait for their predecessors to finish in
> >   order to start processing data. This results to very efficient
> >   handling of large intermediate data sets.
> >
> >
> > Release 0.6
> >
> > Flink 0.6 builds on the latest Stratosphere 0.5 release. It includes
> > many bug fixes and improvements that make the system more stable and
> > robust, as well as breaking API changes.
> > The full release notes are available here:
> >
> >
> https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315522&version=12327101
> >
> > To download the release please visit
> > http://flink.incubator.apache.org/downloads.html
> >
> > Contributors
> >
> > Wilson Cao
> > Ufuk Celebi
> > Stephan Ewen
> > Jonathan Hasenberg
> > Markus Holzemer
> > Fabian Hueske
> > Sebastian Kunert
> > Vikhyat Korrapati
> > Aljoscha Krettek
> > Sebastian Kruse
> > Raymond Liu
> > Robert Metzger
> > Minglianh Qi
> > Till Rohrmann
> > Henry Saputra
> > Chesnay Schepler
> > Kostas Tzoumas
> > Robert Waury
> > Timo Walther
> > Daniel Warneke
> > Tobias Wiens
> >
>

[Aljoscha Krettek-2]

I like it. :D


On Tue, Aug 26, 2014 at 3:58 PM, Kostas Tzoumas <[hidden email]> wrote:

> Oops, thank you, sorry for the typos.
>
>
> On Tue, Aug 26, 2014 at 3:57 PM, Robert Metzger <[hidden email]>
> wrote:
>
> > Well written, thank you!
> > I think you have a typo in Mingliangs name: "Mingliang Qi", Jonathan's
> last
> > name is "Hasenburg"
> >
> >
> > On Tue, Aug 26, 2014 at 3:42 PM, Kostas Tzoumas <[hidden email]>
> > wrote:
> >
> > > Hey folks,
> > >
> > > The Flink 0.6 release is out, and it would be great to announce it at
> the
> > > Flink blog and perhaps at announcements@apache among other places. I
> > > drafted a small text that also explains a bit what is Flink since this
> is
> > > the first Apache release.
> > >
> > > Would be great to get some comments on the text! Also, if you
> contributed
> > > in the release and your name is missing, please respond.
> > >
> > > Best,
> > > Kostas
> > >
> > > We are happy to announce the availability of Flink 0.6. This is the
> > > first release of the system inside the Apache Incubator and under the
> > > name Flink. Releases up to 0.5 were under the name Stratosphere, from
> > > the academic and open source project that Flink originates from.
> > >
> > > What is Flink?
> > >
> > > Apache Flink is a new general-purpose computation engine that can run
> > > on YARN clusters on top of data stored in Hadoop. Flink currently has
> > > programming APIs in Java and Scala. Jobs are executed via Flink's own
> > > runtime engine. Flink features:
> > >
> > > - Robust in-memory and out-of-core processing: once read, data stays
> > >   in memory as much as possible, and is gracefully de-staged to disk in
> > >   the presence of memory pressure from limited memory or other
> > >   applications. The runtime is designed to perform very well both in
> > >   setups with abundant memory and in setups where memory is scarce.
> > >
> > > - Object-based APIs: when programming, you do not have to pack your
> > >   data into key-value pairs or some other framework-specific data
> > >   model. Rather, you can use arbitrary Java and Scala types to model
> > >   your data.
> > >
> > > - A modular system stack: Flink is not a direct implementation of its
> > >   APIs but a layered system. All programming APIs are translated to an
> > >   intermediate program representation that is compiled and optimized
> > >   via a cost-based optimizer. Lower-level layers of Flink also expose
> > >   programming APIs for extending the system.
> > >
> > > - Data pipelining/streaming: Flink's runtime is designed as a
> > >   pipelined data processing engine rather than a batch processing
> > >   engine. Operators do not wait for their predecessors to finish in
> > >   order to start processing data. This results to very efficient
> > >   handling of large intermediate data sets.
> > >
> > >
> > > Release 0.6
> > >
> > > Flink 0.6 builds on the latest Stratosphere 0.5 release. It includes
> > > many bug fixes and improvements that make the system more stable and
> > > robust, as well as breaking API changes.
> > > The full release notes are available here:
> > >
> > >
> >
> https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315522&version=12327101
> > >
> > > To download the release please visit
> > > http://flink.incubator.apache.org/downloads.html
> > >
> > > Contributors
> > >
> > > Wilson Cao
> > > Ufuk Celebi
> > > Stephan Ewen
> > > Jonathan Hasenberg
> > > Markus Holzemer
> > > Fabian Hueske
> > > Sebastian Kunert
> > > Vikhyat Korrapati
> > > Aljoscha Krettek
> > > Sebastian Kruse
> > > Raymond Liu
> > > Robert Metzger
> > > Minglianh Qi
> > > Till Rohrmann
> > > Henry Saputra
> > > Chesnay Schepler
> > > Kostas Tzoumas
> > > Robert Waury
> > > Timo Walther
> > > Daniel Warneke
> > > Tobias Wiens
> > >
> >
>

[Stephan Ewen]

I like it, too

[Vasiliki Kalavri]

/me likes it too :)

Just a minor comment, in the end of the first paragraph:
"Releases up to 0.5 were under the name Stratosphere, from
the academic and open source project that Flink originates from
​"
​I don't think you need the second "from".

V.


On 26 August 2014 16:02, Stephan Ewen <[hidden email]> wrote:

> I like it, too
>

[Sebastian Schelter]

Sounds very good. I would just change "large intermediate datasets" to
"large datasets"

Best,
Sebastian


2014-08-26 7:29 GMT-07:00 Vasiliki Kalavri <[hidden email]>:

> /me likes it too :)
>
> Just a minor comment, in the end of the first paragraph:
> "Releases up to 0.5 were under the name Stratosphere, from
> the academic and open source project that Flink originates from
> ​"
> ​I don't think you need the second "from".
>
> V.
>
>
> On 26 August 2014 16:02, Stephan Ewen <[hidden email]> wrote:
>
> > I like it, too
> >
>

[Fabian Hueske]

well written.
I would rewrite the first sentence of the "What is Flink?" section to
something like:

"Apache Flink is a general-purpose data processing engine for clusters. It
runs on Hadoop components such as HDFS and YARN, as well as stand-alone."

Rational:
- Flink isn't new (and immature). The code is out for several years.
- Compute engine might also be understood as something like an MPI
framework.
- I would emphasize that we do not require HDFS and YARN, but play nice
with it.

Do we want to highlight the iterative data flows?


2014-08-26 16:02 GMT+02:00 Stephan Ewen <[hidden email]>:

> I like it, too
>

[Stephan Ewen]

Good idea to point out that it supports iterative flows efficiently...


On Tue, Aug 26, 2014 at 4:33 PM, Fabian Hueske <[hidden email]> wrote:

> well written.
> I would rewrite the first sentence of the "What is Flink?" section to
> something like:
>
> "Apache Flink is a general-purpose data processing engine for clusters. It
> runs on Hadoop components such as HDFS and YARN, as well as stand-alone."
>
> Rational:
> - Flink isn't new (and immature). The code is out for several years.
> - Compute engine might also be understood as something like an MPI
> framework.
> - I would emphasize that we do not require HDFS and YARN, but play nice
> with it.
>
> Do we want to highlight the iterative data flows?
>
>
> 2014-08-26 16:02 GMT+02:00 Stephan Ewen <[hidden email]>:
>
> > I like it, too
> >
>

[Sean Owen]

(A minor minor comment then -- the first 'from' should go, but the
second is necessary to the meaning. Flink hasn't originated
something.)

On Tue, Aug 26, 2014 at 3:29 PM, Vasiliki Kalavri
<[hidden email]> wrote:
> /me likes it too :)
>
> Just a minor comment, in the end of the first paragraph:
> "Releases up to 0.5 were under the name Stratosphere, from
> the academic and open source project that Flink originates from
> "
> I don't think you need the second "from".

[Kostas Tzoumas-2]

Thanks everyone! I uploaded a changed post here:
http://flink.incubator.apache.org/news/2014/08/26/release-0.6.html


On Tue, Aug 26, 2014 at 4:42 PM, Sean Owen <[hidden email]> wrote:

> (A minor minor comment then -- the first 'from' should go, but the
> second is necessary to the meaning. Flink hasn't originated
> something.)
>
> On Tue, Aug 26, 2014 at 3:29 PM, Vasiliki Kalavri
> <[hidden email]> wrote:
> > /me likes it too :)
> >
> > Just a minor comment, in the end of the first paragraph:
> > "Releases up to 0.5 were under the name Stratosphere, from
> > the academic and open source project that Flink originates from
> > "
> > I don't think you need the second "from".
>

[Sebastian Schelter]

One more thing to consider: I would change Object-based APIs to POJO-based
APIs as this is a known term in the developing world. Just a thought.

Best,
Sebastian


2014-08-26 9:15 GMT-07:00 Kostas Tzoumas <[hidden email]>:

> Thanks everyone! I uploaded a changed post here:
> http://flink.incubator.apache.org/news/2014/08/26/release-0.6.html
>
>
> On Tue, Aug 26, 2014 at 4:42 PM, Sean Owen <[hidden email]> wrote:
>
> > (A minor minor comment then -- the first 'from' should go, but the
> > second is necessary to the meaning. Flink hasn't originated
> > something.)
> >
> > On Tue, Aug 26, 2014 at 3:29 PM, Vasiliki Kalavri
> > <[hidden email]> wrote:
> > > /me likes it too :)
> > >
> > > Just a minor comment, in the end of the first paragraph:
> > > "Releases up to 0.5 were under the name Stratosphere, from
> > > the academic and open source project that Flink originates from
> > > "
> > > I don't think you need the second "from".
> >
>

[Till Rohrmann]

I like the announcement :-) Well written Kostas!


On Tue, Aug 26, 2014 at 4:42 PM, Sean Owen <[hidden email]> wrote:

> (A minor minor comment then -- the first 'from' should go, but the
> second is necessary to the meaning. Flink hasn't originated
> something.)
>
> On Tue, Aug 26, 2014 at 3:29 PM, Vasiliki Kalavri
> <[hidden email]> wrote:
> > /me likes it too :)
> >
> > Just a minor comment, in the end of the first paragraph:
> > "Releases up to 0.5 were under the name Stratosphere, from
> > the academic and open source project that Flink originates from
> > "
> > I don't think you need the second "from".
>

[Kostas Tzoumas-2]

Hi everyone,

One more question: how about posting an announcement at [hidden email]
since this is the first release? Does any of the mentors know the process
to do that?

Kostas


On Tue, Aug 26, 2014 at 6:18 PM, Till Rohrmann <[hidden email]> wrote:

> I like the announcement :-) Well written Kostas!
>
>
> On Tue, Aug 26, 2014 at 4:42 PM, Sean Owen <[hidden email]> wrote:
>
> > (A minor minor comment then -- the first 'from' should go, but the
> > second is necessary to the meaning. Flink hasn't originated
> > something.)
> >
> > On Tue, Aug 26, 2014 at 3:29 PM, Vasiliki Kalavri
> > <[hidden email]> wrote:
> > > /me likes it too :)
> > >
> > > Just a minor comment, in the end of the first paragraph:
> > > "Releases up to 0.5 were under the name Stratosphere, from
> > > the academic and open source project that Flink originates from
> > > "
> > > I don't think you need the second "from".
> >
>

[Henry Saputra]

+1 looks great! Thanks for the write up Kostas.

Love to add the iterative programming support as one of the
differentiator feature.

- Henry

On Tue, Aug 26, 2014 at 6:42 AM, Kostas Tzoumas <[hidden email]> wrote:

> Hey folks,
>
> The Flink 0.6 release is out, and it would be great to announce it at the
> Flink blog and perhaps at announcements@apache among other places. I
> drafted a small text that also explains a bit what is Flink since this is
> the first Apache release.
>
> Would be great to get some comments on the text! Also, if you contributed
> in the release and your name is missing, please respond.
>
> Best,
> Kostas
>
> We are happy to announce the availability of Flink 0.6. This is the
> first release of the system inside the Apache Incubator and under the
> name Flink. Releases up to 0.5 were under the name Stratosphere, from
> the academic and open source project that Flink originates from.
>
> What is Flink?
>
> Apache Flink is a new general-purpose computation engine that can run
> on YARN clusters on top of data stored in Hadoop. Flink currently has
> programming APIs in Java and Scala. Jobs are executed via Flink's own
> runtime engine. Flink features:
>
> - Robust in-memory and out-of-core processing: once read, data stays
>   in memory as much as possible, and is gracefully de-staged to disk in
>   the presence of memory pressure from limited memory or other
>   applications. The runtime is designed to perform very well both in
>   setups with abundant memory and in setups where memory is scarce.
>
> - Object-based APIs: when programming, you do not have to pack your
>   data into key-value pairs or some other framework-specific data
>   model. Rather, you can use arbitrary Java and Scala types to model
>   your data.
>
> - A modular system stack: Flink is not a direct implementation of its
>   APIs but a layered system. All programming APIs are translated to an
>   intermediate program representation that is compiled and optimized
>   via a cost-based optimizer. Lower-level layers of Flink also expose
>   programming APIs for extending the system.
>
> - Data pipelining/streaming: Flink's runtime is designed as a
>   pipelined data processing engine rather than a batch processing
>   engine. Operators do not wait for their predecessors to finish in
>   order to start processing data. This results to very efficient
>   handling of large intermediate data sets.
>
>
> Release 0.6
>
> Flink 0.6 builds on the latest Stratosphere 0.5 release. It includes
> many bug fixes and improvements that make the system more stable and
> robust, as well as breaking API changes.
> The full release notes are available here:
> https://issues.apache.org/jira/secure/ReleaseNote.jspa?projectId=12315522&version=12327101
>
> To download the release please visit
> http://flink.incubator.apache.org/downloads.html
>
> Contributors
>
> Wilson Cao
> Ufuk Celebi
> Stephan Ewen
> Jonathan Hasenberg
> Markus Holzemer
> Fabian Hueske
> Sebastian Kunert
> Vikhyat Korrapati
> Aljoscha Krettek
> Sebastian Kruse
> Raymond Liu
> Robert Metzger
> Minglianh Qi
> Till Rohrmann
> Henry Saputra
> Chesnay Schepler
> Kostas Tzoumas
> Robert Waury
> Timo Walther
> Daniel Warneke
> Tobias Wiens

[Henry Saputra]

HI Kostas,

As far as I know, you can send the announcement to [hidden email]
(cc dev at flink and general @) by using your @apache.org account.

- Henry

On Tue, Aug 26, 2014 at 9:32 AM, Kostas Tzoumas <[hidden email]> wrote:

> Hi everyone,
>
> One more question: how about posting an announcement at [hidden email]
> since this is the first release? Does any of the mentors know the process
> to do that?
>
> Kostas
>
>
> On Tue, Aug 26, 2014 at 6:18 PM, Till Rohrmann <[hidden email]> wrote:
>
>> I like the announcement :-) Well written Kostas!
>>
>>
>> On Tue, Aug 26, 2014 at 4:42 PM, Sean Owen <[hidden email]> wrote:
>>
>> > (A minor minor comment then -- the first 'from' should go, but the
>> > second is necessary to the meaning. Flink hasn't originated
>> > something.)
>> >
>> > On Tue, Aug 26, 2014 at 3:29 PM, Vasiliki Kalavri
>> > <[hidden email]> wrote:
>> > > /me likes it too :)
>> > >
>> > > Just a minor comment, in the end of the first paragraph:
>> > > "Releases up to 0.5 were under the name Stratosphere, from
>> > > the academic and open source project that Flink originates from
>> > > "
>> > > I don't think you need the second "from".
>> >
>>

[Henry Saputra]

Some tips on posting the announcement:

Be sure to a) use plain text, b) use the word [ANNOUNCE] or [ANN] in
the subject line, and c) include a link for more information/download.

Also do include the word "(Incubating)" after the first mention of
Apache Flink, to read "Apache Flink (Incubating)".

- Henry

On Tue, Aug 26, 2014 at 10:27 AM, Henry Saputra <[hidden email]> wrote:

> HI Kostas,
>
> As far as I know, you can send the announcement to [hidden email]
> (cc dev at flink and general @) by using your @apache.org account.
>
> - Henry
>
> On Tue, Aug 26, 2014 at 9:32 AM, Kostas Tzoumas <[hidden email]> wrote:
>> Hi everyone,
>>
>> One more question: how about posting an announcement at [hidden email]
>> since this is the first release? Does any of the mentors know the process
>> to do that?
>>
>> Kostas
>>
>>
>> On Tue, Aug 26, 2014 at 6:18 PM, Till Rohrmann <[hidden email]> wrote:
>>
>>> I like the announcement :-) Well written Kostas!
>>>
>>>
>>> On Tue, Aug 26, 2014 at 4:42 PM, Sean Owen <[hidden email]> wrote:
>>>
>>> > (A minor minor comment then -- the first 'from' should go, but the
>>> > second is necessary to the meaning. Flink hasn't originated
>>> > something.)
>>> >
>>> > On Tue, Aug 26, 2014 at 3:29 PM, Vasiliki Kalavri
>>> > <[hidden email]> wrote:
>>> > > /me likes it too :)
>>> > >
>>> > > Just a minor comment, in the end of the first paragraph:
>>> > > "Releases up to 0.5 were under the name Stratosphere, from
>>> > > the academic and open source project that Flink originates from
>>> > > "
>>> > > I don't think you need the second "from".
>>> >
>>>