BCE HackDay

[aculich]

After various discussions I've had with @davclark , @paciorek , @ck37 , and @ryanlovett , I wonder if it's time to have a BCE HackDay ahead of doing a Fall release? Is there a day (or half day or even just 2 hours) that we could get together in the next couple of weeks? Preferably before the Oct 15th Cloud WG topic which is: BCE on AWS?

Maybe Friday 10/2 or Friday 10/9?

I'll start triaging all the issues we have here and in the documentation issues ahead of that HackDay.

I'd like to address issues like this one that Dav and I discussed on the uc-juptyer slack:

aculich [6:49 PM]
@Davlab Is anaconda the approach used in http://bce.berkeley.edu ?

aculich [6:49 PM]
Whatever the approach, I wonder if we ought to include a short description in the https://github.com/ucberkeley/bce/blob/dev/README.md of the high-level decisions and reasons for making them so that:

aculich [6:49 PM]

• folks can (without looking at the build scripts) see what our approach is; and

aculich [6:49 PM]

• can see our rationale (which won’t necessarily be obvious when looking at the build scripts)
  NEW MESSAGES

----- Yesterday September 20th, 2015 -----

davlab [7:42 PM]
@aculich: yes, we use anaconda now. We aren’t using it as well as we could (i.e., I’d like to enable conda install bce=fall-2015 for at least arbitrary-ish unix). Such a thing is not hard, we just haven’t gotten to it.

davlab [7:43 PM]
I think a great first step would be simply documenting the choices we made even without the rationale! I’ve been meaning to do this, and I think we should explain how to install most of the major tools - like PyCharm, RStudio, anaconda, etc.

[paciorek]

Sounds like a good idea to me. 10/2 and 10/9 both work though I teach
2-3:30.

On Mon, Sep 21, 2015 at 7:45 AM, Aaron Culich notifications@github.com
wrote:

After various discussions I've had with @davclark
https://github.com/davclark , @paciorek https://github.com/paciorek ,
@ck37 https://github.com/ck37 , and @ryanlovett
https://github.com/ryanlovett , I wonder if it's time to have a BCE
HackDay ahead of doing a Fall release? Is there a day (or half day or even
just 2 hours) that we could get together in the next couple of weeks?
Preferably before the Oct 15th Cloud WG topic which is: BCE on AWS?

Maybe Friday 10/2 or Friday 10/9?

I'll start triaging all the issues we have here
https://github.com/ucberkeley/bce/issues and in the documentation issues
https://github.com/ucberkeley/bce/issues ahead of that HackDay.

I'd like to address issues like this one that Dav and I discussed on the uc-juptyer
slack https://uc-jupyter.slack.com/messages/general/:

aculich [6:49 PM]
@Davlab https://github.com/davlab Is anaconda the approach used in
http://bce.berkeley.edu ?

aculich [6:49 PM]
Whatever the approach, I wonder if we ought to include a short description
in the https://github.com/ucberkeley/bce/blob/dev/README.md of the
high-level decisions and reasons for making them so that:

aculich [6:49 PM]

• folks can (without looking at the build scripts) see what our
  approach is; and

aculich [6:49 PM]

• can see our rationale (which won’t necessarily be obvious when
  looking at the build scripts) NEW MESSAGES

----- Yesterday September 20th, 2015 -----

davlab [7:42 PM]
@aculich https://github.com/aculich: yes, we use anaconda now. We
aren’t using it as well as we could (i.e., I’d like to enable conda
install bce=fall-2015 for at least arbitrary-ish unix). Such a thing is
not hard, we just haven’t gotten to it.

davlab [7:43 PM]
I think a great first step would be simply documenting the choices we made
even without the rationale! I’ve been meaning to do this, and I think we
should explain how to install most of the major tools - like PyCharm,
RStudio, anaconda, etc.

—
Reply to this email directly or view it on GitHub
#65.

[ck37]

That would be fun - I’m open 12:30-3pm on 10/2 and 10/9.

On Mon, Sep 21, 2015 at 7:58 AM, Christopher Paciorek <
notifications@github.com> wrote:

Sounds like a good idea to me. 10/2 and 10/9 both work though I teach
2-3:30.

On Mon, Sep 21, 2015 at 7:45 AM, Aaron Culich notifications@github.com
wrote:

After various discussions I've had with @davclark
https://github.com/davclark , @paciorek https://github.com/paciorek
,
@ck37 https://github.com/ck37 , and @ryanlovett
https://github.com/ryanlovett , I wonder if it's time to have a BCE
HackDay ahead of doing a Fall release? Is there a day (or half day or
even
just 2 hours) that we could get together in the next couple of weeks?
Preferably before the Oct 15th Cloud WG topic which is: BCE on AWS?

Maybe Friday 10/2 or Friday 10/9?

I'll start triaging all the issues we have here
https://github.com/ucberkeley/bce/issues and in the documentation
issues
https://github.com/ucberkeley/bce/issues ahead of that HackDay.

I'd like to address issues like this one that Dav and I discussed on the
uc-juptyer
slack https://uc-jupyter.slack.com/messages/general/:

aculich [6:49 PM]
@Davlab https://github.com/davlab Is anaconda the approach used in
http://bce.berkeley.edu ?

aculich [6:49 PM]
Whatever the approach, I wonder if we ought to include a short
description
in the https://github.com/ucberkeley/bce/blob/dev/README.md of the
high-level decisions and reasons for making them so that:

aculich [6:49 PM]

• folks can (without looking at the build scripts) see what our
  approach is; and

aculich [6:49 PM]

• can see our rationale (which won’t necessarily be obvious when
  looking at the build scripts) NEW MESSAGES

----- Yesterday September 20th, 2015 -----

davlab [7:42 PM]
@aculich https://github.com/aculich: yes, we use anaconda now. We
aren’t using it as well as we could (i.e., I’d like to enable conda
install bce=fall-2015 for at least arbitrary-ish unix). Such a thing is
not hard, we just haven’t gotten to it.

davlab [7:43 PM]
I think a great first step would be simply documenting the choices we
made
even without the rationale! I’ve been meaning to do this, and I think we
should explain how to install most of the major tools - like PyCharm,
RStudio, anaconda, etc.

—
Reply to this email directly or view it on GitHub
#65.

—
Reply to this email directly or view it on GitHub
#65 (comment).

[ryanlovett]

Hi all,

Yes, I'm available. And I apologize again for the delay. I created a test
build for fall on Saturday and need to clean up some of the FAILs.

If 10/15 is the target date, would you be against basing Fall on Ubuntu
15.10 which comes out on 10/22? Would it be okay for the 10/15 event to be
based on a Fall preview based on 15.10 beta?

Ryan

On Mon, Sep 21, 2015 at 9:08 AM, Chris Kennedy notifications@github.com
wrote:

That would be fun - I’m open 12:30-3pm on 10/2 and 10/9.

On Mon, Sep 21, 2015 at 7:58 AM, Christopher Paciorek <
notifications@github.com> wrote:

Sounds like a good idea to me. 10/2 and 10/9 both work though I teach
2-3:30.

On Mon, Sep 21, 2015 at 7:45 AM, Aaron Culich notifications@github.com
wrote:

After various discussions I've had with @davclark
https://github.com/davclark , @paciorek <https://github.com/paciorek

,

@ck37 https://github.com/ck37 , and @ryanlovett
https://github.com/ryanlovett , I wonder if it's time to have a BCE
HackDay ahead of doing a Fall release? Is there a day (or half day or
even
just 2 hours) that we could get together in the next couple of weeks?
Preferably before the Oct 15th Cloud WG topic which is: BCE on AWS?

Maybe Friday 10/2 or Friday 10/9?

I'll start triaging all the issues we have here
https://github.com/ucberkeley/bce/issues and in the documentation
issues
https://github.com/ucberkeley/bce/issues ahead of that HackDay.

I'd like to address issues like this one that Dav and I discussed on
the
uc-juptyer
slack https://uc-jupyter.slack.com/messages/general/:

aculich [6:49 PM]
@Davlab https://github.com/davlab Is anaconda the approach used in
http://bce.berkeley.edu ?

aculich [6:49 PM]
Whatever the approach, I wonder if we ought to include a short
description
in the https://github.com/ucberkeley/bce/blob/dev/README.md of the
high-level decisions and reasons for making them so that:

aculich [6:49 PM]

• folks can (without looking at the build scripts) see what our
  approach is; and

aculich [6:49 PM]

• can see our rationale (which won’t necessarily be obvious when
  looking at the build scripts) NEW MESSAGES

----- Yesterday September 20th, 2015 -----

davlab [7:42 PM]
@aculich https://github.com/aculich: yes, we use anaconda now. We
aren’t using it as well as we could (i.e., I’d like to enable conda
install bce=fall-2015 for at least arbitrary-ish unix). Such a thing is
not hard, we just haven’t gotten to it.

davlab [7:43 PM]
I think a great first step would be simply documenting the choices we
made
even without the rationale! I’ve been meaning to do this, and I think
we
should explain how to install most of the major tools - like PyCharm,
RStudio, anaconda, etc.

—
Reply to this email directly or view it on GitHub
#65.

—
Reply to this email directly or view it on GitHub
#65 (comment).

—
Reply to this email directly or view it on GitHub
#65 (comment).

[aculich]

@ryanlovett making it a preview release based on Wily Warewolf (15.10) would be good-- the working group can help with some of preview-release testing for the final fall release!

[paciorek]

I'm wary of basing on a release of Ubuntu that is not yet out. Is there a
compelling reason to wait as opposed to just using 15.04?

On Mon, Sep 21, 2015 at 10:06 AM, Aaron Culich notifications@github.com
wrote:

@ryanlovett https://github.com/ryanlovett making it a preview release
based on Wily Warewolf (15.10) would be good-- the working group can help
with some of preview-release testing for the final fall release!

—
Reply to this email directly or view it on GitHub
#65 (comment).

[davclark]

I'm mostly interested in getting BCE onto an LTS release, and keeping it that way. I've run into annoying issues surrounding things like CUDA.