Re: Dataflow final (?) hackathon for release - poll

3 views
Skip to first unread message

Graham Klyne

unread,
Nov 6, 2012, 6:13:53 PM11/6/12
to Katherine Fletcher, Bhavana Nanjangud Ananda, Anusha Ranganathan, Richard Jones, dataflo...@googlegroups.com
On 06/11/2012 13:31, Katherine Fletcher wrote:
> Hackathon goals:
>
> Day 1: Integration
>
> Day 2: Preparing for next steps: Packaging, benchamarking, get continuous
> testing/version management/workflow management set up. Update documentation
> if time.

My take is that, if 2 days are available, the first should focus on integrated
testing and tidying any loose ends. Hopefully, most of the substantive work on
the TODO list (see below) will be done beforehand. That would include
packaging, which I anticipate is substantially ready to go, but may need a few
configuration tweaks.

If that goes well, the second day could start looking at plans for allowing
development to proceed without us running into the kind of hiatus we've been
experiencing with this release. We need a framework in which we can make
incremental changes and roll them out with a good degree of confidence we're not
introducing regressions. I see re-establishing the continuous integration
environment as a key to achieving this, but this needs to be considered an
integral part of ongoing development. Developer and maintenance documentation
would also be part of this.

So, my suggestion for day 2 targets would be:
* identify priorities for next stages of development
* plan, and if possible deploy, a continuous integration environment. Be clear
how developers can add to the range of tests covered by CI
* discuss use of version management to better support development and release
requirements (I'd suggest gitflow, in full or part) (this would also address the
"forking" aspect of item 3 on the TODO list)
* check state of developer and maintenance documentation

#g
--

> To-do list for release:
>
>
>
> 1) SWORD multiple request problem fixed, and rolled into main DataBank
> code.
>
> 2) SWORD package configuration sorted and rolled into main DB code.
>
> 3) Working DataStage submission to single repository (no silo
> selection), rolled into main DataStage code and fork set up for work on
> multiple repositories.
>
> 4) Use DataStage to make updates to a data package already submitted to
> DataBank.
>
> 5) Install docs updated as needed (note about MySQL password
> selection?).
>
> 6) Install packages (debian at least) tested on Ubuntu 12.04 server
>
> 7) Testing
>
> 8) Quick check that DataStage
> <https://github.com/dataflow/DataStage/wiki> and DataBank
> <https://github.com/dataflow/RDFDatabank/wiki> documentation wikis still
> broadly relevant to newest version of code. Anything you can tick off the
> list in Jira <http://dataflow-jira.bodleian.ox.ac.uk/jira/browse/DATAFLOW> ?
> You should be able to pull down a filter for all outstanding issues, as in
> the picture below.
>
>
>
> Keep up the good work!
>
>
>
>
>
> Katherine
>
>
>
>
>
>
>
>
>
>
>
>
>
>

Katherine Fletcher

unread,
Nov 6, 2012, 8:31:59 AM11/6/12
to Bhavana Nanjangud Ananda, Anusha Ranganathan, Richard Jones, Graham Klyne, dataflo...@googlegroups.com

Dear Bhavana, Anusha, Richard and Graham (also copied to list, just in case):

 

From what I’ve heard, you’ve all been making progress since the last hackathon and there is a realistic chance that you’ll be able to knock this on the head in the next few weeks.

 

Can I propose one last hackathon, for you to test all of your updates and maybe even finish up the release?  I think we mainly need Bhavana, Anusha and Graham, but if Richard could join in (even for part) that might be handy.  If Richard can’t come, please make sure you’re in touch some other way, in case of SWORD updates.  Here’s a poll (typed with trembling hands…).

 

http://www.doodle.com/nd2pr9wucrtappvu

 

Hackathon goals:

Day 1: Integration

Day 2: Preparing for next steps: Packaging, benchamarking, get continuous testing/version management/workflow management set up.  Update documentation if time.

 

To-do list for release:

 

1)      SWORD multiple request problem fixed, and rolled into main DataBank code.

2)      SWORD package configuration sorted and rolled into main DB code.

3)      Working DataStage submission to single repository (no silo selection), rolled into main DataStage code and fork set up for work on multiple repositories.

4)      Use DataStage to make updates to a data package already submitted to DataBank.

5)      Install docs updated as needed (note about MySQL password selection?).

6)      Install packages (debian at least) tested on Ubuntu 12.04 server

7)      Testing

8)      Quick check that DataStage and DataBank documentation wikis still broadly relevant to newest version of code.  Anything you can tick off the list in Jira?  You should be able to pull down a filter for all outstanding issues, as in the picture below.

 

Keep up the good work!

 

 

Katherine

 

 

 

 

 

image003.jpg
Reply all
Reply to author
Forward
0 new messages