v1.4 Planning

[Greg Pendlebury]

Howdy all,

I can see the light at the end of the tunnel... I get to go back to coding soon, but I have just one or two documents left. This one is focusing on the jobs that are on the list for inclusion (or potential inclusion) in v1.4, scheduled for release after the Xmas / New Year break.

Typically our agile project approach means that jobs get laid out in priority order, and whichever ones we finish in time make it into the release. We front-load this order with the most important elements of the release, so...

• v1.2 was about Curation Management.
  
• v1.3 was about Fedora 3 Upgrading.
• v1.4 is more about back-end housekeeping.

Because of this approach, a few low priority jobs have been continually shuffled further down the list. Going into v1.4, since we don't have a single high profile goal, we hope to push these to the front of the queue so they don't keep getting pushed around. The list below represents a rough outline of jobs on the table right now... and as per above, not all of them will make it into the list, so please take this as an oppurtunity to provide feedback if priorities should be shifted, as well as provide feedback and/or just express an interest in any tickets here.

1. RIF-CS Updates... including v1.3 compatibility, and expanding XML ingests to accept RIF-CS.
2. Centralising storage and configuration of persistent identifiers. Currently there are references scattered through-out the code.
3. Add SRU Implementation to Mint... some feedback may be useful here. Currently the v2.0 SRU spec is only in draft, but it is desirable, since it adds support for JSON. This would be good for us, since we have extensive use of JSON throughout the system.
4. Overhaul of build processes. I'd like to implement institutional builds in Maven for implementers, capturing how you like to do curation along with local customisations; this should make everyone's life easier during deployment and upgrades. Sample build would be available for common setups (eg. Handle, VITAL and such).
   
5. Also move the ZIPs to be basically a demo system using 'local' curation. Some feedback during the workshop from David Beitey @ JCU has also highlighted some great improvements we can make to deploying the ZIPs in a more streamlined fashion.
6. Integrate with NLA on Party infrastructure. Dialog is currently ongoing with NLA and ANDS, but at the very minimum we intend to proxy NLA searches to allow cross-institutional links for Parties that already exist Nationally.
7. Investigate OpenLayers for geospatial map work in the forms. I'm informed there's already some experience out there in the eResearch community on this, so I may start picking others brains, or even (optimistically) invite some contributions.
8. A couple of minor bugs I've got on my list to look into regarding Geonames search errors.
9. If time permits, I've got some Fascinator re-engineering I'd like to get done. Ideas that got whiteboarded by the team at USQ, but we never got around to. If I get this far, I'll post more notes on the Fascinator list.

As always, there may be a variety of more trivial five minute fixes that come up, and make their way into the build, but please let me know if you have any feedback.

Ta,
Greg

[Greg Pendlebury]

>> RIF-CS Updates... including v1.3 compatibility, and expanding XML ingests to accept RIF-CS.

That would the newly released RIF-CS v1.3 by the way, nothing to do with ReDBox 1.3

[Grant Jackson]

Hi Greg, Duncan,

Firstly I wanted to say thanks for all your excellent and timely assistance
over the past months with getting ReDBox going at Flinders. We just moved
ReDBox-Mint into production on Friday and I'm very pleased with its
functionality. (We still have a few loose ends to tidy up so it is
currently empty.) I think its a great product and the 2 of you plus past
and present team members from USQ, UoN, QCIF and ANDS have done a great job.

Secondly I have attached my 2 cents worth of suggestions for future versions.

I will still be at Flinders but will mostly be involved with other work for
the next few months so am less likely to be active on this list. Thanks
again.

Cheers, Grant

Quoting Greg Pendlebury <greg.pe...@gmail.com>:

> >> RIF-CS Updates... including v1.3 compatibility, and expanding XML
> ingests to accept RIF-CS.
>
> That would the newly released RIF-CS v1.3 by the way, nothing to do with
> ReDBox 1.3
>
> On 25 November 2011 11:14, Greg Pendlebury <greg.pe...@gmail.com>wrote:
>
> > Howdy all,
> >
> > I can see the light at the end of the tunnel... I get to go back to coding
> > soon, but I have just one or two documents left. This one is focusing on
> > the jobs that are on the list for inclusion (or potential inclusion) in
> > v1.4, scheduled for release after the Xmas / New Year break.
> >
> > Typically our agile project approach means that jobs get laid out in
> > priority order, and whichever ones we finish in time make it into the
> > release. We front-load this order with the most important elements of the
> > release, so...
> >

> > - v1.2 was about Curation Management.
> > - v1.3 was about Fedora 3 Upgrading.
> > - v1.4 is more about back-end housekeeping.

> >
> > Because of this approach, a few low priority jobs have been continually
> > shuffled further down the list. Going into v1.4, since we don't have a
> > single high profile goal, we hope to push these to the front of the queue
> > so they don't keep getting pushed around. The list below represents a
> rough
> > outline of jobs on the table right now... and as per above, not all of
> them
> > will make it into the list, so please take this as an oppurtunity to
> > provide feedback if priorities should be shifted, as well as provide
> > feedback and/or just express an interest in any tickets here.
> >

> > 1. RIF-CS Updates... including v1.3 compatibility, and expanding XML
> > ingests to accept RIF-CS.
> > 2. Centralising storage and configuration of persistent identifiers.

> > Currently there are references scattered through-out the code.

> > 3. Add SRU Implementation to Mint... some feedback may be useful here.

> > Currently the v2.0 SRU spec is only in draft, but it is desirable, since
> it
> > adds support for JSON. This would be good for us, since we have
> extensive
> > use of JSON throughout the system.

> > 4. Overhaul of build processes. I'd like to implement institutional

> > builds in Maven for implementers, capturing how you like to do curation
> > along with local customisations; this should make everyone's life
> easier
> > during deployment and upgrades. Sample build would be available for
> common
> > setups (eg. Handle, VITAL and such).

> > 5. Also move the ZIPs to be basically a demo system using 'local'

> > curation. Some feedback during the workshop from David Beitey @ JCU has
> > also highlighted some great improvements we can make to deploying the
> ZIPs
> > in a more streamlined fashion.

> > 6. Integrate with NLA on Party infrastructure. Dialog is currently
> > ongoing with NLA and ANDS, but at the very *minimum* we intend to

> > proxy NLA searches to allow cross-institutional links for Parties that
> > already exist Nationally.

> > 7. Investigate OpenLayers for geospatial map work in the forms. I'm

> > informed there's already some experience out there in the eResearch
> > community on this, so I may start picking others brains, or even
> > (optimistically) invite some contributions.

> > 8. A couple of minor bugs I've got on my list to look into regarding
> > Geonames search errors.
> > 9. If time permits, I've got some Fascinator re-engineering I'd like

[Greg Pendlebury]

Thanks for that Grant.

It looks like most of these are customisations that will become easier with the new institutional builds, along with a bunch of polish we can add to the system. I'll have a look in more detail later and open tickets for any new works.

Ta,
Greg

[Duncan Dickinson]

Hi Grant,

Thank you for your support and efforts it has really helped us ensure that ReDBox can meet institutional needs.

All the best for your new endeavours!

Cheers,

Duncan

[Greg Pendlebury]

Hey Grant,

Thanks again, and he's the responses I've taken:

No work required / Already planned for v1.4:

• "Support for RIF-CS v1.3"
• "Document how to delete references: ..."
• "For each release-type ... provide sample config files. ..."
• "Support for ARDC Party Infrastructure Project harvest"

Institutional build samples:
These sorts of front-end customisations are exactly what I'd suggest for institutional builds. I will provide samples.

• "Allow handle corresponding to the ReDBox-Mint record to be visible without login as Admin."
• "Allow links to related objects in ReDBox-Mint with handles to be visible without login as Admin."
• "Allow some dataset Management info to be visible without login."

Doco Tickets
Not sure when these will be addressed, but I've ticketed them:

• #51 "Document how to configure non-ARC/NHMRC project ingestion..."
• #52 "Provide user documentation... (various)"
  

Bugs / Improvements
Higher priority tickets

• #53 "ReDBox will not let you attach a file to a dataset if it is already attached to another dataset."
• #54 "Consider providing replacement template-fields in various config files for replacing UoN names with those of your own institution."
• #55 "ReDBox will not allow the user to delete more than one attachment file unless you "Save and Close" the dataset"
• #56 "If you are logged into ReDBox; you click on Published ... that sometimes you see all datasets." << I've seen this too
  
• #57 "If records are deleted in the Mint by Admin, I believe they are still visible at Home > Latest Additions" << This could just be a Solr cache
  

Unknown
This has always been talked about, but I have no idea how/when this will arise, or what it will look like.

• "Support for RIF-CS service metadata records"

Ta,
Greg

[Grant Jackson]

Hi Greg,

Thanks for the update. These responses look very good.

Cheers, Grant

Quoting Greg Pendlebury <greg.pe...@gmail.com>:

> Hey Grant,
>
> Thanks again, and he's the responses I've taken:
>

> *No work required / Already planned for v1.4:*
>
> - "Support for RIF-CS v1.3"
> - "Document how to delete references: ..."
> - "For each release-type ... provide sample config files. ..."
> - "Support for ARDC Party Infrastructure Project harvest"
>
> *Institutional build samples:*

> These sorts of front-end customisations are exactly what I'd suggest for
> institutional builds. I will provide samples.
>

> - "Allow handle corresponding to the ReDBox-Mint record to be visible
> without login as Admin."
> - "Allow links to related objects in ReDBox-Mint with handles to be

> visible without login as Admin."

> - "Allow some dataset Management info to be visible without login."
>
> *Doco Tickets*

> Not sure when these will be addressed, but I've ticketed them:
>

> - #51 <http://code.google.com/p/redbox-mint/issues/detail?id=51>"Document

> how to configure non-ARC/NHMRC project ingestion..."

> - #52 <http://code.google.com/p/redbox-mint/issues/detail?id=52>"Provide
> user documentation... (various)"
>
> *Bugs / Improvements*
> Higher priority tickets
>
> - #53 <http://code.google.com/p/redbox-mint/issues/detail?id=53> "ReDBox

> will not let you attach a file to a dataset if it is already attached to
> another dataset."

> - #54 <http://code.google.com/p/redbox-mint/issues/detail?id=54>"Consider

> providing replacement template-fields in various config files for
> replacing UoN names with those of your own institution."

> - #55 <http://code.google.com/p/redbox-mint/issues/detail?id=55> "ReDBox

> will not allow the user to delete more than one attachment file unless
> you
> "Save and Close" the dataset"

> - #56 <http://code.google.com/p/redbox-mint/issues/detail?id=56> "If you

> are logged into ReDBox; you click on Published ... that sometimes you see
> all datasets." << I've seen this too

> - #57 <http://code.google.com/p/redbox-mint/issues/detail?id=57> "If

> records are deleted in the Mint by Admin, I believe they are still
> visible
> at Home > Latest Additions" << This could just be a Solr cache
>

> *Unknown*

> This has always been talked about, but I have no idea how/when this will
> arise, or what it will look like.
>

> - "Support for RIF-CS service metadata records"

[Peter Sefton]

Given that NLA integration is a feature of the new metadata stores funding would it be better to put that on hold, research the full scope of what needs to be done and resource it using money contributed by RedBOX sites receiving that funding? Complicated to orchestrate I know, but it might be a way to stretch out the funding to get more features.

--

Peter Sefton +61410326955 p...@ptsefton.com http://ptsefton.com
Gmail, Twitter & Skype name: ptsefton

[Greg Pendlebury]

Hi Peter,

v1.4 won't be solving the bugger NLA integration issues, just a simple search proxy for the (admittedly limited) number of researchers already present in NLA systems.

There is ongoing dialogue with NLA and ANDS that I suspect will take some time before this can even start in earnest on more serious work. What version of ReDBox (if any) and what funding it occurs under is A) beyond my ability to predict, and B) beyond my pay grade to espouse :) But more funding and more developers is always good...

Ta,
Greg

[Greg Pendlebury]

>> "v1.4 won't be solving the bugger NLA integration issues"

Ummm, that would be "v1.4 won't be solving the bigger* NLA integration issues"

Almost as good as a "kid in the pants".

Ta,
Greg

[Vicki Picasso]

Hi Greg,

Feedback on future planning from Newcastle, in no particular order:

1. 'ALT tags' in interface forms and any other usability issues that can be addressed
2. LDAP
3. availability of HTML datastream - to provide additional option to facilitate researcher review
4. Inclusion of Embargo date in forms, mappings, etc - in tandem with Access Rights/Conditions (Management Tab)
5. Inclusion of Ethics application/information in forms (Management Tab)
6. Revisit availability of Australian Gazetteer? And improved assistance with geospatial - widgets perhaps
7. Modify behaviour of the widget associated with 'Date Coverage' (Coverage Tab)
8. Link validation checker (Admin)
9. RIF-CS 1.3 compliance
10. Issue with attached same file/same file name - system doesn't allow

We plan to do 1.3 UAT for our VITAL/Fedora upgrade tomorrow so may send some more feedback through based on outcomes.

cheers, Vicki

Vicki Picasso
Senior Librarian, Research Support and Bibliographic Services
University Library
Academic and Global Relations Division
University of Newcastle
Callaghan NSW 2308 Australia

P. 02 4921 5861
F. 02 4921 5833
E. Vicki....@newcastle.edu.au

>>> Greg Pendlebury <greg.pe...@gmail.com> 25/11/2011 12:14 pm >>>
Howdy all,

I can see the light at the end of the tunnel... I get to go back to coding
soon, but I have just one or two documents left. This one is focusing on
the jobs that are on the list for inclusion (or potential inclusion) in
v1.4, scheduled for release after the Xmas / New Year break.

Typically our agile project approach means that jobs get laid out in
priority order, and whichever ones we finish in time make it into the
release. We front-load this order with the most important elements of the
release, so...

- v1.2 was about Curation Management.
- v1.3 was about Fedora 3 Upgrading.
- v1.4 is more about back-end housekeeping.

Because of this approach, a few low priority jobs have been continually
shuffled further down the list. Going into v1.4, since we don't have a
single high profile goal, we hope to push these to the front of the queue
so they don't keep getting pushed around. The list below represents a rough
outline of jobs on the table right now... and as per above, not all of them
will make it into the list, so please take this as an oppurtunity to
provide feedback if priorities should be shifted, as well as provide
feedback and/or just express an interest in any tickets here.

1. RIF-CS Updates... including v1.3 compatibility, and expanding XML
ingests to accept RIF-CS.
2. Centralising storage and configuration of persistent identifiers.

Currently there are references scattered through-out the code.

3. Add SRU Implementation to Mint... some feedback may be useful here.

Currently the v2.0 SRU spec is only in draft, but it is desirable, since it
adds support for JSON. This would be good for us, since we have extensive
use of JSON throughout the system.

4. Overhaul of build processes. I'd like to implement institutional

builds in Maven for implementers, capturing how you like to do curation
along with local customisations; this should make everyone's life easier
during deployment and upgrades. Sample build would be available for common
setups (eg. Handle, VITAL and such).

5. Also move the ZIPs to be basically a demo system using 'local'

curation. Some feedback during the workshop from David Beitey @ JCU has
also highlighted some great improvements we can make to deploying the ZIPs
in a more streamlined fashion.

6. Integrate with NLA on Party infrastructure. Dialog is currently
ongoing with NLA and ANDS, but at the very *minimum* we intend to proxy

NLA searches to allow cross-institutional links for Parties that already
exist Nationally.

7. Investigate OpenLayers for geospatial map work in the forms. I'm

informed there's already some experience out there in the eResearch
community on this, so I may start picking others brains, or even
(optimistically) invite some contributions.

8. A couple of minor bugs I've got on my list to look into regarding
Geonames search errors.
9. If time permits, I've got some Fascinator re-engineering I'd like to

[Greg Pendlebury]

Hi All,

Just a quick update on where v1.4 is up to. Last week was a return to work for QCIF in 2012 and I've spent the time tidying up what tickets made the cut into v1.4. Summary below, listed in the same order as original planning.

1. (Done) RIF-CS v1.3: There is a reference sample for basic v1.3 compliance in place for users to refer to. It can be swapped in place with the v1.2 template if desired. More v1.3 work will come later (see the other threads on this list), moving from more then just 'compliance' in the next version of ReDBox (1.5).
   
2. (Done) Central config is referred to more religiously through-out the system(s). I also addressed Grant's feedback (#54) regarding institutional labelling whilst doing this, as it is related.
3. (Skipped) We received no feedback on this, and no interest was expressed. Given the scope of the work involved, we are starting to rethink whether there is any urgency to this idea at all. Please let us know if you want to see this pushed forward, since at the moment, it looks like this is going to go back onto a back-burner... at least until SRU v2.0 is moves from a draft spec.
4. (Done... or will be) Institutional builds... this is what I'm working on right now. I'd hoped to wrap it up on Friday and release but hit a snag. Expect it today or tomorrow.
5. (Will be done... as above) This is going to be the last part of my institutional build work.
6. (Missed) This work didn't make it into v1.4, but just last week we verified with ANDS and NLA exactly what the complete 'integration' will look like. I'll probably post more detail on this in its own thread and it will be the first work implemented in v1.5 (my guess, not based on planning notes) if institutions are looking to bolt on to existing installations. As a sidenote however, I've already posted some experimental work that allows integration with National infrastructure via RDA instead of NLA.
7. and 8. and 9. (Missed) We just didn't get this far down the list. Although of course, minor bugs (something point 8 was about) are always addressed if found.

As per point 4, expect a proper release either today or tomorrow.

Ta,
Greg

On 25 November 2011 11:14, Greg Pendlebury <greg.pe...@gmail.com> wrote: