What would you like to see from CollectiveAccess 2.0?

We'd like to hear from users as we plan for CollectiveAccess 2.0. What would you like to see from the next major release? 

All ideas are welcome from features to new protocols to big picture wish lists.

Thanks in advance for weighing in!


  • I would like to have plugins to store images somewhere other than the file system. I can't run it on a cloud service, since I can't upload to Amazon S3 or a similar service.
  • edited July 2017
    Some off line functions to allow adding or updating items on a local machine without internet access and then syncing when internet access is available?
    Because of limited/nonexisting internet access I have to make notes on paper and enter later. This would save lots of time.

  • I have to deal with a shared server.  There are a number of limitations that I have found partial workaround,but it would be nice to have the functionality built in.

    - It is very easy to run up against too many mySQL requests per user, and timeouts.  It would be wonderful to be able to break up long operations to avoid this, such as reindexing, etc

    -It would be really useful to be able to run all caUtils function from a browser - shared server often do not have shell access. I have always wondered why they cant be.

    - Built-in backup/restore functions runnable from a browser.

  • I would like it to be easier to run under https. I have to disable a lot of warnings due to scripts and images being loaded as http. I eventually gave up and went back to http in order to keep going.
  • That's something we can get into 1.7.4 or 1.7.5... it has been on the todo list forever.
  • Support for the following:

    php as a service  - certain batch processing would timeout when running php as a service
    php 7
    maraidb 10.x  listed as supported
  • Hi cisaksen ,

    When you say PHP as a service, so you mean under Windows? Or PHP-fpm?

    We already support PHP 7 and 7.1, so I'm not sure what else there is to do there.

    Your suggestions that we explicitly support nginx and MariaDB 10.x are good and will be done sooner rather than later.

    • the ability to do regex string replacements in batch editing.
    • the ability to search on the CA record cataloger, and the date created/modified of the CA record
    • command line or API media import (which I understand is in the to-do list)
    • an entity and list item merge feature

  • What do the entity and list item merge features entail?
  • I like the way ArchivesSpace does it.   If you have John A. Smith and John Andrew Smith (who are the same person) to merge,  you go to the edit screen of the authoritative one, select a merge button which dropdowns a search for the non-authoritative name, you select it and then click merge.  Then all the records previously associated with the non-authoritative name now  have it replaced with the authoritative one.   You could also have a checkbox or something to create the non-authoritative name as an alternate name. 
  • That's already possible in the current version of CA.  The only thing we don't do, which perhaps we should, is the automated creation of the non-authoritative name. (At least in our experience you're usually consolidating typos, so you don't necessarily want to keep the you're removing as non-authoritative).

    Try deleting an entity. You should see the merge controls as part of the delete confirmation screen.

  • Very nice.  I just didn't think to look for it there and apparently haven't read the wiki close enough.
  • Ability to attach more than one media item per object from the media importer. (So we often have the main image, a detail, maybe the back, etc. It'd be cool if we could set up the media "type" to correspond somehow so it could do it. So item 2017.1.1.back it would know would be the back of 2017.1.1 and 2017.1.1.detail and so forth. It's a drag to have to upload those all individually.)
  • The big one for me would be the ability to run "big" jobs like imports or exports in the background on the server itself, but using the web interface to set it up.

    Right now when you go to run an import from the web interface the import runs right away, and the webpage keeps loading until it is done (or times out). The only way to do "big" jobs, especially on a slow network, is to do it directly on the server in the command line or through ssh, which is really not the best, especially if you want users to be able to do this but not have access to server settings.

    Ideally, there would be a way to set up a job, test it (i.e. import the first 100 items from the list and see if it works), and then either tell the server to go away and do it now, or schedule it to run later during a period of low demand. The server would then email you when the job is finished, or if there is an error, etc.


    Chris Litfin
  • Chris: you can do this in the current version of CA by enabling background processing. To do this you'll need to __CA_QUEUE_ENABLED__ in your setup.php file and then set up a cron task to periodically run caUtils process-task-queue
  • How about a plug in to create thank you letters, deed of gifts, and loan agreements? I am envisioning being able to upload the latest letter head and filling out a few forms in Providence with the relevant information specific to that museum. When a donation arrives, I could then add the details of the specific donation and the program generates a deed of gift and a customized thank you letter. Yes, I probably could create a similar experience using displays, but having a more focused experience would make a nice quality of life improvement. 
  • I would like to see updated documentation or Pawtucket2. The previous version docs seem totally useless for version2. I would be willing to help if you like.
  • BCHSDA: I totally agree. We'd love your help. Contact me at seth@collectiveaccess.org to discuss when you can.
  • I went to a set to see if I can print out labels for a set of storage locations, and it would not give me that option; therefore, I would like to see if you can add an option to export labels from a user set.
  • +1 for sjenson's idea!
  • Re: My comment on 2 August about background processing

    Hi Seth,

    I hadn't heard of this feature before and I have been playing around with it a bit. __CA_QUEUE_ENABLED__ only seems to handle single imports of media.

    What I want is essentially something that allows you to set up a cron job to run anything that catools/cautils can run, but from the browser instead of the command line. (I know that actually building cron jobs from the web is not a good idea, but something that can duplicate that functionality.) The ability to "preview" the first x% of the job (i.e. the first 100 lines to import) would be a real plus.

    Actually, bringing all the catools/cautils functionality out to the web browser would be nice. I know some of it (like rebuilding search indices) is available in the web browser, but (correct me if I am wrong) not all functionality is.


    Chris Litfin
  • I'll just come with a first suggestion : opening an external system for plugin or theme distribution (kind of marketplace just like wordpress or drupal plugins, but without the money question for me, only GPL and github/public repos), and allow users to share notes & evaluation those. CollectiveAccess is a marvelous platform, I've two or three plugins that I'd like to share for Providence : one using phantomjs to automatically create Catalogue Raisonné as a pdf, think the one of Noguchi, builded from the Providence interface ; one to edit simple lists (one language, using same term for singular/plural/idno/value with jstree) ; one tourML export for TAP mobile app ; another for jewish/arabian/masonic/french revolutionary calendars...

    I've even made tests to connect an autonomous raspberry hygro/thermo box, reporting to CA storage locations :-)

    Those are try-outs, mostly oneshot project, they are not ready to be properly diffused inside CA zip, and it's hard to think of them inside some branch.
    Maintaining and expending those little repositories outside could be a good way to do thing, expanding the community and sharing more of the profile.

    It could be the same for the current Providence Statistics plugin (things I'd like to change : proper feet to CA look'n feel, defaulting back to Google Charts, exports to XLSX and PDF...).

    For Pawtucket 2, I didn't take much time on the forum those last times, so I'm not sure to be a very reliable documentation asset, but the content management aspects are really interesting to dig in.

    An international version of the website too ? I'm still willing to translate to french the exact content of it (aka without any credit or link for me), it would be great for CA to be visible in other languages, spanish and german would be great additions too.

    As always, what Whirl-i-gig and this community has build is marvelous :-) Best to all of you.
  • I'd like to see the displaynameFormat parameter added to the entityJoiner refinery so the display name format can be customized in the same way as it can be for entitySplitter.
    Additionally, for entities, I've always found that using a single field called 'surname/organization' to hold surname OR organization to be a bit restrictive as I typically want to hold the name of an actual person at the organization in the firstname/middlename/surname fields and then the name of the organization itself in a separate field, possibly a new one, e.g. 'orgname', or maybe in the 'displayname' field. By using one field for surname/organization this complicates things so what I want to do is not easily possible. I also don't want to see a field called 'surname/organization' on my user interface - can this not be configurable, perhaps via the profile?
  • I would like CA searches to use 'like'
  • Hi,

    My institution does not use Pawtucket, but rather takes advantage of CA's synchronization service. Will you continue to support and update that service? I noticed on the development page that this service is run on a web-service API that is now deprecated. 
  • Caroline: actually the synchronization service does not use a deprecated API. We have no plans to stop supporting it.
  • Hi Seth,

    Thanks, this is great news for us!
  • edited September 2017
    Defaults that can be set specifically for users or groups of users.
    Ability to add the entire search/browse results to an existing set rather than just the toggle checked (this gets tedious when you have many pages of results.)
    Ability to name and save media import settings and choose from saved options in the gui.
  • It would be great to allow users to access directories on their local machines when importing media via the web gui (either the IMPORT -> Media or mapped IMPORT -> Data).

    Right now, we're struggling with how we allow users access to the /import directory on the linux server, which seems to be where the images to import must go.
Sign In or Register to comment.