{23} Trac comments (3729 matches)

Results (3101 - 3200 of 3729)

Ticket Posixtime Author Newvalue
#665 1294247602000000 thejimmyg Duplicate of #884
#665 1297268097000000 thejimmyg The latest version of the DMS collection interface spec says that this should be a manual process. No work to be done.
#667 1294414834000000 thejimmyg This is something we'll shortly be looking into for DGU work anyway once we switch to the "thin" system so don't need a separate ticket for it here too. Also, the upgrade of SQLAlchmey and test performance improvements may also make this less of a problem now anyway. Marking wontfix.
#668 1294412060000000 thejimmyg Assigning this to myself but won't be looking at it any time soon, if anyone else wants to take it on I'd be very grateful.
#668 1311176649000000 thejimmyg This ticket is more than 6 months old so marking as invalid in line with our ticketing policy. The API v1 and 2 were completely re-written as part of the logic layer refactor so it is likely this issue is resolved anyway.
#676 1294413641000000 thejimmyg There was discussion of this on the list a few months ago, was there a conclusion?
#691 1294412765000000 thejimmyg This is more a long term idea, taking ownership for the timebeing but happy for someone else to take this on.
#691 1294412929000000 thejimmyg If we did implement this we'd need to link new sample package to previous sample package of continuous series. So closing #748.
#691 1294413010000000 thejimmyg We'd also fold up continuous series in search results behind newest sample package. So closing #749.
#691 1294413993000000 thejimmyg We'd also discuss package relationships ideas with JF so closing #444.
#692 1294247841000000 thejimmyg CKAN doesn't need to implement this, Drupal does. Incidentally, the initial version is implemented on UAT anyway.
#698 1291832133000000 thejimmyg Actually we've implemented a first version which doesn't store the data. See this post: http://blog.ckan.org/2010/12/04/open-data-day-announcing-ckan-data-proxy/ You can get data like this: http://1.latest.jsonpdataproxy.appspot.com/?sheet=1&indent=4&url=http://research.dwp.gov.uk/asd/asd4/r1_values.xls
#699 1294412194000000 thejimmyg I've just verified that the search pages navigation does not respect the checkboxes. Could you have a look please Friedrich?
#711 1294232521000000 thejimmyg At the moment DGU has locations such as Wales, England etc. This won't change and won't try to be merged with INSPIRE datasets or bounding boxes. What we do need is a flag for INSPIRE so that the different types of package can be highlighted.
#711 1296592589000000 thejimmyg Our new approach agreed with Rod is to start with a minimum set of package extras and expand them in line with feedback from the user group. We can now show the source GEMINI record as XML (and also a very basic HTML view) for anyone who wants the full detail. No need for this ticket anymore.
#712 1294408364000000 thejimmyg There is no 0.4 metadata spec yet so we can't model it. If there was, the discussion would be on #711 anyway.
#728 1294232752000000 thejimmyg WAF records may always need to be re-harvested to see if they have changed. Does CSW provide any functionality that allows us to see what has changed?
#728 1304936482000000 thejimmyg Actually, these are a different type of filter. We need to send a filter to only get modified documents, to save processing our end.
#728 1310124784000000 thejimmyg Isn't this already completed? Putting in the backlog but I think it is already implemented. Adria, could you please update accordingly?
#736 1294409747000000 thejimmyg The tests we write for this would need to: * Submit harvest source form with valid CSW location * Submit harvest source form with invalid CSW location We can now close #790 and #791. Note that discussion on end-to-end testing is now at #784.
#736 1296592767000000 thejimmyg We also need to check WMS URLs added to package resources to check they are genuine WMSs.
#736 1304937755000000 thejimmyg We now check WMS links. The CSW link checking could be more robust. New proposal is: Check that a URL with a / is treated the same way as one without. Treated as the same if same port, scheme, domain and path (ie query string and fragment are ignored, port is normalised away)
#738 1294408864000000 thejimmyg This needs more discussion. It is a separate problem from deleting a harvest source from the admin interface which is dealt with in #884.
#738 1294409432000000 thejimmyg Note from John Bywater on duplicate ticket #783 says: "We need a condition for 'disappears'. Asked AW about this (prob. in July) but should chase for an answer."
#738 1297685069000000 thejimmyg It may be handy to have a history of harvested documents and also be able to delete documents but still get their revisions. The package comparison code will need to look at these revisions.
#738 1299761436000000 thejimmyg This is now complete and on UAT.
#739 1294247758000000 thejimmyg This needs to be implemented via an API call that Drupal can use
#739 1296592842000000 thejimmyg Implemented today. May want a better implementation later.
#740 1294408931000000 thejimmyg This will have to be done via the API though for DGU/UKLP. Should probably have it on the package interface in main CKAN too though.
#740 1296592889000000 thejimmyg Duplicate of #739.
#743 1294412807000000 thejimmyg This is a Drupal requirement rather than a CKAN one for the timebeing.
#744 1294412813000000 thejimmyg This is a Drupal requirement rather than a CKAN one for the timebeing.
#745 1294412820000000 thejimmyg This is a Drupal requirement rather than a CKAN one for the timebeing.
#746 1294412827000000 thejimmyg This is a Drupal requirement rather than a CKAN one for the timebeing.
#747 1294412834000000 thejimmyg This is a Drupal requirement rather than a CKAN one for the timebeing.
#748 1294412916000000 thejimmyg Merging with #676.
#748 1294412976000000 thejimmyg Sorry, that should have been #691.
#749 1294412986000000 thejimmyg Merge with #691
#750 1294408970000000 thejimmyg How would this be done? Is it part of the CSW spec?
#750 1296592940000000 thejimmyg Duplicate of #728
#752 1294408472000000 thejimmyg Duplicate of #712. See #711 now.
#753 1294247922000000 thejimmyg Not an explicit requirement yet so closing ticket. Will re-open if needed.
#754 1294408726000000 thejimmyg We may not need to do any filtering in practice but if we did, the implementation would have these stages: * Add filter attribute to harvest source entity * Add filter field to harvest source form * Construct and send filtered CSW GetRecords request Thus these tickets can now be marked as duplicated: #755 #756 #759.
#754 1294409145000000 thejimmyg Also: * Change API documentation to indicate harvest source entity has filter attribute * Change "CSW Get Records" request to accept and use given CSW filter Now #758 and #760 can be closed too.
#755 1294408632000000 thejimmyg Duplicate of #754
#756 1294408642000000 thejimmyg Duplicate of #754
#757 1294233016000000 thejimmyg Don't fully understand this ticket. Will return to it.
#757 1296593448000000 thejimmyg Duplicate of #754. If we added filtering we'd need to write a migration script anyway.
#758 1294409053000000 thejimmyg Also a duplicate of #754.
#759 1294408652000000 thejimmyg Duplicate of #754
#760 1294409111000000 thejimmyg Duplicate of #754
#763 1292775248000000 thejimmyg Notes from my discussion with David a while ago: Following irc discussion, it looks like read-only mode is simply achieved by Apache config: * 503 for PUT/POST operations - stops writes * 503 for GETs to URIs containing: /edit, /create, /new, /authz - stops providing forms that lead to a write * setenv CKAN_READONLY="Undergoing maintenance 12.00 UTC for one hour" which can be picked up by CKAN to be displayed as we see fit in the future. Friedrich's current IATI sprint may link into this and grey out edit links etc.
#766 1294248066000000 thejimmyg I haven't heard this mentioned yet, but yes, let's try to implement it if possible. Appears to expose a CSW interface? http://webhelp.esri.com/geoportal_extension/9.3.1/index.htm#ext_csw_clnts.htm
#766 1296593519000000 thejimmyg This is a duplicate of #767
#767 1294409336000000 thejimmyg Our CSW code should be able to cope with ArcGIS too though surely? Worth testing though. To do this we'd need to: * Obtain example ArcGIS CSW server * Write a test Now we can close #768 and #769.
#767 1296593504000000 thejimmyg I haven't heard this mentioned yet, but yes, let's try to implement it if possible. Appears to expose a CSW interface? http://webhelp.esri.com/geoportal_extension/9.3.1/index.htm#ext_csw_clnts.htm
#767 1304936413000000 thejimmyg We are now live and have used a variety of servers in testing.
#768 1294409308000000 thejimmyg Duplicate of #767
#769 1294409318000000 thejimmyg Duplicate of #767
#770 1294413390000000 thejimmyg Running CLI harvester command without arguments or with --help would return help for the harvester (currently crashes). We can therefore close #772, #773 and #774
#770 1296593925000000 thejimmyg I don't care about this. It is too trivial. If I get around to fixing it they great but I don't need it logged as a ticket.
#771 1294413256000000 thejimmyg This works for me. {{{ paster help }}} and {{{ paster --help }}} both show help text for me.
#772 1294413313000000 thejimmyg Merging with #770
#773 1294413400000000 thejimmyg Merging with #770
#774 1294413412000000 thejimmyg Merging with #770
#778 1294416567000000 thejimmyg This is now done.
#779 1294416581000000 thejimmyg This is now done.
#780 1294416608000000 thejimmyg This is now done.
#781 1294415081000000 thejimmyg This is a duplicate of #812 so closing.
#783 1294409399000000 thejimmyg Exact duplicate of #738.
#784 1294410089000000 thejimmyg In order to have end to end testing for UKLII we need to: * Review document from PP with the different things we need to test * Set up a GeoNetwork instance with sample documents to be tested We may also want to test ArcGIS, see #767
#784 1304936251000000 thejimmyg The UKLP test manager is looking at a new test proposal, we have had approval to go live without this, we can re-open later if necessary.
#786 1294410004000000 thejimmyg Duplicate of #884.
#787 1303118054000000 thejimmyg The AuthAPI now exists as an IMiddleware plugin, we really need the permission system moved into CKAN before it is useful though and this depends on a refactor of the Auth system. See #1094
#787 1315821118000000 thejimmyg The joint authentication was implemented a long time ago and is deployed on catalogue.data.gov.uk. We'll build the authorisation layer in ticket #1326 so marking this as fixed.
#788 1294410007000000 thejimmyg Duplicate of #884.
#789 1294248289000000 thejimmyg Will, this means that we just get given a URL to harvest from and we need to determine whether it is a CSW or a WAF. Can you look at this please? The code may already do it, I'm not sure.
#789 1296593257000000 thejimmyg Covered by #736 now.
#790 1294409709000000 thejimmyg Duplicate of #736.
#791 1294409723000000 thejimmyg Duplicate of #736
#793 1294409836000000 thejimmyg Anyone else know what this means? I can't find changeset aa9aa32e00a9 in the repository so marking as invalid.
#794 1294248216000000 thejimmyg This work is now underway in the underlying DGU implementation.
#794 1297686491000000 thejimmyg Actually, for the timebeing we will match but not do anything with that matched information, until there is a clear use case. Publisher is simply the publisher for which the source was registered. Closing this ticket.
#794 1300877517000000 thejimmyg I think we need to look at this again in relation to scotland.
#795 1296593361000000 thejimmyg This is a duplicate of #794 because if we needed to do this it would be because we need to support harvest sources registered on behalf of publishers by agents,
#799 1294232675000000 thejimmyg How can we tell a WAF document has changed, we simply need to re-harvest it to see surely? Moving the issue to ticket #728 to be dealt with together.
#800 1294245610000000 thejimmyg Discussion for this ticket is now at #728
#801 1294233386000000 thejimmyg Perhaps this should be implemented differently so that each harvest attempt creates its own database entry with a timestamp and a status attached. We'll need to move to this when we move to a queue based system.
#801 1296593735000000 thejimmyg I don't see why we'd really need to do this. Under the current system a job would be created just before the harvesting would be run so the creation date of the job is (give or take a few seconds) the date of the harvesting. If we moved to a queue system this would be done better anyway.
#801 1297686706000000 thejimmyg We do have a requirement for this now. The job model has changed so that it is hidden from the user. We therefore want to know the timestamp the job started and the timestamp it finished. We'll therefore need migrations adding too.
#801 1300196714000000 thejimmyg This will now be solved as part of the larger harvesting refactor. See #921.
#802 1294233294000000 thejimmyg Merging with #801
#804 1294233156000000 thejimmyg This is the document I drafted after discussions with each party which was approved by JF.
#809 1297075561000000 thejimmyg This is something to think about more in the future.
#810 1310128477000000 thejimmyg This ticket is over 6 months old so closing.
#811 1294415158000000 thejimmyg See also related extras field tickets #811 and #893
#813 1294411047000000 thejimmyg This is now done. The text reads "Add a Package" in the main navigation. Thanks!
#814 1294411109000000 thejimmyg This is now done. It's the "About" link in the main application. Thanks!
#815 1294410951000000 thejimmyg This is a duplicate of #234. Please have a look at existing tickets or re-open closed ones rather than adding new ones so that we can keep the old discussions. I've added your point to #234. Thanks!
#816 1311171466000000 thejimmyg John, could you have a look at this one please. We'd need an entry in the logic layer for it and then some autocomplete JavaScript on the format field. Can we standardise on jQueryUI going forward. http://jqueryui.com/demos/autocomplete/
Note: See TracReports for help on using and creating reports.