{23} Trac comments (3729 matches)

Results (1101 - 1200 of 3729)

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Ticket Posixtime Author Newvalue
#876 1293188088000000 anonymous Thanks for your feedback, very useful. I don't really agree with the people in the linked discussion who say it's pointless testing against a different database from production. The goal here is to make it easy enough for people to run as many tests as possible that they actually do so. Even 15 minutes is too long in that case. With sqlite we can get it in at under 5 minutes. I would also like to identify the longest running tests (which I would characterise as "functional" or "integration" tests and make them run as a separate suite, and then encourage a culture of writing true unit tests before functional tests, so that running unit tests can happen in 1 minute and be part of the regular development cycle. That's no replacement for also running *all* tests periodically, and also running tests under postgres, which we can continue to do on the continuous integration server. Longer term I agree that it would be better to run local tests against postgres too, but that will I think involve refactoring many of the tests.
#876 1293218714000000 anonymous I agree with all your points about testing apart form using sqlite, especially splitting out the functional tests and continuous integration. > Longer term I agree that it would be better to run local tests against postgres too, but that will I think involve refactoring many of the tests. Well there are two options 1. refactor the tests 2. refactor the code to use sqlite and postress It is a value judgment as to which is more complicated. I personally think 2 is more complicated but may be wrong on that. The real danger with 2 is that you are needlessly adding complication to production code, with 1 you are only changing the tests. Upgrading to sqlalchemy 0.5+ should happen first regardless. You will need upto date documentation. There is another option too. Put the postgres data directory on tempfs/ramfs and turn off durability [http://www.postgresql.org/docs/9.0/static/non-durability.html here]. We would need a way to db init before the tests where run (or) at boot). This may be the best of both worlds. Anyway Happy xmas!!
#885 1293278564000000 wwaites Friedrich had some mixed experiences with owslib and some German CSW endpoints: http://pudo.okfnpad.org/geodaten
#885 1293314219000000 wwaites See: https://bitbucket.org/ww/ckanext-ows impossible as of yet to test against live server from the OS
#884 1293319984000000 wwaites Implemented in feature-884-rmsource branch at https://bitbucket.org/ww/ckan/changeset/50cbfc5a6bee
#868 1293401808000000 anonymous Attached are the timings I have for the tests after I upgraded to 0.57 and after a few simple test tweaks. They do not include setup and teardown time at the class level as they are not assignable to individual tests.
#698 1293472613000000 anonymous Data proxy documentation: http://democracyfarm.org/dataproxy/api.html (included in sources) Updated ('s' as in structured) data proxy app: http://sdataproxy.appspot.com
#698 1293649815000000 rgrp This ticket is complete: * ckanext-dataapi: working /api/data/{resource-id} with tests * https://bitbucket.org/okfn/dataproxy - the dataproxy code running at http://jsonpdataproxy.appspot.com * functioning but needs tests and improvements There a whole bunch of improvements to be done but these will be in ticket:888
#889 1293715109000000 rgrp Done in cset:cdbb2e6128b3 (had added item to template earlier pulling from 'g' object but had not tied that up to config as yet). Also added test and also '''removed existing google analytics code''' -- so need to explicitly set this in config from now on.
#853 1294047550000000 pudo Looking at the changeset this cannot be functional yet: where is the implementation of the policy document exchange. It seems to me like this is currently adding the actual credentials to the request (self.ofs.conn.add_aws_auth_header(headers, 'PUT', path)). c.f.: * http://doc.s3.amazonaws.com/proposals/post.html#Form_Fields * http://code.google.com/apis/storage/docs/reference-methods.html#postobject
#510 1294138332000000 dread Completed 21st Dec 2010.
#679 1294166120000000 dread I added some extra bits in cset:1ca7ba29d409. Resource formats disagree between DGU and the FAQ - have sided with DGU for now as it's simpler. I think this ticket is complete now.
#563 1294231961000000 thejimmyg There is a choice here as to whether we provide an export to GeoNetwork or support a minimal CSW interface ourselves.
#566 1294232284000000 thejimmyg To a large extent it does now so closing this ticket in favour of more specific ones.
#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.
#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.
#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?
#757 1294233016000000 thejimmyg Don't fully understand this ticket. Will return to it.
#804 1294233156000000 thejimmyg This is the document I drafted after discussions with each party which was approved by JF.
#802 1294233294000000 thejimmyg Merging with #801
#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.
#800 1294245610000000 thejimmyg Discussion for this ticket is now at #728
#665 1294247602000000 thejimmyg Duplicate of #884
#884 1294247654000000 thejimmyg Related to ticket #665, reopening so that I can close it once I've got it working with Drupal too.
#739 1294247758000000 thejimmyg This needs to be implemented via an API call that Drupal can use
#692 1294247841000000 thejimmyg CKAN doesn't need to implement this, Drupal does. Incidentally, the initial version is implemented on UAT anyway.
#753 1294247922000000 thejimmyg Not an explicit requirement yet so closing ticket. Will re-open if needed.
#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
#794 1294248216000000 thejimmyg This work is now underway in the underlying DGU implementation.
#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.
#481 1294248359000000 thejimmyg This is now implemented for DGU.
#885 1294253497000000 wwaites done with feature-885-owslib
#884 1294253563000000 wwaites see feature-884-rmsource in the main ckan repo deleting my own repo
#296 1294407032000000 thejimmyg Merging into #896.
#297 1294407051000000 thejimmyg Merging into #896.
#298 1294407080000000 thejimmyg Merging into #896.
#299 1294407099000000 thejimmyg Merging into #896.
#394 1294407189000000 thejimmyg Duplicate of #371.
#398 1294407372000000 thejimmyg Considering this closed because of this blog post documenting the use of the API and linking to the CKAN docs: * http://data.gov.uk/blog/announcing-datagovuk-catalogue-apis
#413 1294407429000000 thejimmyg There is no caching in the current DGU CKAN code.
#413 1294407458000000 thejimmyg (or rather all caching is disabled). If we re-enable it we'll want to test its behaviour.
#427 1294407533000000 thejimmyg We may need an updated form of the license service to support UKLP.
#497 1294407705000000 thejimmyg Merged with #497
#497 1294407718000000 thejimmyg Sorry that should have been #496.
#496 1294407899000000 thejimmyg This would only apply to packages that had a HarvestedDcoument record (ie had been harvested). There are two approaches for this: * Build the capability to respond to CSW "GetCapabilities" and "Get Records" requests * Export to an existing CSW server to provide the mechanism for us
#570 1294407974000000 thejimmyg Merged with #569.
#569 1294408084000000 thejimmyg Our metadata validation will consist of: * Testing the schema * Testing the schematron * Checking we can get out the data we need
#617 1294408164000000 thejimmyg Duplicate of #569.
#612 1294408188000000 thejimmyg Duplicate of #569.
#611 1294408239000000 thejimmyg I don't understand this task, but suspect it can be treated as a duplicate of #569.
#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.
#752 1294408472000000 thejimmyg Duplicate of #712. See #711 now.
#755 1294408632000000 thejimmyg Duplicate of #754
#756 1294408642000000 thejimmyg Duplicate of #754
#759 1294408652000000 thejimmyg Duplicate of #754
#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.
#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.
#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.
#750 1294408970000000 thejimmyg How would this be done? Is it part of the CSW spec?
#758 1294409053000000 thejimmyg Also a duplicate of #754.
#760 1294409111000000 thejimmyg Duplicate of #754
#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.
#768 1294409308000000 thejimmyg Duplicate of #767
#769 1294409318000000 thejimmyg 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.
#783 1294409399000000 thejimmyg Exact duplicate of #738.
#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."
#790 1294409709000000 thejimmyg Duplicate of #736.
#791 1294409723000000 thejimmyg Duplicate of #736
#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.
#793 1294409836000000 thejimmyg Anyone else know what this means? I can't find changeset aa9aa32e00a9 in the repository so marking as invalid.
#786 1294410004000000 thejimmyg Duplicate of #884.
#788 1294410007000000 thejimmyg Duplicate of #884.
#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
#346 1294410298000000 thejimmyg Could you take a look at this at some point please David? If it is already resolved could you please close the ticket? Thanks!
#252 1294410341000000 thejimmyg David, do you know where this requirement has come from? Is is still relevant?
#381 1294410466000000 thejimmyg There is no mention of what the test defect is. All tests pass so marking invalid.
#881 1294410574000000 thejimmyg Hello fccoelho, Could you please post the output you get from running pip? Thanks, James
#447 1294410666000000 thejimmyg Hi Richard, could you take a look at this please? It is probably invalid now, in which case feel free to close it. Many thanks, James
#447 1294410738000000 memespring Have you got a link or screengrab?
#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!
#234 1294410993000000 thejimmyg Cygri opened #815 which I've closed as a duplicate. He requests: "The search field (on the homepage and in the top right corner of each page) should have autocomplete for package name. If a package name is selected, it should not do a search but go straight to the package page." @memespring - Is this something you are looking at?
#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!
#449 1294411301000000 thejimmyg Will, if this is something you are interested in do you want to kick it off. Otherwise maybe close the ticket?
#453 1294411369000000 thejimmyg This isn't really something that should be a CKAN ticket. It is more an ongoing thing.
#465 1294411534000000 thejimmyg Let's reopen this ticket if it comes up again. For now not an issue. Maybe also related to caching tickets #840 #841 #856 #668 #543 #540 #537.
#466 1294411633000000 thejimmyg What was the original requirement? What's wrong with the current HTTP header name? Can we not mark this as wontfix for now and re-open if needed?
#467 1294411656000000 thejimmyg Duplicate of #467
#467 1294411681000000 thejimmyg Sorry, that should have been #466.
#476 1294411741000000 thejimmyg Where did this requirement come from? Setting to wontfix for the timebeing. Feel free to re-open if there is a clear use case for this.
#477 1294411761000000 thejimmyg Duplicate of #476
#480 1294411813000000 thejimmyg We are currently updating the specification and the servers this runs on.
#371 1294411939000000 thejimmyg It is implied in this that the performance of sites should beat the QoS criteria, therefore closing #485. Ensuring this happens is an ongoing process.
#485 1294411946000000 thejimmyg Merging with #371.
#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.
#699 1294412194000000 thejimmyg I've just verified that the search pages navigation does not respect the checkboxes. Could you have a look please Friedrich?
#498 1294412520000000 thejimmyg This is something that the Drupal team would need to do. There is no ability or requirement to embed the map widget into the CKAN search system yet.
#538 1294412635000000 thejimmyg Rufus, is this something you want to take on or shall we close the ticket as wontfix?
#565 1294412708000000 thejimmyg Don't understand this one? Anyone else know about it?
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Note: See TracReports for help on using and creating reports.