{23} Trac comments (3729 matches)

Results (3501 - 3600 of 3729)

Ticket Posixtime Author Newvalue
#2703 1342624713000000 aron.carroll Just needs a new action that loads the same resource_form.html template with an empty stages array.
#2703 1342687332000000 toby created on dev branch
#2704 1342707398000000 toby @ira, again not sure this makes sense from a ux perspective @aron, reassign to me if you want it doing should be trivial
#2704 1342709081000000 shevski Not an ideal UX solution, but important in the interim!
#2704 1342710830000000 toby @ira, this isn't really needed then! there is no point doing work if we need to do it again editing resources is easy enough now anyway I suggest we just close this ticket as a won't fix and #2705 as well while we're at it. I'm also not happy about these being added to the phase 2 milestone when we are at the end of finishing it and these are not bugfixes or critical functionality
#2704 1342716968000000 toby need some thought
#2704 1343914890000000 aron.carroll This relates to #2705. We need to decide what happens when the user clicks on one of the sidebar links. At the moment they'll get a warning if they've edited the form telling them they'll lose thier changes. Perhaps we should submit the form using ajax, showing a spinner to the user, then send them on to the next resource. This is the same situation as if we add an "add new resource" button to the listing.
#2704 1343948227000000 toby @aron, I don't like auto submit as people may think (reasonably) that their changes are discarded. We could catch all clicks and then ask the user if they've changed something but this feels a little crazy - better than the browser dialog. hmmm maybe we should discuss this ticket with the other backlog of undecided/unsure personally I think we need to review the whole edit/view/delete logic as it feels real poor from my perspective - I need to edit something to delete it - that is a wtf imho let's work out what we are doing before wasting energy on this
#2704 1352205939000000 johnmartin Fixed in master
#2705 1342707341000000 toby @ira, I'm not sure about this if some one is editing a resource moving to edit the dataset seems wrong - a back to resource makes sense to me but not moving to the edit page @aron, If you want this then assign it to me if you're busy
#2705 1342716980000000 toby need some thought
#2705 1352205954000000 johnmartin Fixed in master
#2706 1342707718000000 toby @aron, I've done this for you
#2707 1344504846000000 toby Formmating done other requests moved to new tickets as low priority for effort needed Title fixes translations etc http://trac.ckan.org/ticket/2822 Ordering of fields http://trac.ckan.org/ticket/2823
#2709 1342626212000000 markw but some of the relevant info is here: http://ckan.org/2012/05/30/atom-feeds/
#2710 1342630558000000 aron.carroll Displayed in a lightbox as of f347f11
#2710 1342779460000000 toby @aron, we need to have this styled a little better when non-js user maybe it needs a little thought so that the lightbox gets a snippet and the non-js user gets the snippet embedded in a page so consisten styling and looks like part of ckan
#2710 1342784018000000 toby @aron I've done this - it needs a bit of work ie breadcrumb and you won't like what I did with the css classes/js- a better solution is needed but this gives a quick example that we can improve on let me know your thoughts https://github.com/okfn/ckan/commit/cec57f36094327611ff9bc8935c061a61ca3e961
#2710 1343233018000000 aron.carroll Fixed this up, assigning back to you for one small issue. The h.url_for('datastore_read') is returning a broken url: https://github.com/okfn/ckan/blob/2710-nicer-api-page/ckan/controllers/package.py#L1139-1140 It looks like this: {{{http://localhost:5000datastore_read?id=794dab29-703f-4ab5-9043-ace58c87a2ce}}}
#2710 1343318314000000 aron.carroll Ignore above comment, I didn't have datastore.enabled = 1 in the ini
#2711 1352205968000000 johnmartin Fixed in master
#2713 1343656865000000 aron.carroll Fixed in 32110e7
#2714 1342706752000000 toby @ira, this isn't a phase 2 item - moving to phase 3 can you view the document when you download it? The message says it thinks it is corrupt
#2714 1342794133000000 shevski Got it working http://demo.ckan.org/dataset/test-dataset/resource/14244c9c-7dcd-43f0-9df0-b363baa2fa8d
#2714 1342794216000000 toby @aron what was the problem?
#2714 1352206138000000 johnmartin This seems fixed... I'm closing this.
#2715 1342716931000000 toby live on s031
#2716 1342716474000000 toby fixed and on s031
#2718 1342785406000000 toby @shevski can you use links to s031 for stuff like this as it is the testing serve and i can play with the data more - unless the error is specific to demo.ckan.org ------------- with this issue it is actually wider than you suggest as I cannot edit the group that the dataset is in either. We need to discuss this one as I don't know the best approach should groups select datasets or datasets groups or both - how do we remove datasets from groups - who can do this etc - maybe next week?
#2718 1342788150000000 toby related to #2699
#2720 1343214149000000 toby minor fix so seems good to do it now
#2720 1343219999000000 aron.carroll Padding fixed in 1b41a9b the user list is part of #2451
#2721 1342948657000000 shevski E.g. 'test-group' was created & deleted
#2721 1343031299000000 toby @ira, this is an issue for sysadmins only - i'll look at it anyhow
#2721 1343037422000000 toby moving to phase 3 a s fundamental ckan issue needs team discussions
#2722 1342948991000000 shevski In fact, this is the wrong bug. Search doesn't seem to work *at all* in groups Tried to search http://s031.okserver.org:2375/group/data-explorer?q=&sort=views_recent+desc Bug 1: no results Bug 2: the text display covers the order by box
#2722 1343036871000000 toby @ira I have disabled the sort order of group datasets in phase 2 as it is totally broken - I'll re-add it in phase 3 Also the comment you made you added bug2 - text covers order by box -PLEASE DO NOT DO THIS - all it does is make the tickets really hard to deal with - if this becomes a habit I will close all such tickets as INVALID as it makes trac unusable for developers - THIS IS REALLY IMPORTANT - cheers @aron, can you watch for feature creep in tickets and let me know if it happens - thanks
#2722 1343060105000000 shevski Hey, I only mentioned bug 2 here because I'm pretty certain this already exists as a separate ticket. Have definitely reported it before, but i think someone else created the ticket - so just wanted to remind that this hasn't been fixed either yet. What's the best way to do this in future?
#2722 1343062752000000 toby @ira, It is much easier to work with tickets that only cover one thing - If something is found then the easiest thing is to add a new ticket. If you feel it's related to another ticket then you can add that to the new ticket eg related to #1234 If tickets need combining it is easy for us devs to do it. Basically adding something to a ticket is generally a bad thing. Sometimes it's fair enough if something is missing for a new feature being reviewed etc - but even then a new ticket is generally not a problem. Small tickets are nice to work on as we can close them quick etc. The old [super] style tickets have been banned as they generally lead to owners becoming demotivated and unproductive. It's always nice if you can find if a ticket already exists but I don't mind you opening a few duplicates if you can't find the ticket - for the demo you can probably just skim the tickets in the milestones as there aren't too many - but as you know I try to review them quite often. Also please don't reopen really old tickets just add a new one If it's been closed in the last month or so then feel free to re-open if you're sure it's not just waiting for release - you can always check with the last owner of the ticket. hope that makes sense
#2722 1344358667000000 toby @aron, can you re-add the dropdown and sort the overlap issue not sure which template it's in
#2722 1344510010000000 aron.carroll Done in 8769233. I think the overlap issue was fixed for the dataset search a while ago. Let me know if it remains.
#2724 1343030687000000 toby 1) this is because it is an incomplete dataset that has not been fully added 2) please explain where you mean - we need a url 3) These should not be seen reassigning to toby
#2725 1343030773000000 toby @ira, This is a ckan issue not demo-theme I think we need a ckan day - wider team to look at many of these style issues
#2726 1342949999000000 shevski Point 2 relies on the Chrome previews of PDFs, but can we not use this ability?
#2726 1343030906000000 toby @ira, I think this is probably linked to the qa plugin and auto detecting formats rather than believing the user.
#2727 1343031501000000 toby i'll do this
#2728 1343031111000000 toby @ira, This is because you are a sysadmin - not sure why ckan does this but i'll look at it with the other group search tickets moving to phase 3 as not a user viewable issue
#2728 1345023944000000 toby Issue with not re-indexing datasets when group is updated - needs proper fix as part of phase 4
#2729 1342951176000000 shevski Also, why does searchign for format:csv not work? http://s031.okserver.org:2375/dataset?q=format%3ACSV&sort=relevance+asc
#2730 1343031481000000 toby i'll do this
#2730 1343032612000000 toby fixed will put on demo site later today
#2731 1343051608000000 markw Also, http://offener.datenkatalog.at, also called http://at.ckan.net, gives me a 500 error, including the words "Please contact the server administrator, [no address given]".
#2734 1343061727000000 toby Shevski, I think this is an old dataset it works fine with new ones. I think this is a corrupt data issue. also HTML, is fine as a format please reopen if you can reproduce with a dataset that has been created since the update today
#2734 1343061994000000 shevski I did this with a dataset I just created, 5 mins before creating this ticket Tried it again with a new datasrt, happened again. It's when you're editing a resource
#2734 1343063266000000 toby @ira, can you try to work out exactly how to cause this problem starting with `add a dataset` and let me know how you break things tell me all the fields you filled in - also try to get it to happen using as little effort as possible eg add dataset http://s031.okserver.org:2375/dataset/new title = mooooo click next:add data resource = moo format = csv click `save & add another` click on resource just created in sidebar to edit change format to `HTML,` click `Save & add another` etc that all works fine for me what am i doing wrong? thanks also please try using a non sysadmin account
#2734 1343121017000000 shevski Wierd. It always breaks for me. Maybe we go through this on skype. I just did it with a non-admin account. Steps: 1. Create new dataset. Name whatever, resource it a link (I use http://trac.ckan.org/ticket/2734) e.g. http://s031.okserver.org:2375/dataset/format-error-test 2. Once you've *finished* the dataset creation process you are on the main dataset page. Click on the resource you added. 3. Click on "edit" 4. Change resource field to "html," for example and click on "update resource" 5. Error shows, clicking on "update resource" doesn't do anything. How about I do it and you watch via screen share?
#2734 1343121694000000 toby cool I can reproduce fix will be today
#2734 1343126896000000 toby @ira, fixed and uploaded to s031 re-open if still broken
#2734 1343127101000000 shevski ace!
#2736 1346670097000000 ross Going to be reworked without celery.
#2737 1343135305000000 toby @ira, that url is broken. I'll fix this
#2737 1343136141000000 aron.carroll Done in bc93031
#2738 1343124741000000 toby the resource has no webstore_url and therefore no data api I assume that some data is in the datastore but personnaly I don't know the data closing as this is the correct behaviour as far as is known - Ie complete guess ;p but that is how it is done in ckan core
#2738 1343125335000000 shevski in CKAN core there is a tickbox to add to datastore (which we don't have anywhere) David: how should this work?
#2739 1343135348000000 toby we need all the help text - moving to phase 3
#2739 1343817919000000 aron.carroll Added in 038f4f2
#2740 1343123431000000 shevski breadcrumb should be /datasets/ dataset name / add data
#2740 1343135455000000 toby i'll fix this
#2741 1346841059000000 johnmartin Awaiting merge into demo branch
#2741 1352205988000000 johnmartin Fixed in master
#2742 1352206081000000 johnmartin Fixed in master
#2743 1343130830000000 toby @ira, I've changed the wording on the button too 'Add Dataset to Group' Adds the group auto-magically this is live on s031 for testing fun
#2743 1343131092000000 shevski brill, works great, thanks!
#2744 1343211810000000 shevski should probably be same width of format field, e.g. here: http://s031.okserver.org:2375/en/dataset/new_resource/gold-prices
#2744 1343219298000000 aron.carroll Fixed in 9472a94
#2746 1343211146000000 aron.carroll For some reason the test servers aren't serving the demo.js file. Assigning to Toby for investigation.
#2748 1344503744000000 aron.carroll Moving this to phase 4 until we decide what to do about the edit/save workflow
#2749 1343229716000000 aron.carroll Fixed in 27af570. Fix was simple just don't define a title in the base template. An alternative would have been to wrap the title portion in a block {{{ {% block title %}{% block page_title %}My Subtitle{% endblock %} - {{ super() }}{% endblock %} }}}
#2749 1343232304000000 toby I like the block page_title approach but not too fussed as long as this instance is working
#2751 1343597577000000 shevski Do we need two versions? 1 for thedatahub - where we stick to having "groups" resources etc and 1 for datasuite where groups are re-named as "publishers" and more emphasis on data files vs resources?
#2751 1344015002000000 toby This is now possible via the `paster trans mangler` the initial verdict is that our translation coverage is extremely poor and we need a full template review - maybe there is a jinja2 i18n solution but I think not without `{% trans %}` Moving to phase 4 - also I'm just holding this ticket and will be allocating to aron's successor Publishers vs Groups is more complex than I think is appreciated, is it really worth the time and effort just to make everyone using ckan confused? Anyhow that discussion is not part of this ticket as it is not in-scope
#2751 1344240974000000 aron.carroll I tried {{{paster trans mangler -c datahub.ini}} I also had to {{{ pip install polib }}} and run {{{ python setup.py egg_inf }}} I find it hard to believe out coverage is that poor, I've just been through a random selection of templates and they all have appropriate calls to {{{_()}} or {{{ {% trans %} }}}. Will look once I get this running.
#2751 1344243046000000 toby @aron, The pip install polib is needed this is a deployment requirement so not in the pip-requirements.txt Not sure where it should live we use it elsewhere so may already be in there somewhere not sure why you needed the egg_inf
#2752 1343219871000000 aron.carroll Assigning to toby to look at this crazyness
#2752 1343222618000000 toby fixed in stable
#2754 1343226309000000 toby @shevski as a reminder url please - I'll find the page but it makes my job harder
#2755 1344505304000000 toby this is dependent on #2790 (pending) and #2356 (waiting merge) other fixes have been done therefore closing this even if the effects won't be seen till those fixes are merged
#2756 1343231863000000 seanh See commits: f5043a2, 2280428, a8eedda (the ticket number wasn't in the commit messages, sorry)
#2759 1343651082000000 aron.carroll Fixed in 16bb516
#2764 1343320440000000 rgrp Duplicate of #2732
#2765 1343735917000000 toby @aron, I've added a data summary The data errors for empty fields are not happening due to our use of repoze.who for logins which is crap and doesn't provide feedback re loggin details so not possible without hacking repoze which i'm not doing. Could you look at cleaning up the error summary we don't supply a field name (key) as not relevant it would be nice if we could hide the : also removing the invalid fields message would be good to hide maybe add this as an option to the macro
#2765 1343817858000000 aron.carroll Cool, closed as of 99b8c4f. I fixed the error summary display and highlighted the fields in red to show they were incorrect.
#2766 1343318862000000 shevski e.g. see 'ff' dataset http://s031.okserver.org:2375/en/user/shevski shows up in activity stream
#2767 1343645517000000 aron.carroll Closed as of 25d7499. All form.select() calls now require a tuple/list in the form. {{{ options = ( {"value": "opt1-value", "text": "Option One Text"}, {"value": "opt2-value", "text": "Option Two Text"} ) }}
#2769 1343328989000000 ross You can view this at http://test.ckan.org/apps to get an idea of how simple the form is. Have also moved the templates out of the package folder into their own now.
#2769 1343377760000000 toby @shevski, we will probably want a design for this - maybe wait till we have the basics up and running
#2769 1343734083000000 toby rough cut down needs some design love but new ticket for that http://trac.ckan.org/ticket/2778
#2770 1343389747000000 seanh Done (but not log messages on history page, won't do those)
#2771 1343392265000000 seanh I've started working on this on a branch, will get it finished as soon as I can
#2771 1350303564000000 seanh Branch is here, but not finished yet: https://github.com/okfn/ckan/tree/2750-add-docs-and-examples-for-idatasetform-and-igroupform
Note: See TracReports for help on using and creating reports.