You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@bloodhound.apache.org by Apache Bloodhound <de...@bloodhound.apache.org> on 2013/07/30 18:29:59 UTC
[Apache Bloodhound] #618: Parent/child updates for QCT
#618: Parent/child updates for QCT <select />
--------------------------+-----------------
Reporter: olemis | Owner:
Type: enhancement | Status: new
Priority: major | Version: 0.6.0
Resolution: |
--------------------------+-----------------
There's something wrong with quick ticket user experience once multi-
product support is installed in place .
The fact is that QCT will display field values for the current
(product/global) context . If a ticket is created for a different product
the set of components / versions / milestones / ... will not be the same .
This might lead to one of (at least) three cases .
1. missing value UI due to resource name mismatch
2. wrong/invalid field value reflected in UI , in cases where field
value is
shown anyway even on mismatch
3. Trac errors e.g. resource not found
--
Ticket URL: <https://issues.apache.org/bloodhound/ticket/618>
Apache Bloodhound <https://issues.apache.org/bloodhound/>
The Apache Bloodhound issue tracker
Re: [Apache Bloodhound] #618: Parent/child updates for QCT
Posted by Apache Bloodhound <de...@bloodhound.apache.org>.
#618: Parent/child updates for QCT <select />
--------------------------+-------------------------------
Reporter: olemis | Owner: nobody
Type: enhancement | Status: new
Priority: major | Milestone:
Component: ui design | Version: 0.6.0
Resolution: | Keywords: qct ticket fields
--------------------------+-------------------------------
Changes (by olemis):
* keywords: => qct ticket fields
* owner: => nobody
* component: => ui design
Old description:
> There's something wrong with quick ticket user experience once multi-
> product support is installed in place .
>
> The fact is that QCT will display field values for the current
> (product/global) context . If a ticket is created for a different product
> the set of components / versions / milestones / ... will not be the same
> . This might lead to one of (at least) three cases .
>
> 1. missing value UI due to resource name mismatch
> 2. wrong/invalid field value reflected in UI , in cases where field
> value is
> shown anyway even on mismatch
> 3. Trac errors e.g. resource not found
New description:
There's something wrong with quick ticket user experience once multi-
product support is installed in place .
The fact is that QCT will display field values for the current
(product/global) context . If a ticket is created for a different product
the set of components / versions / milestones / ... will not be the same .
This might lead to one of (at least) three cases .
1. missing value UI due to resource name mismatch
2. wrong/invalid field value reflected in UI , in cases where field
value is shown anyway even on mismatch
3. Trac errors e.g. resource not found
--
--
Ticket URL: <https://issues.apache.org/bloodhound/ticket/618#comment:1>
Apache Bloodhound <https://issues.apache.org/bloodhound/>
The Apache Bloodhound issue tracker
Re: [Apache Bloodhound] #618: Parent/child updates for QCT
Posted by Apache Bloodhound <de...@bloodhound.apache.org>.
Re: [Apache Bloodhound] #618: Parent/child updates for QCT
Posted by Apache Bloodhound <de...@bloodhound.apache.org>.
#618: Parent/child updates for QCT <select />
--------------------------+---------------------------------
Reporter: olemis | Owner: gjm
Type: enhancement | Status: accepted
Priority: critical | Milestone: Release 8
Component: ui design | Version: 0.6.0
Resolution: | Keywords: qct, ticket, fields
--------------------------+---------------------------------
Comment (by gjm):
r1541127 changes the qct behaviour to start with no optional selections
and revert to this state after a timeout on either submission or
cancelling the form (other closing methods should keep the current
selections).
--
Ticket URL: <https://issues.apache.org/bloodhound/ticket/618#comment:9>
Apache Bloodhound <https://issues.apache.org/bloodhound/>
The Apache Bloodhound issue tracker
Re: [Apache Bloodhound] #618: Parent/child updates for QCT
Posted by Olemis Lang <ol...@gmail.com>.
On 10/7/13, Apache Bloodhound <de...@bloodhound.apache.org> wrote:
> #618: Parent/child updates for QCT <select />
[...]
>
> > Prioritize work on this ticket considering [http://goo.gl/NiZ96W this
> message].
>
> I don't suppose we could stop using shortened links where they are not
> necessary?
>
> The thread in question appears to have been silent after [http://mail-
> archives.apache.org/mod_mbox/bloodhound-
> dev/201309.mbox/%3C523B38B1.9010104@wandisco.com%3E this message]
jftr , links to archives.apache.org like the one above are not
rendered correctly by e-mail clients hence being hard to follow ...
> which
> suggests the following behaviour for QCT fields:
>
> * On page load, QCT forces the user to make selections;
> * Immediately after a ticket submission, the QCT retains the previous
> ticket selections for some specified timeout;
> * If the QCT is reopened before the timeout ends, the timeout is ignored;
> * If the QCT is again closed instead of submitting, the timeout is reset.
>
That's ok . Nevertheless I think it's worth to mention that the
initial aim of this ticket was to update component / version / ...
fields in QCT after changing product selection so that fields included
in form submission requests will match the definitions in selected
product context ... but I guess this will be the next move .
PS: From a QA perspective I'm not sure that we can write a functional
test atm for this kind of dynamic client-side JS behaviors .
--
Regards,
Olemis - @olemislc
Re: [Apache Bloodhound] #618: Parent/child updates for QCT
Posted by Apache Bloodhound <de...@bloodhound.apache.org>.
#618: Parent/child updates for QCT <select />
--------------------------+---------------------------------
Reporter: olemis | Owner: gjm
Type: enhancement | Status: accepted
Priority: critical | Milestone: Release 8
Component: ui design | Version: 0.6.0
Resolution: | Keywords: qct, ticket, fields
--------------------------+---------------------------------
Changes (by gjm):
* owner: nobody => gjm
* status: new => accepted
Comment:
Replying to [comment:5 olemis]:
> Prioritize work on this ticket considering [http://goo.gl/NiZ96W this
message].
I don't suppose we could stop using shortened links where they are not
necessary?
The thread in question appears to have been silent after [http://mail-
archives.apache.org/mod_mbox/bloodhound-
dev/201309.mbox/%3C523B38B1.9010104@wandisco.com%3E this message] which
suggests the following behaviour for QCT fields:
* On page load, QCT forces the user to make selections;
* Immediately after a ticket submission, the QCT retains the previous
ticket selections for some specified timeout;
* If the QCT is reopened before the timeout ends, the timeout is ignored;
* If the QCT is again closed instead of submitting, the timeout is reset.
--
Ticket URL: <https://issues.apache.org/bloodhound/ticket/618#comment:6>
Apache Bloodhound <https://issues.apache.org/bloodhound/>
The Apache Bloodhound issue tracker
Re: [Apache Bloodhound] #618: Parent/child updates for QCT
Posted by Apache Bloodhound <de...@bloodhound.apache.org>.
Re: [Apache Bloodhound] #618: Parent/child updates for QCT
Posted by Apache Bloodhound <de...@bloodhound.apache.org>.
#618: Parent/child updates for QCT <select />
--------------------------+---------------------------------
Reporter: olemis | Owner: gjm
Type: enhancement | Status: accepted
Priority: critical | Milestone: Release 8
Component: ui design | Version: 0.6.0
Resolution: | Keywords: qct, ticket, fields
--------------------------+---------------------------------
Comment (by olemis):
Replying to [comment:7 gjm]:
> Something like the following might do the job if we want to include
timeouts (2 minutes here):
>
[...]
>
> If the behaviour from this is acceptable
I've been playing with it for a few minutes and found nothing weird ...
> that still seems to leave a question of whether we should treat non-
cancelling form closure in a different way.
... IMO better discussed via mailing list .
--
Ticket URL: <https://issues.apache.org/bloodhound/ticket/618#comment:8>
Apache Bloodhound <https://issues.apache.org/bloodhound/>
The Apache Bloodhound issue tracker
Re: [Apache Bloodhound] #618: Parent/child updates for QCT
Posted by Apache Bloodhound <de...@bloodhound.apache.org>.
#618: Parent/child updates for QCT <select />
--------------------------+---------------------------------
Reporter: olemis | Owner: nobody
Type: enhancement | Status: new
Priority: critical | Milestone: Release 8
Component: ui design | Version: 0.6.0
Resolution: | Keywords: qct, ticket, fields
--------------------------+---------------------------------
Changes (by olemis):
* priority: major => critical
* milestone: => Release 8
Comment:
Prioritize work on this ticket considering [http://goo.gl/NiZ96W this
message].
--
Ticket URL: <https://issues.apache.org/bloodhound/ticket/618#comment:5>
Apache Bloodhound <https://issues.apache.org/bloodhound/>
The Apache Bloodhound issue tracker
Re: [Apache Bloodhound] #618: Parent/child updates for QCT
Posted by Apache Bloodhound <de...@bloodhound.apache.org>.
#618: Parent/child updates for QCT <select />
--------------------------+-------------------------------
Reporter: olemis | Owner: nobody
Type: enhancement | Status: new
Priority: major | Milestone:
Component: ui design | Version: 0.6.0
Resolution: | Keywords: qct ticket fields
--------------------------+-------------------------------
Description changed by olemis:
Old description:
> There's something wrong with quick ticket user experience once multi-
> product support is installed in place .
>
> The fact is that QCT will display field values for the current
> (product/global) context . If a ticket is created for a different product
> the set of components / versions / milestones / ... will not be the same
> . This might lead to one of (at least) three cases .
>
> 1. missing value UI due to resource name mismatch
> 2. wrong/invalid field value reflected in UI , in cases where field
> value is shown anyway even on mismatch
> 3. Trac errors e.g. resource not found
New description:
There's something wrong with quick ticket user experience once multi-
product support is installed in place .
The fact is that QCT will display field values for the current
(product/global) context . If a ticket is created for a different product
the set of components / versions / milestones / ... will not be the same .
This might lead to one of (at least) three cases .
1. missing value in UI due to resource name mismatch
2. wrong/invalid field value reflected in UI , in case it is shown
anyway even on mismatch
3. Trac errors e.g. resource not found
--
--
Ticket URL: <https://issues.apache.org/bloodhound/ticket/618#comment:2>
Apache Bloodhound <https://issues.apache.org/bloodhound/>
The Apache Bloodhound issue tracker
Re: [Apache Bloodhound] #618: Parent/child updates for QCT
Posted by Apache Bloodhound <de...@bloodhound.apache.org>.
#618: Parent/child updates for QCT <select />
--------------------------+---------------------------------
Reporter: olemis | Owner: gjm
Type: enhancement | Status: accepted
Priority: critical | Milestone: Release 8
Component: ui design | Version: 0.6.0
Resolution: | Keywords: qct, ticket, fields
--------------------------+---------------------------------
Comment (by gjm):
Something like the following might do the job if we want to include
timeouts (2 minutes here):
{{{
#!diff
Index: bloodhound_theme/bhtheme/htdocs/js/theme.js
===================================================================
--- bloodhound_theme/bhtheme/htdocs/js/theme.js (revision 1529833)
+++ bloodhound_theme/bhtheme/htdocs/js/theme.js (working copy)
@@ -50,6 +50,14 @@
});
*/
+ function checkSelections () {
+ return $.inArray('', $('#qct-box select').map(function() {return
$(this).val();})) == -1;
+ }
+ $('#qct-create').attr("disabled", !checkSelections());
+ $('#qct-box select').change(function () {
+ $('#qct-create').attr("disabled", !checkSelections());
+ });
+
$('#qct-inline-newticket').click(function() {
$('#qct-inline-notice-success, #qct-inline-notice-error').hide();
@@ -100,9 +108,24 @@
}
// Clear input controls inside quick create box
+ var timeout;
+ $('#qct-newticket').click(function () {
+ if (timeout) {
+ clearTimeout(timeout);
+ }
+ });
function qct_clearui() {
- $('#qct-form input[name!="__FORM_TOKEN"], #qct-form select, #qct-
form textarea').val('');
- $('#qct-inline-form input[name!="__FORM_TOKEN"], #qct-inline-form
select, #qct-inline-form textarea').val('');
+ $('#qct-form input[name!="__FORM_TOKEN"], #qct-form
textarea').val('');
+ $('#qct-inline-form input[name!="__FORM_TOKEN"], #qct-inline-form
textarea').val('');
+ $('#qct-create').attr("disabled", !checkSelections());
+ if (timeout) {
+ clearTimeout(timeout);
+ }
+ timeout = setTimeout(function () {
+ $('#qct-form select').val('');
+ $('#qct-inline-form select').val('');
+ $('#qct-create').attr("disabled", !checkSelections());
+ }, 120000);
}
// We want to submit via #qct-create
}}}
If the behaviour from this is acceptable that still seems to leave a
question of whether we should treat non-cancelling form closure in a
different way.
--
Ticket URL: <https://issues.apache.org/bloodhound/ticket/618#comment:7>
Apache Bloodhound <https://issues.apache.org/bloodhound/>
The Apache Bloodhound issue tracker
Re: [Apache Bloodhound] #618: Parent/child updates for QCT
Posted by Apache Bloodhound <de...@bloodhound.apache.org>.
#618: Parent/child updates for QCT <select />
--------------------------+---------------------------------
Reporter: olemis | Owner: antony
Type: enhancement | Status: assigned
Priority: critical | Milestone: next 0.x
Component: ui design | Version: 0.6.0
Resolution: | Keywords: qct, ticket, fields
--------------------------+---------------------------------
Comment (by antony):
This should now be fixed by:
https://svn.apache.org/viewvc?view=revision&revision=r1599098
--
Ticket URL: <https://issues.apache.org/bloodhound/ticket/618#comment:13>
Apache Bloodhound <https://issues.apache.org/bloodhound/>
The Apache Bloodhound issue tracker
Re: [Apache Bloodhound] #618: Parent/child updates for QCT
Posted by Apache Bloodhound <de...@bloodhound.apache.org>.