telling a data engine to reload a source

[Weng Xuetian]

On Fri, Feb 10, 2012 at 7:03 AM, Eric Mesa <ericsbinaryworld at gmail.com> wrote:
> On Thu, Feb 9, 2012 at 8:48 AM, Aaron J. Seigo <aseigo at kde.org> wrote:
>>
>>
>> the plasmoid (almost) never tells the engine when to reload. how can it
>> when
>> it is simply a visualization of the data? if the visualization knows such
>> things, it is not longer a visualization. if the dataengine does not know
>> such
>> things then it is no longer a true manager of data.
>>
>> no. the dataengine itself must know when and how to reload data. the only
>> thing a visualization can do is request updates every N ms (time based
>> updates).
>>
>> --
>> Aaron J. Seigo
>
>
> So what I really should do is set the plasmoid to request updates every
> 86,400,000 seconds (1 day) if I'd want the data to be fresh every day?
>
Well, yes.

You could check the microblog plasmod:
http://quickgit.kde.org/index.php?p=kdeplasma-addons.git&a=blob&hb=HEAD&f=applets%2Fmicroblog%2Fmicroblog.cpp
It's using: m_engine->connectSource(profileQuery, this,
m_historyRefresh * 60 * 1000);

If you need some feature such as refresh manually, you could add a
ServiceJob to the data engine.
http://quickgit.kde.org/index.php?p=kdeplasma-addons.git&a=tree&f=dataengines%2Fmicroblog
Microblog dataengine also has a ServiceJob called refresh, in order to
trigger refresh on the plasmoid side.