Please keep me in CC since i'm not subscribed to the list.
Hi.
Just assume a lot of cheering for the approach of a simple but complete IMAP
client and fancy stories about me ever once again looking at Trojitá here.
Attached are 3 small patches, mostly regarding the UI (git format)
Notice
Am 09.07.2012, 16:47 Uhr, schrieb Jan Kundrát :
- Please use "0" instead of "NULL" (see [1], even though it isn't
terribly explicit in there).
No problem. I used to prefer 0 (or rather 0L) anyway, but and
FYI the /current/ talk around KDE is to use NULL because it can easily be
sed'ed into nu
Am 10.07.2012, 20:30 Uhr, schrieb Jan Kundrát :
- Starting in the wide layout, going to compact, the message widget gets
wider. That's correct as it has to fill up space which is suddenly
bigger. However, going back to wide, its "big width" remains and one has
to scroll even though it isn't real
Am 15.07.2012, 11:40 Uhr, schrieb Jan Kundrát :
Well, you aren't going to get away from QWidget weirdness so easy :).
Yes, am :-P
If you open a message which is called the MIME torture test [1]
"My Innocent Machine Exploded" torture test?
How did they do that in 1993?
might be even more ob
Am 15.07.2012, 14:07 Uhr, schrieb Jan Kundrát :
I see; that's probably what causes these gray areas around any messages
(right?).
Gray?
There's padding but it ought to be base colored (ie. message text
background) and is here (in any style)
But yes, this causes the padding.
So it looks
Am 16.07.2012, 02:07 Uhr, schrieb Jan Kundrát :
I'm not sure what you mean by input-transparent -- if you mean that
input events shall be delivered to the widget "below"
yes.
then nope, even a simple widget being placed in a QHBoxLayout suddenly
getting visible would be OK.
OK.
If you m
Am 16.07.2012, 12:29 Uhr, schrieb Jan Kundrát :
You'll be happy to see that support for building messages with
attachments which landed in master last night. I hope I'll have it ready
today.
Yes am.
Do you have done further work on the composer GUI or intend to or is the
field open atm?
Blast - i forgot to attach patches.
No. 1 should fix paddings (we scan up on reparenting anyway, so there's no
need for API changes at all)
No. 2 invokes link handling of the emptyView for the header - i made the
slot public but since it's stateless (could be a static if it wasn't a
slot) t
Am 19.07.2012, 12:48 Uhr, schrieb Jan Kundrát :
Hmm, so it looks like the paddings are still subtly broken under certain
circumstances -- please see a bug report [1] with an attached screenshot
[2].
Seen the shot.
Iirc the emptyView never gets a handlePageLoadFinished (thus no initial
res
Am 19.07.2012, 12:48 Uhr, schrieb Jan Kundrát :
Hmm, so it looks like the paddings are still subtly broken under certain
circumstances -- please see a bug report [1] with an attached screenshot
[2].
Could you please have a look when you have some time?
Patch attached - I *assume* that th
1) addressbook interface
2) addressbook completion
Quick Copy & Paste from my mailer - nearly no time lost ;-)
please notice that the "custom" completer was pretty much a result of the
wish to complete comma separated lists
this works, but the parser will likely rank them invalid.
I'd still
Attached is fixed patch - i forgot the "return" ...
Cheers,
Thomas
0001-replace-tr-ToCcBcc-with-RecipientKind.patch
Description: Binary data
Am 20.07.2012, 16:38 Uhr, schrieb Jan Kundrát :
Thanks, that's it. I like the QMetaObject::invokeMethod bit -- I though
one had to use QTimer::singleShot )and an explicit slot due to the
arguments) for that.
QTimer::singleShot(0, qobject, SLOT(doSome())); is in fact just a wrapper
around
Am 20.07.2012, 20:00 Uhr, schrieb Jan Kundrát :
it looks like my changes are conflicting, so patches 2-5 won't apply
i'll do a rebase on the comments then.
This is QStringBuilder, right?
Yes.
AFAIK you cannot use it in Qt 4.6
"In 4.6, an internal template class QStringBuilder has been ad
Am 21.07.2012, 00:48 Uhr, schrieb Thomas Gahr :
Someone first had to fall out of the beds - i hope that's the weather and
not age ...
It shows the default cursor if no modifier is pressed, since in this
case the user will have to pick what action to take upon dropEvent -
i.e. during drag
Am 21.07.2012, 15:51 Uhr, schrieb Thomas Gahr :
(And no, abusing and altering the shape for the LinkAction cursor
won't work in that case either ;-)
Yeah, thinking about that idea makes shrug, though I give myself half a
credit for creativity ;)
There /is/ a "legal" way abusing the LinkAction
Am 23.07.2012, 22:12 Uhr, schrieb Jan Kundrát :
Hi Thomas/Thomas,
there's a bug in how the header view works
Is there a particular reason why you put the SimplePartWidget (being an
EmbeddedWebView being a QWbview, providing nice scrolling facilities) into
an extra scrollarea?
Ie. do you in
Am 13.08.2012, 19:54 Uhr, schrieb Jan Kundrát:
Your code obviously provides better functionality; having a full-blown
address book holding structured information is great. On the other hand,
in its present form, it removes the support for this remembering of
entered data.
I'm not sure wha
on is subject to change, so if you don't agree with
my stance, please feel free and welcome to share yours.
1. Multple accounts
This is something which is being requested pretty often. I don't have a
use for this feature myself and prefer to concentrate on other areas,
but Thomas Lüb
Am 16.08.2012, 16:39 Uhr, schrieb R Ceunt :
Therefore Trojita should not only allow me to open a
meeting request (attachment) with my calendar software, but Trojita
should also allow my calendar software to send meeting
requests/updates. However, I can imagine that this would then be
implemente
Am 16.08.2012, 12:47 Uhr, schrieb Jan Kundrát :
Since he's using gmail: "no", but the mails are date sorted (the only
"nasty" thing is trojotá's queue ordering instead of stack ordering in
MVC, but that's minor and could be fixed easily with a config option)
I have to admit I don't understand
Am 16.08.2012, 12:51 Uhr, schrieb Jan Kundrát :
On 08/15/12 22:30, Peter Morgan wrote:
Since we cannot know whether the user is interested in learning an
address
or even interested in NOT adding it my proposal would be to only learn
recipient addresses with a name attached and in addition "
Am 18.08.2012, 01:27 Uhr, schrieb Peter Morgan :
For the GUI we at least agree there should be autolearning for
recipients
(outgoing mail) with a human readable name and an explicit way in the
mail
header.
I dont think that adding an outbound email should add to address book.
Eg misht...@e
Am 18.08.2012, 02:21 Uhr, schrieb Peter Morgan :
The idea is to only learn mails with a name attached, ie.
Homer Simpson
but not
al.coho...@moes.com
to precisely prevent the case where you've a mystic mail contact
without a human readable name for it.
Thats a good i
Am 18.08.2012, 13:26 Uhr, schrieb Jan Kundrát :
On 08/17/12 22:07, Thomas Lübking wrote:
Attached.
The GUI looks very simple.
Yes, i don't like it complex ;-)
Originally I assumed that you'd like to maintain separate releases of
the application
I had no concrete release pl
Am 18.08.2012, 20:08 Uhr, schrieb Peter Morgan :
hi guys
Am I out of the loop here ?
this conversation is re the address book, but where is it.. am curious
;-)
Pete
Was attached to a former mail.
Attached slightly fixed version here
Cheers,
Thomas
be.contacts.txz
Description: applicati
Am 19.08.2012, 21:59 Uhr, schrieb Peter Morgan :
hi guys,
Been hacking away at looking at the MessageView.
I think I would like to change this a little..
The FROM, TO, CC etc.. Dates, Subject I would make into a widget at the
top.
and the WebView would remain the same below.. I hasten to ad
On Freitag, 5. Oktober 2012 00:39:26 CEST Jan Kundrát wrote:
> Let's use the multiple widgets approach at first to see how that works.
Not if you'd in general prefer the container model - adding interim stuff ends
up in bit rot and makes gittannoyance when having to be reverted out of
follow-up
On Samstag, 6. Oktober 2012 15:00:28 CEST Jan Kundrát wrote:
> - *bold*, /italic/ and _underline_ formatting is lost when
> replying (they appear as "bold, italic and underline" in the
> response, without any extra formatting)
Fixed, was related to the 2nd "issue" (and i didn't care much about pr
The reply rewrapping patch had a bug (i somehow committed an interimg state
while reducing variable amount) leading to an infinite loop.
Sorry about that, attached is the correct patch (replacment, not addition)
The other patch prevents the focus from being passed around on wheel events
(what s
... affecting the composer.
1. Moved the attachments to the top right shoulder to get things more compact
and esp. save precious vertical space
2. Attachments have now full DND support - one can drag them into the text area
and the attachment view to add them and drop them out of the view to re
"Blast"
On Mittwoch, 10. Oktober 2012 15:29:11 CEST Jan Kundrát wrote:
> What is this doing -- I assume this is not related to focus on wheel?
It's the "conditional seachline deactivation" (yes, bed because of non atomic
commit)
The issue with the current behavior emerges if you filter mails (
On Mittwoch, 10. Oktober 2012 15:54:18 CEST Jan Kundrát wrote:
> ./ui_ComposeWidget.h:113:19: error: no member named 'setBuddy' in
> 'QGroupBox'
> groupBox->setBuddy("subject");
> ^
>
> Fixed manually, the patch is in.
Yes, got that as well (comes from a merge conflict), i guess i forgo
On Donnerstag, 11. Oktober 2012 18:25:37 CEST, Jan Kundrát wrote:
> On 10/10/12 18:42, Thomas Lübking wrote:
>> I don't get the assert, but run into
>> qDebug()<< "drag-and-drop: mailbox not found";
>
> Are you sure you're building in the debu
I finally figured how to false positively send by ctrl+enter - happens if you
quickly enter (newline) and then use some ctr+something shortcut (usually
ctrl+v)
Acting on keyPressEvent doesn't help either, because you can cause false
positives the other way (ctrl+v, then quickly press enter)
So
On Montag, 15. Oktober 2012 15:37:23 CEST, Jan Kundrát wrote:
>> +static QDateTime lastVersion;
>> +QDateTime lastTouched(file.exists() ?
>> QFileInfo(file).lastModified() : QDateTime());
>> +if (lastVersion < lastTouched) {
>
> If I'm reading this correctly, this will evaluate to tru
Attached is the hardened ctrl+enter patch, uses now a static inline bool
instead of the macro.
Regarding the search config / generator toolbutton
1. i don't mind the icon
2. i don't intend to discuss on whether it should be on the left or the right,
but should probably explain why i initially mo
On Freitag, 19. Oktober 2012 03:38:36 CEST, Jan Kundrát wrote:
> Anyway, there's supposed to
> be an ExLineEdit class in Qt4's demo browser application; this
> class should support embedding icons (or buttons?) in a
> QLineEdit. That would look better, in my opinion. Would you like
> to try to
On Samstag, 20. Oktober 2012 13:46:15 CEST, Jan Kundrát wrote:
> On 10/19/12 22:03, Thomas Lübking wrote:
>> Icons, I'd say.
>> This is not hard to do, but unfortunately the AutoRaise
>> buttons do not look that great in oxygen& qtcurve, ahve a
>> look.
&g
On Montag, 22. Oktober 2012 19:22:20 CEST, Jan Kundrát wrote:
> Hi Thomas,
> what is the purpose of ComposerTextEdit::notify
> (ComposerTextEdit.cpp:41) and related methods?
Copy-over from one of my own classes.
Provides some in-window notification, i thought it could be an alternative to
the c
On Sonntag, 28. Oktober 2012 07:00:45 CEST, wrote:
> First, and the biggest one, what keyboard shortcuts are defined
> (and how do I find them)
Shortcuts should be displayed next to the (context-)menu entries (Qt operates
on actions which can have various triggers such as popup menus, shortcuts
On Sonntag, 28. Oktober 2012 18:24:34 CEST, Jan Kundrát wrote:
> Fixing c) would not be trivial, but should be doable. It's just
> that it's not easy to define what the sort order should be -- I
> can definitely make it behave like THREAD=REFS by default, but
> if the user selects e.g. sort by
On Sonntag, 28. Oktober 2012 21:08:17 CEST, wrote:
>> b) Threading doesn't work at all -- that would be a new bug,
>
> threading doesn't work at all.
What capabilities are provided by your mail server?
(If you don't know and don't know hot to figure: is it a public server?)
Cheers,
Thomas
On Sonntag, 28. Oktober 2012 19:01:40 CEST, Jan Kundrát wrote:
> Thomas, do you know about some semi-reasonable Qt library which
> would manage the GUI for configuring keyboard shortcuts for me?
You mean *like* kactioncollection? No :-(
> (Short of the whole of KDE, if possible -- that's a dep
On Montag, 29. Oktober 2012 17:50:31 CEST, Jan Kundrát wrote:
> Hi Thomas,
> the attached message is a multipart/alternative of a text/plain
> and a text/html.
FTR: trojita shows "Attachment (attachment/,0) and i get an emtpy file as
Download (i fetched the message via the webmailer ;-)
> Troji
Fine with me
(reminds me to finish the multiaccount patch - i've written a complete
metamodel a month ago and just not invoked it - can please someone write me
buttkicking mails if i don't get it done this week =)
Thomas
Sorry, took me some time to finally read this ;-)
On Dienstag, 18. Dezember 2012 00:47:26 CEST, Jan Kundrát wrote:
> In Trojitá, the sender identities are not tied to the IMAP
> [...]
> a feature which picks up the correct identity when I click
> "reply".
> 1) Add a list of domains (or regula
On Donnerstag, 21. Februar 2013 15:24:58 CEST, Jan Kundrát wrote:
Thomas Lubking is currently working on this feature. We don't
have a timeframe on when it will be ready, though.
/o!oo ;-P
You're kinda lucky because i rebased it on current master last night.
I can send around a plain patch fo
On Mittwoch, 1. Mai 2013 12:04:16 CEST, Jan Kundrát wrote:
https://github.com/sreepriyachalakkal/trojita get's me ObiWan ("This is not the
webpage you're looking for"), but this:
if (edit->text().isEmpty() || parseRecipients(recipients, errorMessage)) {
p.setColor(QPalette::Base, palette(
On Mittwoch, 1. Mai 2013 13:55:00 CEST, Sreepriya Chalakkal wrote:
That link doesnt work. This is the new link to the new function
https://github.com/spchal/trojita/blob/master/src/Gui/ComposeWidget.cpp#L560
Jan, i saw the same in FindBar.cpp - doesn't fix it's wrong ;-)
Ok, whenever you wan
On Donnerstag, 2. Mai 2013 17:23:12 CEST, Jan Kundrát wrote:
An address book integration is missing from Trojita. If you
feel like contributing patches
Ah, sh** - i forgot to send around the sources for the abook GUI, did I? (Yes,
i certainly did)
Attached.
It's not integrated into trojita i
On Donnerstag, 2. Mai 2013 20:19:56 CEST, Jan Kundrát wrote:
Hi Thomas, based on your User-Agent header, you're running the
multiaccounts-2 branch. For how long have you been using it?
"Months"
* The "bye" issue and general offline state was never really resolved (you get one bye,
you're "off
On Donnerstag, 2. Mai 2013 20:20:10 CEST, Jan Kundrát wrote:
Damn, you've ruined my fundraiser. Now I'll have to do all the
boring work like adding fancy buttons to the message viewer for
free :).
Pfff... it's time consuming and requires "motivation" nevertheless.
Not the workload determines t
On Sonntag, 5. Mai 2013 04:38:53 CEST, may...@maykel.sytes.net wrote:
I have configured the thunderbird, kmail with the same account
and same settings and not problems...
a) Do you know what BURL is?
b) have you tried disabling it?
Thomas
Counterproposal attached, briefly tested.diff --git a/src/Gui/CompleteMessageWidget.cpp b/src/Gui/CompleteMessageWidget.cpp
index 6eb25d1..dcaa632 100644
--- a/src/Gui/CompleteMessageWidget.cpp
+++ b/src/Gui/CompleteMessageWidget.cpp
@@ -21,7 +21,10 @@
*/
#include "CompleteMessageWidget.h"
+#i
Hi, sorry for being late.
On Dienstag, 21. Mai 2013 14:39:36 CEST, Francesco Riosa wrote:
It was not totally satisfying for me too but it was easy and fast, I will
try to mimic the account selection next
My initial approach was to just have a QListView containing the account names with
"Add"
On Mittwoch, 5. Juni 2013 15:53:32 CEST, Pali Rohár wrote:
Hello,
in this Google Summer of Code I'd like to integrate Trojita into
KDE. This means support for KWallet, KAddressBook, dbus, mailto:,
... and also integration to Kontact (like KMail).
KWallet is accessible via DBus (which is itse
> On Wed Jun 12 06:30:02 PM Jan Kundrát wrote:
>> On Wednesday, 12 June 2013 18:02:13 CEST, Vayu wrote:
>> > Thanks, I just compiled from Git and now it is saving my
>> > positions. It doesn't save my column sort order though.
>>
>> I guess this could be added, even though it's a rather expensive
On Freitag, 14. Juni 2013 16:30:16 CEST, Jan Kundrát wrote:
I'm not sure whether DBUS is what we want here.
The traditional way for POSIX systems (including mac, but not windows) was to
have a PID lockfile and SIGHUP the running process to make it raise/activate
itself.
It's also poosible to
On Donnerstag, 13. Juni 2013 09:08:57 CEST, Jan Kundrát wrote:
Thinking about this and expanding on my domain-based filtering
idea, perhaps the button could tell the user what domains the
mail is trying to fetch from, and upon pressing it, save these
settings (the domain of the image URL) pers
On Freitag, 14. Juni 2013 19:31:50 CEST, Jan Kundrát wrote:
I was actually aware of this threat
Ok, proof damanged.
and expected the users to be aware of the risks
Depends on "users" - as long as pot. facebook an trojitá users are disjunct,
this is probably true.
But the majority of all use
On Freitag, 14. Juni 2013 21:41:14 CEST, Kevin Krammer wrote:
As far as I know there are only two contact data services: Evolution Data
Server (EDS) and Akonadi.
Data service is not excatly the same as "addressbook", where there are hundreds.
So if Trojitá was eg. to (by desktop service plugin)
On Freitag, 14. Juni 2013 22:12:42 CEST, David Lang wrote:
repeated scary warnings are worse than useless. They
desensitize users about other, more important, warnings.
I did not seriously suggest that as viable option (the workflow alone would be
nasty)
You cannot protect all the users. If
Am Samstag, 15. Juni 2013 schrieb David Lang
>> Just to avoid misunderstandings:
>> this is not about blocking image attachments (that's not a MUAs job at
all) or rendering attached images inline with textmails, but fetching and
rendering images with html mails that are *not* attached to the mail b
On Samstag, 15. Juni 2013 09:04:56 CEST, David Lang wrote:
hey, I'm running pine here, images are something I save to open
in a different tool most of the time :-) however, at $work it's
hard to stick to that and I end up having to use OWA fairly
frequently to deal with HTML messages.
pine -
On Sonntag, 16. Juni 2013 03:19:05 CEST, David Lang wrote:
arguably you can run into this with the embedded images as well.
No. If an image is actually attached (regardless of the Disposition) it resides
at your MSP.
If you mean that you can receive and display base64 encoded CP - yes, that's
On Sonntag, 16. Juni 2013 10:03:32 CEST, Kevin Krammer wrote:
KMail has a global setting to enable this but does not check it by default.
By default it displays an option inline at the top of each mail
which contains
external references.
Do you happen to know whether there's been a similar d
On Montag, 17. Juni 2013 13:48:39 CEST, Jan Kundrát wrote:
- if there's no QtKeychain at build time, disable password
storage altogether and have Trojita ask for one when needed
Wouldn't that render Trojitá "unusable" for "non-DE" users (openbox or so), eventually
even the "minor" ones (xfce,
On Sunday, 16 June 2013 13:26:24 CEST, Pali Rohár wrote:
24.6 - 30.6
* change trojita code to use functions from loaded plugins
* move internal password code into trojita plugin (and drop
internal addressbook for now, it will be ported in august)
I'll be predictably not (very) available July
On Montag, 17. Juni 2013 18:20:37 CEST, Jan Kundrát wrote:
On Monday, 17 June 2013 14:41:16 CEST, Thomas Lübking wrote:
Wouldn't that render Trojitá "unusable" for "non-DE" users
(openbox or so), eventually even the "minor" ones (xfce, lxde)?
That depend
On Dienstag, 18. Juni 2013 03:05:05 CEST, David Lang wrote:
On Mon, 17 Jun 2013, Thomas L?bking wrote:
Subject: [trojita] Re: Using QtKeychain ...
This seems like familiar territory :-)
It's not the best way possible, but do you really want to drive
away users because of this? Does that re
On Mittwoch, 19. Juni 2013 13:15:46 CEST, Jan Kundrát wrote:
It seems that QtKeychain won't help users of "lightweight" DEs
and that there are people out there who really, really like
password storage to be available all the time.
Just to reason that:
"password"
"password1"
"12345"
"secret"
On Donnerstag, 20. Juni 2013 10:52:05 CEST, David Lang wrote:
So please try and keep the addressbook plugin from being too Qt
specific and make it be something that can support a simple use
case like this or a simple local file.
Trojita uses Qt and the plugin interface will likely be the Qt o
On Donnerstag, 20. Juni 2013 12:18:19 CEST, Pali Rohár wrote:
On Thursday 20 June 2013 10:47:27 Jan Kundrát wrote:
I look at qt plugins howto and there is also used inlined virtual
destructor: https://qt-project.org/doc/qt-4.8/plugins-howto.html
Because there mentioned "FilterInterface" clas
On Donnerstag, 20. Juni 2013 14:09:05 CEST, Jan Kundrát wrote:
sort-of-kinda-solves this via qobject_cast, but that's a band
aid. The lesson I've learned is to never use an inlined virtual
destructor.
I do not recall myself suggesting such ;-)
You can omit the empty virtual destructor (when
Am Donnerstag, 20. Juni 2013 schrieb Jan Kundrát :
> Now I'm confused -- your mail suggests to omit the destructor altogether,
and at the same time you quote a document saying that one should always
include it. I agree with the quoted document :).
The important part here is that there /is/ a virt
On Donnerstag, 20. Juni 2013 16:53:15 CEST, Jan Kundrát wrote:
What I'm saying is that if you have class Foo: public QObject,
you have to declare and define a non-inlined virtual destructor
~Foo(). You're right that not marking it as "virtual" doesn't
matter. If there's no explicit ~Foo(), you
On Donnerstag, 20. Juni 2013 20:42:34 CEST, Jan Kundrát wrote:
You can also kill the ignores parameter, it doesn't appear to
be used. Thomas, do you have a comment on that, given that
originally added this?
The ignorelist is meant to have the addressbook skip entries already used in
TO/CC/BCC
On Freitag, 21. Juni 2013 11:10:49 CEST, Pali Rohár wrote:
With isValid() I mean if all required programs/libraries/other
dependences for plugin are installed.
Package resolution is a distro job (in the linux/bsd world) ie. a *plugin* that
requires akonadi should simply not be installed when
On Freitag, 21. Juni 2013 11:17:24 CEST, Pali Rohár wrote:
Hm, I think it is not needed. In plugin interface we need only
function which open contact window. And there could be edit
button or window will be itself editable (like kaddressbook or
internal trojita addressbook).
What if the user
On Freitag, 21. Juni 2013 12:17:38 CEST, Pali Rohár wrote:
On Friday 21 June 2013 12:06:52 Thomas Lübking wrote:
On Freitag, 21. Juni 2013 11:10:49 CEST, Pali Rohár wrote: ...
With akonadi I mean: There are akonadi client libraries and
akonadi server (process). You need to have akonadi
On Freitag, 21. Juni 2013 17:25:56 CEST, Kevin Krammer wrote:
Hmm, how often does it check that, or is a plugin only ReadOnly if it can
never ever write, e.g. addressbook file on a read-only optical drive?
I'd say, a simple member return will do.
The plugin can either write or not by backend d
On Freitag, 28. Juni 2013 12:35:19 CEST, Kevin Krammer wrote:
http://quickgit.kde.org/?p=clones/trojita/pali/trojita.git
I would still recommend to do at least the completion task in form of a job.
I frankly was under the impression there was agreement on this anyway?
ALso I definitly saw so
On Freitag, 28. Juni 2013 11:50:13 CEST, Pali Rohár wrote:
In interface are virtual pure functions. So then remove protected
constructor (and compiler will generate one default)?
If you can compile this, you'll be my hero ;-)
-- snip
class Pure {
public:
virtual void fo
On Freitag, 28. Juni 2013 14:20:33 CEST, Pali Rohár wrote:
On Friday 28 June 2013 14:10:35 Thomas Lübking wrote:
On Freitag, 28. Juni 2013 12:35:19 CEST, Kevin Krammer wrote: ...
Now I'm rewriting addressbook requests to use jobs. But I do not
think that we need common interface for
On Freitag, 28. Juni 2013 14:27:34 CEST, Pali Rohár wrote:
On Friday 28 June 2013 14:20:34 Thomas Lübking wrote:
On Freitag, 28. Juni 2013 11:50:13 CEST, Pali Rohár wrote: ...
I know, so I'm asking if I should remove constructor (and
compiler will generate one without parent qo
On Freitag, 28. Juni 2013 14:53:10 CEST, Pali Rohár wrote:
And for example QDBusAbstractAdaptor has constructor protected:
http://doc.qt.digia.com/4.7/qdbusabstractadaptor.html
And exactly how many pure virtual functions?
;-P
Thomas
On Freitag, 28. Juni 2013 15:06:41 CEST, Jan Kundrát wrote:
On Friday, 28 June 2013 14:31:21 CEST, Thomas Lübking wrote:
The point about a common class is to allow generic job
handlers as well as not to have to write the same stupid
"stop", "cancelled", "finished&
On Mittwoch, 10. Juli 2013 20:52:20 CEST, Caspar Schutijser wrote:
I don't know if someone has another opinion on this?
The problem with the KABC plugin is that it seems to be unusable with a few
months (but on older distros) - i don't have to sell that to google but it
sounds like a pointle
On Mittwoch, 10. Juli 2013 21:41:30 CEST, Pali Rohár wrote:
I played a lot with minimal size and on my machine there are no
scrollbars. But now I think that minimal size depends on used qt
theme and so we cannot predicate correct minmal size.
So is there any way to tell general settings tab w
Hi Pali
Following some comments on the branch.
I read the diff and wrote them in a train so it's rather brief ;-)
On a general note, anyc API in a loop of course sucks but it might still
worth trying to get rid of the nested event loops
Cheers,
Thomas
+
+namespace Common {
+
+PluginLoader::Plug
>
>> +disconnect(job, 0, 0, 0);
>> +job->stop();
>> +++ // TODO how do the get off the hash?
>
> QLineEdit is asociated with job. So other option could be to move
> job pointer to new class inhered by QLineEdit.
That doesn't scale because everytime you want to change the
assiciati
On Montag, 22. Juli 2013 22:00:30 CEST, Pali Rohár wrote:
I need to add widget m_composeButton to actionCollection().
I think you want to add the actions attached to the widget - not the widget
which is just a button but could also be a combobox or a menubar or whatever
representing a hook to
On Donnerstag, 25. Juli 2013 11:53:20 CEST, Pali Rohár wrote:
In our case (in Trojita), we have KGlobal::hasMainComponent()
What sets KGlobal::newComponentData() and why does it pass a junk name?
I'd suggest to fix that.
(The KComponentData constructors do so, when passed RegisterAsMainCompon
On Donnerstag, 25. Juli 2013 14:06:44 CEST, Pali Rohár wrote:
It looks like that shared kde4 kabc library which is linked to
kde addressbook trojita plugin (which was loaded before kwallet
plugin). Order of loading trojita plugins is now undefined. And I
do not want to have some linear orderin
On Montag, 29. Juli 2013 11:01:44 CEST, Jan Kundrát wrote:
On Wednesday, 24 July 2013 11:08:09 CEST, Pali Rohár wrote:
I rebased pali-gsoc-merge branch again with squashed fixes. It
contains replaced QEventLoop code, fixed hardcoded libraries in
CMakeLists.txt for building KDE plugins and other
Set #2 of comments:
+QAction *actionComposeMail = m_window->findChild(QLatin1String("action_compose_mail"));
+QAction *actionComposeDraft = m_window->findChild(QLatin1String("action_compose_draft"));
...
+QAction *actionReplyPrivate = m_window->findChild(QLatin1String("action_reply_pr
On Dienstag, 30. Juli 2013 23:23:13 CEST, Kevin Krammer wrote:
+void KDEAddressbookProcess::slotApplicationLoaded()
this is superfluous, you can bind the finished signal to
Might even connect the signal to deleteLater directly?
That's what i meant - it's rarely required to write a slot to del
And the rest:
+void KWalletManager::open()
+{
+if (m_isOpening)
+return;
+if (m_wallet && m_wallet->isOpen()) {
+QTimer::singleShot(0, this, SLOT(walletOpened()));
*cough* - and why does the call need to be deferred anyway?
+void KWalletManager::walletOpened(bool opened)
On Montag, 5. August 2013 16:40:02 CEST, Jan Kundrát wrote:
- The fact that qobject_cast doesn't work across shared library
boundaries is suspicious. Wasn't supporting *that* use case one
of the reasons (besides not requiring RTTI) for introducing
qobject_cast over dynamic_cast?
Did i miss th
1 - 100 of 202 matches
Mail list logo
