Python3 API generation

[Hans-Peter Jansen]

Hi Detlef,

I just updated the eric package for openSUSE Tumbleweed. 

During that process I noticed, that you omitted the python3 API files,
which I do understand, since they vary greatly between versions. 

Can you give me an example, how you create the API files for a python 
interpreter? My plan is to generate the them for the default interpreter at 
least during packaging.

Thanks and cheers,
Pete
--
Life without chameleons is possible, but pointless.

Re: [Eric] ANN: eric 20.01 released

[Hans-Peter Jansen]

Hi Detlev,

Am Mittwoch, 1. Januar 2020, 12:48:14 CET schrieb Detlev Offenbach:
> Hello Pythonistas,
> 
> I just uploaded eric 20.01. It fixes some bugs and adds these new features.
>  *  MicroPython
>  *  added entries to the ESP menu to show the Chip ID, Flash ID and
> MAC Address *  added entries to the ESP menu to backup and restore a
> firmware *  Web Browser
>  *  added support for viewing PDF documents inside the web
> browser (requires PyQt/Qt 5.13 or better)
> 
> It is available via the eric web site or PyPI.
>
> https://eric-ide.python-projects.org/index.html[1]
> https://pypi.org/project/eric-ide/[2]

There's a minor issue with eric's versioning. As you might know, RPM uses 
something along the lines of:

https://sourceforge.net/projects/eric-ide/files/%{distname}/stable/%{version}/
%{distname}-%{version}.tar.gz

to locate the source, where distname is defined as eric6 and version as 20.1.

This is tested during package submit to Factory (eg. next openSUSE release/
Tumbleweed).

Unfortunately, this fails, because the directory after stable is named 20.01 
now (vs. 20.1). I solved this by inventing distver, but it would be nice to 
avoid such confusion in future releases.

Thanks && Cheers,
Pete


___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] [QScintilla] [eric6] small glitch with fix and some notes

[Hans-Peter Jansen]

Am Mittwoch, 4. Dezember 2019, 18:52:41 CET schrieb Detlev Offenbach:
> Am Mittwoch, 4. Dezember 2019, 10:28:43 CET schrieb Hans-Peter Jansen:
> > 
> > Habe leider gestern ein paar Mal den Fall gehabt, dass Ctrl+Alt+TAB wieder
> > nicht funktionierte (jetzt allerdings mit aktivierten jedi calltips und
> > completions). Die beiden focus-betreffenden Patches habe ich natürlich
> > übernommen.
> 
> Calltips und Completions waren das andere Fokus Problem. Bitte schick mir
> die Schrittfolge, um dieses Phänomen zu reproduzieren.

Hab dazu noch keine Systematik gefunden, weil es nur sehr sporadisch auftritt.

Hier mal zwei reproduzierbare Probleme (habe jetzt erst einmal die 
Jedi-Calltips 
deaktiviert, also nur noch QScintilla calltips/completions aktiv).

Wenn eine completion angezeigt wird, und man in ein anderes Fenster umschaltet, 
dann bleibt diese aktiv. Selbst wenn man auf einen anderen Desktop umschaltet, 
also etwa so: (bitte ungewohntes Farbschema ignorieren!)



Und ein weiteres Problem: Funktion "Neue Dokumentenansicht" macht nichts, 
"Neue Dokumentenansicht (in neuem Abschnitt)" erzeugt einen neuen Splitter, 
aber 
keine Dokumentenansicht, d.h. ich habe bislang keinen Weg gefunden, zwei 
Ansichten auf eine Datei zu bekommen.

Diese beiden behindern den Workflow doch schon erheblich.

> > Heute morgen hat mich noch dies hier angesprungen:
> > :
> > wrapped C/C++ object of type Editor has been deleted
> > --
> > --  File
> > "/usr/lib/python3.7/site-packages/eric6/ViewManager/ViewManager.py", line
> > 5160, in saveCurrentEditor
> > 
> > self.saveEditorEd(aw)
> >   
> >   File
> >   "/usr/lib/python3.7/site-packages/eric6/ViewManager/ViewManager.py",
> > 
> > line 5148, in saveEditorEd
> > 
> > ok = ed.saveFile()
> >   
> >   File "/usr/lib/python3.7/site-packages/eric6/QScintilla/Editor.py", line
> > 
> > 3399, in saveFile
> > 
> > self.extractTasks()
> >   
> >   File "/usr/lib/python3.7/site-packages/eric6/QScintilla/Editor.py", line
> > 
> > 2791, in extractTasks
> > 
> > txtList = self.text().split(self.getLineSeparator())
> 
> Das hatte ich hier noch auf keinem meiner Systeme (Windows 10, versch.
> Linux, macOS). Ist das Verhalten reproduzierbar? Welche Schritte muss ich
> ausführen, um das zu erzeugen?

Ist nicht mehr aufgetreten seitdem... :-( 

BTW: Nach einem Tumbleweed Upgrade hatte zypper python3-qt4 deinstalliert (TW 
selbst ist mittlererweise Qt4 free). Das hat einen Traceback in irgendwelchen 
Plugins getriggert. Hab python3-qt4 wieder von meinen KDE4/Qt4 Repos 
installiert, 
aber vielleicht sollte Eric diese Situation irgendwann mal besser handhaben...

> > Erinnert mich daran, dass es toll wäre, wenn eric sich die
> > Splitterposition
> > merken würde!
> 
> Welche Splitter meinst du? Es werden verschiedene benutzt.

Die Splitterposition einer geteilten Ansicht.

Beim Testen der Ansichtsteilung, und anschließendem Beenden trat jetzt zweimal 
der angehängte Fehler auf.

> > BTW, die MicroPython-Unterstützung von eric finde ich supercool. Habe
> > schon
> > längst mal vor, damit ein bisschen zu spielen (vielleicht kann ich damit
> > meinem 14 jährigen Sohn die embedded-Welt und Python damit ein wenig 
näher
> > bringen).. Welche µPython boards benutzt/bevorzugst Du so?
> 
> Zum Entwickeln und Testen habe ich verschiedene Boards. Habe aber nicht
> wirklich groß etwas entwickelt. Am besten ausgestattet für den Einstieg
> scheint mir das Circuit Playground Express von Adafruit zu sein.

Schau ich mir mal an, Danke für die Info.

Grüße,
Hans-Peter

2019-12-05, 11:10:57

: 
'Editor' object has no attribute '_Editor__inSpellLanguageChanged'

  File "/usr/lib/python3.7/site-packages/eric6/QScintilla/Editor.py", line 5747, in __newViewNewSplit
self.vm.newEditorView(self.fileName, self, self.filetype)
  File "/usr/lib/python3.7/site-packages/eric6/ViewManager/ViewManager.py", line 4788, in newEditorView
editor, assembly = self.cloneEditor(caller, filetype, fn)
  File "/usr/lib/python3.7/site-packages/eric6/ViewManager/ViewManager.py", line 4809, in cloneEditor
tv=e5App().getObject("TaskViewer"))
  File "/usr/lib/python3.7/site-packages/eric6/QScintilla/EditorAssembly.py", line 46, in __init__
self.__editor = Editor(dbs, fn, vm, filetype, editor, tv)
  File &quo

Re: [Eric] [eric6] small glitch with fix and some notes

[Hans-Peter Jansen]

[Fabian: if you want to leave the loop, let me know]

Am Freitag, 29. November 2019, 20:22:22 CET schrieb Hans-Peter Jansen:
> Am Donnerstag, 28. November 2019, 20:12:25 CET schrieb Hans-Peter Jansen:
> > Am Donnerstag, 28. November 2019, 19:27:51 CET schrieb Detlev Offenbach:
> > 
> > Dear @audience, would you test these ill-behaviors and report, please?

I just reproduced this behavior with:

Leap 15.0:

Python 3.6.5, 64-Bit
Qt 5.9.4
PyQt 5.10.1
sip 4.19.7
QScintilla 2.10.4
WebEngine 61.0.3163.140
WebKit 602.1
eric6 19.04 (rev. abfed806d4ae)

Leap 15.1:

Python 3.6.5, 64-Bit
Qt 5.9.4
PyQt 5.10.1
sip 4.19.7
QScintilla 2.10.4
WebEngine 61.0.3163.140
WebKit 602.1
eric6 19.04 (rev. abfed806d4ae)

Leap 15.2 (Alpha Build 523.2 from 2019/11/18):

Python 3.6.5, 64-Bit
Qt 5.12.5
PyQt 5.13.1
sip 4.19.19
QScintilla 2.10.4
WebKit 602.1
eric6 19.04 (rev. abfed806d4ae)

Created two new files (shown as utf-8 text in status line), added some üzkül 
text, and switched tabs with shortcut Ctrl-Alt-TAB: cursor disappeared. Typing 
doesn't work until I click into the editor widget. (All tests done in 
Virtualbox 6.0.14, 15.1 and 15.2 are pristine ootb installations with just 
additional selection of eric and python3-Markdown packages, and all updates 
installed).

It starts to smell like a logical problem, somewhere between QScintilla and 
Eric. 

@Detlev: the openSUSE default setup for eric uses the builtin defaults, as 
provided from you¹. So it should behave similar for most current out of the 
box installations of everything Linux/X11. (Ubuntu makes a difference of 
course due to their policy to patch *everything*, that doesn't refuge in trees 
on count to three).

To be fair: openSUSE uses one QScintilla patch, that adapts to the Qt5 
installation, and Qt5 (as of Tumbleweed) contains a couple of (essential) 
patches. The one, that comes nearest this issue is 0001-QWidget-setFocusProxy-
adjust-focus-widget-properly.patch, that fixes the input field behavior of 
QWebEngine (as mentioned already).

https://build.opensuse.org/package/show/KDE:Qt5/libqt5-qtbase

Since the issue spans from 5.9 to 5.13, the arguments being a Qt issue 
dissolve.

Cheers,
Pete

¹) Reminds me, that I need to disable the check for updates by default.
   Would be nice, if you could point my to the bool in question.


___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] [eric6] small glitch with fix and some notes

[Hans-Peter Jansen]

Am Donnerstag, 28. November 2019, 20:12:25 CET schrieb Hans-Peter Jansen:
> Am Donnerstag, 28. November 2019, 19:27:51 CET schrieb Detlev Offenbach:
> 
> Dear @audience, would you test these ill-behaviors and report, please?
> 
> > I have the strong impression, that this is a Qt issue.
> 
> Yeah, smells like that. Unfortunately, without a minimal example, it's
> rather pointless to report such issues (even with code, it's a matter of
> luck to find somebody who cares).
> 
> Will start with a report to openSUSE Factory.

Okay, here's the current state of affairs:

There's a tab view focus issue in eric: when switching tabs with Ctrl-Alt{,-
Shift}-TAB, the focus (cursor) from editor widgets disappear. A click into the 
widget make the cursor reappear.

There's a focus related Qt 5.13.1 regression known so far:

 * https://bugreports.qt.io/browse/QTBUG-77364

Applying the fix doesn't help with this issue. BTW, Tumbleweeds Qt 5.13.1 
carries a patch for this already:

0001-QWidget-setFocusProxy-adjust-focus-widget-properly.patch

but only QWebEngine related issues are reported so far concerning this.

Reverting this, as well as reverting 

https://codereview.qt-project.org/c/qt/qtbase/+/268743

doesn't solve this issue.

Another test, using Qt 5.14 Beta 3 from

http://download.opensuse.org/repositories/KDE:/Medias/images/iso/?
P=*5.14.*.iso

shows the same behavior. Hence, it is to be expected, that any new Qt version 
will suffer from this issue.

Further investigation showed some churn in the src/widgets/kernel/qwidget.cpp 
module, related to focus and focus proxies during this year, so it's very 
likely to be a Qt regression.

@Detlev: have your tried to reproduce the issue already? 

I've tried some hacks on QScintilla/Editor.py, but no cure so far..

Sorry && Cheers,
Pete


___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] [eric6] small glitch with fix and some notes

[Hans-Peter Jansen]

Am Donnerstag, 28. November 2019, 19:27:51 CET schrieb Detlev Offenbach:
> Hi Pete,
> 
> thanks for the patch. I just added it. It will be part of the next release
> scheduled for next Sunday (Dec, 1st).

Nice.

> With regard to the tab switching shortcut, which PyQt/Qt version are you
> using? I noticed that some focus related issues appeared after I switched to
> PyQt/Qt 5.13.1. When I switched back to 5.13.0 they were gone again. I
> looked at the code (E5TabWidget.py) and the code to pass focus to the new
> 'current' widget is there.

BTW: there's another focus related issue: when editing some files, I often 
select parts with Ctrl-Shift cursor keys and cut/copy the selection (Ctrl-
{X,C}. I noticed, that it doesn't perform the cut/copy operation *sometimes*. 
The right click context menu shows most items disabled, then. Two Alt-TAB 
shortcuts later, all is set again. (But again, rather inconvenient).

Yes, I'm on Qt 5.13.1, but unlike documented in my last mail, using qscintilla 
2.11.3 meanwhile. 

Dear @audience, would you test these ill-behaviors and report, please?

> I have the strong impression, that this is a Qt issue.

Yeah, smells like that. Unfortunately, without a minimal example, it's rather 
pointless to report such issues (even with code, it's a matter of luck to find 
somebody who cares).

Will start with a report to openSUSE Factory.

Thanks,
Pete


___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

[Eric] [eric6] small glitch with fix and some notes

[Hans-Peter Jansen]

Hi Detlev,

the glitch below is easily fixed with:

--- /usr/lib/python3.7/site-packages/eric6/QScintilla/Editor.py~
2019-11-23 11:22:49.0 +0100
+++ /usr/lib/python3.7/site-packages/eric6/QScintilla/Editor.py 2019-11-28 
11:23:31.698303982 +0100
@@ -1697,7 +1697,7 @@ class Editor(QsciScintillaCompat):
 @type bool
 """
 self.__setSpellingLanguage(language)
-self.__checkSpellLanguage(language)
+self.__checkSpellLanguage()
 
 if not self.__inSpellLanguageChanged and propagate:
 self.__inSpellLanguageChanged = True


After some longer break, I'm back to work with Eric on a daily base. Now with
openSUSE Tumbleweed, where I maintain the eric package meanwhile..

While at it, I noticed an inconvenient behavior: when switching between tabs
with shortcuts (Ctrl-Alt-TAB, Ctrl-Alt-Shift-TAB), There's no active cursor
in this editor shown, but after a click into it only. That defeats the purpose 
of those shortcuts significantly. ;-)

Would be very nice, if you could look into this some time soon.

Thanks,
Pete



2019-11-27, 08:29:57

: 
__checkSpellLanguage() takes 1 positional argument but 2 were given

  File "/usr/lib/python3.7/site-packages/eric6/QScintilla/Editor.py", line 
5747, in __newViewNewSplit
self.vm.newEditorView(self.fileName, self, self.filetype)
  File "/usr/lib/python3.7/site-packages/eric6/ViewManager/ViewManager.py", 
line 4788, in newEditorView
editor, assembly = self.cloneEditor(caller, filetype, fn)
  File "/usr/lib/python3.7/site-packages/eric6/ViewManager/ViewManager.py", 
line 4809, in cloneEditor
tv=e5App().getObject("TaskViewer"))
  File "/usr/lib/python3.7/site-packages/eric6/QScintilla/EditorAssembly.py", 
line 46, in __init__
self.__editor = Editor(dbs, fn, vm, filetype, editor, tv)
  File "/usr/lib/python3.7/site-packages/eric6/QScintilla/Editor.py", line 493, 
in __init__
propagate=False)
  File "/usr/lib/python3.7/site-packages/eric6/QScintilla/Editor.py", line 
1700, in __spellLanguageChanged
self.__checkSpellLanguage(language)


Version Numbers:
  Python 3.7.3, 64-Bit
  Qt 5.13.1
  PyQt 5.13.1
  sip 4.19.19
  QScintilla 2.11.2
  WebEngine 73.0.3683.105
  eric6 19.11 (rev. 83df7d69e61b)

Platform: linux
3.7.3 (default, Apr 09 2019, 05:18:21) [GCC]

Desktop: KDE

Plugins Version Numbers:
  PluginAbout 19.11
  PluginApis 2.1.0
  PluginAssistantEric 4.1.3
  PluginCodeStyleChecker 19.11
  PluginColorString 2.2.3
  PluginCompletionJedi 3.5.0
  PluginCxFreeze 6.0.13
  PluginDocumentationSets 1.0.0
  PluginEricapi 19.11
  PluginEricdoc 19.11
  PluginInvertHexColor 2.1.3
  PluginMetricsRadon 2.1.0
  PluginMqttMonitor 1.1.2
  PluginPrintRemover 2.1.2
  PluginProjectDjango 5.2.3
  PluginProjectDjangoTagsMenu 2.0.4
  PluginProjectKivy 2.0.3
  PluginProjectPyramid 2.4.3
  PluginProjectWeb 1.1.4
  PluginPyInstaller 1.0.2
  PluginPyLint 6.1.11
  PluginPySide2PyQt 2.2.1
  PluginRefactoringRope 6.10.0
  PluginSelectionEncloser 2.1.4
  PluginSplitMergeCamelCase 2.1.1
  PluginSyntaxChecker 19.11
  PluginTabnanny 19.11
  PluginTimeTracker 2.0.10
  PluginToolGenerateHash 2.1.3
  PluginToolPip 1.6.3
  PluginTranslator 19.11
  PluginVcsGit 19.11
  PluginVcsMercurial 19.11
  PluginVcsPySvn 19.11
  PluginVcsSubversion 19.11
  PluginVmListspace 19.11
  PluginVmTabview 19.11
  PluginVulture 2.0.1
  PluginWizardDataUriEncoder 2.0.4
  PluginWizardDotDesktop 19.11
  PluginWizardE5MessageBox 19.11
  PluginWizardEricPlugin 19.11
  PluginWizardPyRegExp 19.11
  PluginWizardQColorDialog 19.11
  PluginWizardQFileDialog 19.11
  PluginWizardQFontDialog 19.11
  PluginWizardQInputDialog 19.11
  PluginWizardQMessageBox 19.11
  PluginWizardQRegExp 19.11
  PluginWizardQRegularExpression 19.11
  PluginWizardSetup 19.11
  PluginWizardVirtualenv 2.2.7

Distribution Info:
  /etc/os-release
  NAME="openSUSE Tumbleweed"
  # VERSION="20191124"
  ID="opensuse-tumbleweed"
  ID_LIKE="opensuse suse"
  VERSION_ID="20191124"
  PRETTY_NAME="openSUSE Tumbleweed"
  ANSI_COLOR="0;32"
  CPE_NAME="cpe:/o:opensuse:tumbleweed:20191124"
  BUG_REPORT_URL="https://bugs.opensuse.org;
  HOME_URL="https://www.opensuse.org/;
  LOGO="distributor-logo"


___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] [PyQt] ANN: eric 18.05 released

[Hans-Peter Jansen]

On Mittwoch, 2. Mai 2018 18:40:42 Detlev Offenbach wrote:
> Hi,
> 
> I just committed a shorter fix for the issue. Please give it a try.

which works like a charm, thanks Detlev.

Cheers,
Pete
 
> Detlev
> 
> Am Dienstag, 1. Mai 2018, 22:33:29 CEST schrieb Hans-Peter Jansen:
> > On Dienstag, 1. Mai 2018 21:42:38 Hans-Peter Jansen wrote:
> > > Hi Detlev,
> > > 
> > > On Dienstag, 1. Mai 2018 09:38:51 Rex Dieter wrote:
> > > > Detlev Offenbach wrote:
> > > > > I just uploaded eric 18.05. It fixes some bugs and adds these new
> > > > > features.
> > > > 
> > > > One buglet, wrt to packaging, I use
> > > > 
> > > > install.py \
> > > > 
> > > >   -i %{buildroot}/ \
> > > >   -b %{_bindir}
> > > > 
> > > > when building an eric rpm in fedora, but the installed
> > > > /usr/share/applications/eric6.desktop
> > > > ends up including the %{buildroot} macro contents in Exec=, like:
> > > > 
> > > > Exec=/builddir/build/BUILDROOT/eric-18.05-1.fc29.noarch/usr/bin/eric6
> > > 
> > > Because I really want executables defined with a path in the desktop
> > > files,
> > 
> > That way, you really get, what you asked for, and not some stray
> > installation from the home dir, for example...
> > 
> > > here's my suggested fix:
> > and here's a fixed and _tested_ version (hrmpf):
> > 
> > Index: b/install.py
> > ===
> > --- a/install.py
> > +++ b/install.py
> > 
> > @@ -302,7 +302,7 @@ def copyDesktopFile(src, dst, marker):
> >  @param dst destination file name (string)
> >  @param marker marker to be used (string)
> >  """
> > 
> > -global cfg
> > +global cfg, distDir
> > 
> >  if sys.version_info[0] == 2:
> >  f = codecs.open(src, "r", "utf-8")
> > 
> > @@ -310,8 +310,12 @@ def copyDesktopFile(src, dst, marker):
> >  f = open(src, "r", encoding="utf-8")
> >  
> >  text = f.read()
> >  f.close()
> > 
> > -
> > -text = text.replace("@BINDIR@", cfg['bindir'])
> > +
> > +bindir = cfg['bindir']
> > +if distDir:
> > +bindir =  bindir.replace(distDir, '')
> > +
> > +text = text.replace("@BINDIR@", bindir)
> > 
> >  text = text.replace("@MARKER@", marker)
> >  
> >  if marker:
> >  t_marker = " ({0})".format(PythonTextMarkers[marker])
> > 
> > Index: b/eric/eric6_browser.desktop
> > ===
> > --- a/eric/eric6_browser.desktop
> > +++ b/eric/eric6_browser.desktop
> > @@ -1,7 +1,7 @@
> > 
> >  [Desktop Entry]
> >  Version=1.0
> >  Type=Application
> > 
> > -Exec=eric6_browser@MARKER@
> > +Exec=@BINDIR@/eric6_browser@MARKER@
> > 
> >  MimeType=text/html;text/xml;application/xhtml+xml;x-scheme-handler/http;x
> >  -s
> > 
> > cheme-handler/https;application/x-mimearchive; Icon=ericWeb@MARKER@
> > 
> >  Terminal=false
> > 
> > Index: b/eric/eric6_webbrowser.desktop
> > ===
> > --- a/eric/eric6_webbrowser.desktop
> > +++ b/eric/eric6_webbrowser.desktop
> > @@ -1,7 +1,7 @@
> > 
> >  [Desktop Entry]
> >  Version=1.0
> >  Type=Application
> > 
> > -Exec=eric6_webbrowser@MARKER@
> > +Exec=@BINDIR@/eric6_webbrowser@MARKER@
> > 
> >  MimeType=text/html;text/xml;application/xhtml+xml;x-scheme-handler/http;x
> >  -s
> > 
> > cheme-handler/https; Icon=ericWeb@MARKER@
> > 
> >  Terminal=false
> > 
> > You can find my openSUSE builds here, as usual:
> > https://build.opensuse.org/project/monitor/home:frispete:PyQt5
> > 
> > 
> > Cheers,
> > Pete

___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] [PyQt] ANN: eric 18.05 released

[Hans-Peter Jansen]

On Dienstag, 1. Mai 2018 21:42:38 Hans-Peter Jansen wrote:
> Hi Detlev,
> 
> On Dienstag, 1. Mai 2018 09:38:51 Rex Dieter wrote:
> > Detlev Offenbach wrote:
> > > I just uploaded eric 18.05. It fixes some bugs and adds these new
> > > features.
> > 
> > One buglet, wrt to packaging, I use
> > 
> > install.py \
> > 
> >   -i %{buildroot}/ \
> >   -b %{_bindir}
> > 
> > when building an eric rpm in fedora, but the installed
> > /usr/share/applications/eric6.desktop
> > ends up including the %{buildroot} macro contents in Exec=, like:
> > 
> > Exec=/builddir/build/BUILDROOT/eric-18.05-1.fc29.noarch/usr/bin/eric6
> 
> Because I really want executables defined with a path in the desktop files,

That way, you really get, what you asked for, and not some stray installation 
from the home dir, for example...

> here's my suggested fix:

and here's a fixed and _tested_ version (hrmpf):

Index: b/install.py
===
--- a/install.py
+++ b/install.py
@@ -302,7 +302,7 @@ def copyDesktopFile(src, dst, marker):
 @param dst destination file name (string)
 @param marker marker to be used (string)
 """
-global cfg
+global cfg, distDir
 
 if sys.version_info[0] == 2:
 f = codecs.open(src, "r", "utf-8")
@@ -310,8 +310,12 @@ def copyDesktopFile(src, dst, marker):
 f = open(src, "r", encoding="utf-8")
 text = f.read()
 f.close()
-
-text = text.replace("@BINDIR@", cfg['bindir'])
+   
+bindir = cfg['bindir']
+if distDir:
+bindir =  bindir.replace(distDir, '')
+
+text = text.replace("@BINDIR@", bindir)
 text = text.replace("@MARKER@", marker)
 if marker:
 t_marker = " ({0})".format(PythonTextMarkers[marker])
Index: b/eric/eric6_browser.desktop
===
--- a/eric/eric6_browser.desktop
+++ b/eric/eric6_browser.desktop
@@ -1,7 +1,7 @@
 [Desktop Entry]
 Version=1.0
 Type=Application
-Exec=eric6_browser@MARKER@
+Exec=@BINDIR@/eric6_browser@MARKER@
 
MimeType=text/html;text/xml;application/xhtml+xml;x-scheme-handler/http;x-scheme-handler/https;application/x-mimearchive;
 Icon=ericWeb@MARKER@
 Terminal=false
Index: b/eric/eric6_webbrowser.desktop
===
--- a/eric/eric6_webbrowser.desktop
+++ b/eric/eric6_webbrowser.desktop
@@ -1,7 +1,7 @@
 [Desktop Entry]
 Version=1.0
 Type=Application
-Exec=eric6_webbrowser@MARKER@
+Exec=@BINDIR@/eric6_webbrowser@MARKER@
 
MimeType=text/html;text/xml;application/xhtml+xml;x-scheme-handler/http;x-scheme-handler/https;
 Icon=ericWeb@MARKER@
 Terminal=false


You can find my openSUSE builds here, as usual:
https://build.opensuse.org/project/monitor/home:frispete:PyQt5


Cheers,
PeteIndex: b/install.py
===
--- a/install.py
+++ b/install.py
@@ -302,7 +302,7 @@ def copyDesktopFile(src, dst, marker):
 @param dst destination file name (string)
 @param marker marker to be used (string)
 """
-global cfg
+global cfg, distDir
 
 if sys.version_info[0] == 2:
 f = codecs.open(src, "r", "utf-8")
@@ -310,8 +310,12 @@ def copyDesktopFile(src, dst, marker):
 f = open(src, "r", encoding="utf-8")
 text = f.read()
 f.close()
-
-text = text.replace("@BINDIR@", cfg['bindir'])
+   
+bindir = cfg['bindir']
+if distDir:
+bindir =  bindir.replace(distDir, '')
+
+text = text.replace("@BINDIR@", bindir)
 text = text.replace("@MARKER@", marker)
 if marker:
 t_marker = " ({0})".format(PythonTextMarkers[marker])
Index: b/eric/eric6_browser.desktop
===
--- a/eric/eric6_browser.desktop
+++ b/eric/eric6_browser.desktop
@@ -1,7 +1,7 @@
 [Desktop Entry]
 Version=1.0
 Type=Application
-Exec=eric6_browser@MARKER@
+Exec=@BINDIR@/eric6_browser@MARKER@
 MimeType=text/html;text/xml;application/xhtml+xml;x-scheme-handler/http;x-scheme-handler/https;application/x-mimearchive;
 Icon=ericWeb@MARKER@
 Terminal=false
Index: b/eric/eric6_webbrowser.desktop
===
--- a/eric/eric6_webbrowser.desktop
+++ b/eric/eric6_webbrowser.desktop
@@ -1,7 +1,7 @@
 [Desktop Entry]
 Version=1.0
 Type=Application
-Exec=eric6_webbrowser@MARKER@
+Exec=@BINDIR@/eric6_webbrowser@MARKER@
 MimeType=text/html;text/xml;application/xhtml+xml;x-scheme-handler/http;x-scheme-handler/https;
 Icon=ericWeb@MARKER@
 Terminal=false
___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] [PyQt] ANN: eric 18.05 released

[Hans-Peter Jansen]

Hi Detlev,

On Dienstag, 1. Mai 2018 09:38:51 Rex Dieter wrote:
> Detlev Offenbach wrote:
> > I just uploaded eric 18.05. It fixes some bugs and adds these new
> > features.
> 
> One buglet, wrt to packaging, I use
> 
> install.py \
>   -i %{buildroot}/ \
>   -b %{_bindir}
> 
> when building an eric rpm in fedora, but the installed
> /usr/share/applications/eric6.desktop
> ends up including the %{buildroot} macro contents in Exec=, like:
> 
> Exec=/builddir/build/BUILDROOT/eric-18.05-1.fc29.noarch/usr/bin/eric6

Because I really want executables defined with a path in the desktop files,
here's my suggested fix:

Index: b/install.py
===
--- a/install.py
+++ b/install.py
@@ -302,7 +302,7 @@ def copyDesktopFile(src, dst, marker):
 @param dst destination file name (string)
 @param marker marker to be used (string)
 """
-global cfg
+global cfg, distDir
 
 if sys.version_info[0] == 2:
 f = codecs.open(src, "r", "utf-8")
@@ -310,8 +310,12 @@ def copyDesktopFile(src, dst, marker):
 f = open(src, "r", encoding="utf-8")
 text = f.read()
 f.close()
-
-text = text.replace("@BINDIR@", cfg['bindir'])
+   
+bindir = cfg['bindir']
+if distDir:
+bindir =  bindir.replace(distDir + os.sep, '')
+
+text = text.replace("@BINDIR@", bindir)
 text = text.replace("@MARKER@", marker)
 if marker:
 t_marker = " ({0})".format(PythonTextMarkers[marker])
Index: b/eric/eric6_browser.desktop
===
--- a/eric/eric6_browser.desktop
+++ b/eric/eric6_browser.desktop
@@ -1,7 +1,7 @@
 [Desktop Entry]
 Version=1.0
 Type=Application
-Exec=eric6_browser@MARKER@
+Exec=@BINDIR@/eric6_browser@MARKER@
 
MimeType=text/html;text/xml;application/xhtml+xml;x-scheme-handler/http;x-scheme-handler/https;application/x-mimearchive;
 Icon=ericWeb@MARKER@
 Terminal=false
Index: b/eric/eric6_webbrowser.desktop
===
--- a/eric/eric6_webbrowser.desktop
+++ b/eric/eric6_webbrowser.desktop
@@ -1,7 +1,7 @@
 [Desktop Entry]
 Version=1.0
 Type=Application
-Exec=eric6_webbrowser@MARKER@
+Exec=@BINDIR@/eric6_webbrowser@MARKER@
 
MimeType=text/html;text/xml;application/xhtml+xml;x-scheme-handler/http;x-scheme-handler/https;
 Icon=ericWeb@MARKER@
 Terminal=false


Cheers,
PeteIndex: b/install.py
===
--- a/install.py
+++ b/install.py
@@ -302,7 +302,7 @@ def copyDesktopFile(src, dst, marker):
 @param dst destination file name (string)
 @param marker marker to be used (string)
 """
-global cfg
+global cfg, distDir
 
 if sys.version_info[0] == 2:
 f = codecs.open(src, "r", "utf-8")
@@ -310,8 +310,12 @@ def copyDesktopFile(src, dst, marker):
 f = open(src, "r", encoding="utf-8")
 text = f.read()
 f.close()
-
-text = text.replace("@BINDIR@", cfg['bindir'])
+   
+bindir = cfg['bindir']
+if distDir:
+bindir =  bindir.replace(distDir + os.sep, '')
+
+text = text.replace("@BINDIR@", bindir)
 text = text.replace("@MARKER@", marker)
 if marker:
 t_marker = " ({0})".format(PythonTextMarkers[marker])
Index: b/eric/eric6_browser.desktop
===
--- a/eric/eric6_browser.desktop
+++ b/eric/eric6_browser.desktop
@@ -1,7 +1,7 @@
 [Desktop Entry]
 Version=1.0
 Type=Application
-Exec=eric6_browser@MARKER@
+Exec=@BINDIR@/eric6_browser@MARKER@
 MimeType=text/html;text/xml;application/xhtml+xml;x-scheme-handler/http;x-scheme-handler/https;application/x-mimearchive;
 Icon=ericWeb@MARKER@
 Terminal=false
Index: b/eric/eric6_webbrowser.desktop
===
--- a/eric/eric6_webbrowser.desktop
+++ b/eric/eric6_webbrowser.desktop
@@ -1,7 +1,7 @@
 [Desktop Entry]
 Version=1.0
 Type=Application
-Exec=eric6_webbrowser@MARKER@
+Exec=@BINDIR@/eric6_webbrowser@MARKER@
 MimeType=text/html;text/xml;application/xhtml+xml;x-scheme-handler/http;x-scheme-handler/https;
 Icon=ericWeb@MARKER@
 Terminal=false
___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] Promote plug-ins to 'core' plug-ins

[Hans-Peter Jansen]

Dear Detlev,

On Mittwoch, 6. Dezember 2017 18:39:38 Detlev Offenbach wrote:
> Hi,
> 
> the setup.py wizard creates the text for a setup.py file. It does not create
> a skeleton project. What should be part of such a skeleton? Maybe I could
> extend the creation of a project (Project->New...).

May I bring the most advanced skeleton generator to your attention: 
cookiecutter (https://github.com/audreyr/cookiecutter).

I could imagine, that integrating cookiecutter, including a UI for querying 
available templates would be another great unique "selling point" of eric.

Pete

___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] Promote plug-ins to 'core' plug-ins

[Hans-Peter Jansen]

On Montag, 4. Dezember 2017 19:57:43 Detlev Offenbach wrote:
> Hello,
> 
> I am thinking about promoting some plug-ins available via the plug-in
> repository dialog to 'core' plug-ins. These would be a permanent part of the
> eric distribution. No separate download would be required. This is the list
 
That would be my choice:

> - translator


Cheers,
Pete
___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] Feature request: run/debug dialog enhancement

[Hans-Peter Jansen]

Hi Detlev,

guess, what Mikhail wants is a possibility to quickly switch between debuggers 
where a user expects it, e.g. a combo in the Debug Project/Script dialog to 
choose the Python version for the debugger. 

Would spare one "change settings" turnaround.

If you switch debugger versions frequently, this combo would make the latest 
setting obvious. 

Cheers,
Pete

On Dienstag, 12. September 2017 06:02:18 Detlev Offenbach wrote:
> Hello Mikhail,
> 
> if your projects are utilizing different Python interpreters you can define
> them on a per project basis. This is done via Project->Debugger->Debugger
> Properties. Via the configuration dialog (Project->Project page, Debugger
> section) the loading and saving of these can be configured.
> 
> I hope, this is what you were looking for. If there is additional demand,
> let me know.
> 
> Detlev
> 
> Am Montag, 11. September 2017, 21:39:22 CEST schrieb Mikhail Terekhov:
> > Detlev,
> > Would it be possible to add an additional combobox to the Run and Debug
> > dialogs
> > where user could select python interpreter? When working on several
> > projects that
> > have different dependencies it is very convenient to use virtual
> > environments. It is
> > also handy when testing/debugging different versions of some long
> > supported
> > project. Changing interpreter through settings dialog quickly became
> > tedious.
> > 
> > Regards,
> > Mikhail

___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] Seg Fault On Opening Dialog Box

[Hans-Peter Jansen]

Hi Ben,

On Freitag, 19. Mai 2017 16:27:08 Helen Ready wrote:
> Let me start by saying Thank You for the amazing IDE, Eric, and its
> community.  

Welcome to the club.

> Recently, I decided to take the plunge into Linux with CentOS 7
> as my first. After installing Python 3.6.1 from source and setting up a
> virtual enviroment, I naturally went to install the Eric IDE as my next
> step.

Installing Python from source need some careful considerations about
 * what _other_ versions of Python are installed in the system?
 * which Python(s) are qscintilla, sip, PyQt build on..
because some parts of it (Qt, qscintilla) cannot be fenced in the VE, they're 
pure C++...

OTOH; your traceback, although missing symbols, points to some interference 
between Qt and Gtk (it looks, it crashes, where the native Gtk file dialog is 
bent into Qt...).

Please try to execute python examples/dialogs/standarddialogs.py (from PyQt5), 
and check, if the QFileDialog functions behave.

Pete
___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] eric debug mode appearance

[Hans-Peter Jansen]

Hi Tobias, hi Detlev,

On Dienstag, 18. April 2017 20:56:25 Tobias Rzepka wrote:
> Hello Pete,
> 
> the first two wishes are possible. Maybe it'll be the easiest, when there
> will be an extra window, showing the actual code line(s) and the
> disassembled bytecode, like the outputs of dis.dis / dis.distb.

The idea of integration with the editor is two-fold: 
 * make the dis.dis() output easily available during development, where
   the 2 window solution would make the feature more expensive to use
   (scrolling/locating the python lines in the dis.dis() output)
 * similar, make the dis.distb() feature more approachable/obvious in the
   exception handling case

> But stepping through single bytecodes isn't possible because the debug
> events (call, line, exception, ...) are only fired on new lines (see
> ceval.c -> maybe_call_line_trace). It would require to change the Python VM
> itself :-(

Yes, that was my guts feeling already ;)

To my humble knowledge, the two possible features form a unique selling point 
already.

Cheers,
Pete
___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] eric debug mode appearance

[Hans-Peter Jansen]

On Montag, 17. April 2017 18:42:20 Detlev Offenbach wrote:
> Hello Pete,
> 
> the changed indication will be part of the next release. If you need this
> urgently, please get the default branch of the eric sources.

Great, thank you.

> Detlev
> 
> PS: Please give a spec for the bytecode disassembler display.

Started from a toggle option (e.g. Show Bytecode), python source code lines 
are amended with the disassembly below (unnumbered). 

In debug mode, when catching an exception and bytecode display is enabled, the 
offending opcode should be emphasized in the bytecode disassembly 
(dis.distb()). 

See Python build-in dis module and e.g. https://pymotw.com/3/dis/. 

In a future version, single stepping through the bytecode would be awesome 
(but I'm not sure, if interfaces for such functionality are available 
already).

Cheers,
Pete

___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

[Eric] Eric pretend, it's crashed, if executed subsequently

[Hans-Peter Jansen]

Hi Detlev,

if executing another eric process, it pretends, that the session data found is 
from crashed processes, although it is just running already.

It would be nice to distinguish this case from crashed sessions (at least on 
the same machine).

Cheers,
Pete
___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] Wish: Debug option to move program counter

[Hans-Peter Jansen]

Hi Tobias, hi Detlev,

On Dienstag, 14. März 2017 19:14:50 Tobias Rzepka wrote:
> Hello Pete, hello Detlev,
> 
> I've done already a proof of concept for the debugger. There are some
> constraints given by Python but it works to change the instruction pointer
> in a function/ method without executing the code between the actual and new
> instruction pointer.
> 
> I think, it could be very helpful a lot of times. I'll work on it further as
> one of the next topics Tobias

Thank you very much for this feature. 

It helps me to test conditions, that usually not happen, as well as repeat 
some code segments with different values in minutes. Hey students, you 
hopefully write test code to get 100% code coverage, do you?

This reminded me about the very beginnings of my hacking career, where I 
programmed 680x0 assembler on the Atari ST. Thanks to a great debugger called 
MonST, I really was able to grasp the inner workings of these CPUs.

Later, I worked for a company, that supplied Ethernet network hardware and 
software for those Ataris, which I also installed in business environments. 
At that time, I always installed a reset resident version of this debugger on 
all systems. Since Atari programs weren't created with networks in mind and 
some aspects of network file systems couldn't be hidden from them, a few rare  
network operations resulted in crashes, where MonST chimed in, and - guess 
what - I often was able to recover such a crashed machine be changing the 
instruction counter and/or a few register values.. Imagine the face 
expressions of users, that watched me doing that for the first time.

Using Python, debugging is a little less important, but occasionally, having 
sophisticated debugging purposes at hand, is most effective. 

Eric is my favorite Python development environment since a long time, and I 
enjoy it nearly every day. Although my contributions to the Eric code base is 
near zero unfortunately (sorry Detlev), I advertise it on a regular base.

Thank you, Detlev, you're doing a great job!

Tobias, thank you for supporting Eric (and Detlev). It really deserves it!

Pete
___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] Wish: Debug option to move program counter

[Hans-Peter Jansen]

Hi Tobias,

I've missed your message somehow.

On Dienstag, 14. März 2017 19:14:50 Tobias Rzepka wrote:
> Hello Pete, hello Detlev,
> 
> I've done already a proof of concept for the debugger. There are some
> constraints given by Python but it works to change the instruction pointer
> in a function/ method without executing the code between the actual and new
> instruction pointer.

This is good news. 

Technically, the current execution frame must match and the move is line 
based.

> I think, it could be very helpful a lot of times. I'll work on it further as
> one of the next topics Tobias

Great, let us know from your advances.

This might look like a minor feature, but it's a real killer feature in some 
circumstances and raises Eric's coolness factor significantly.

Pete
___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] I'm not able to install eric on linux Mint-KDE (bugs.launchpad.net/linuxmint/+bug/1656433)

[Hans-Peter Jansen]

On Samstag, 4. März 2017 17:34:41 GM wrote:
> a workaround was found by "GHPS":
> https://bugs.launchpad.net/linuxmint/+bug/1656433/comments/5
> 
> after upgrading libqt5scintilla2-l10n and libqt5scintilla2-12v5 manually
> to version 2.9.3
> 
> (
> https://www.ubuntuupdates.org/package/core/yakkety/universe/base/libqt5scint
> illa2-12v5
>  ntilla2-12v5>
> https://www.ubuntuupdates.org/package/core/yakkety/universe/base/libqt5scin
> tilla2-l10n
>  ntilla2-l10n> )
> 
> the distro's version of 'eric' installed without problems. The latest
> release of 'eric' (17.03) installed without problems too.

Just for the record: these are packaging problems of the distro in question, 
hence nothing can be done from the Eric side of things. Obviously, the build 
of libqt5scintilla2 is the culprit here.

Eric dependents on matching versions of Qscintilla and PyQt at least. For 
packagers, problems arises from the *freedom* in this ecosystem, since any 
combination of Qt 4/5 with Python 2/3 is supported as a base, but sip, 
qscintilla and PyQt builds *must* match the basic combination.

Pete
___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] Wish: Debug option to move program counter

[Hans-Peter Jansen]

On Dienstag, 14. März 2017 11:43:18 Detlev Offenbach wrote:
> Hi,
> 
> move the cursor to the line you want to stop next and select "Continue to
> cursor" (Shift+F6). Is that what you are looking for?

No, I really mean changing the execution path. 

Have a look, how pdb does it here:

https://svn.python.org/projects/python/trunk/Lib/pdb.py

Search for do_jump implementation.

In that form, it can modify the line to execute next in the current frame 
only, but this is very helpful to skip or repeat some lines in a debug 
session.

Hth,
Pete

> Detlev
> 
> Am Montag, 13. März 2017, 23:22:42 CET schrieb Hans-Peter Jansen:
> > Hi Detlev,
> > 
> > me again :wink:
> > 
> > During a debug session today, the wish to modify the debug program counter
> > arose (similar to pdbs j(ump) command).
> > 
> > Do you think, this is something useful for Eric, or is it available
> > already, and I just haven't figured out, how?
> > 
> > I attempted to add Eric to:
> > https://wiki.python.org/moin/PythonDebuggingTools
> > but the wiki don't let me authenticate at the moment.
> > 
> > If somebody in the audience already has a functioning python wiki account,
> > please add Eric there...
> > 
> > Cheers,
> > Pete
> > ___
> > Eric mailing list
> > Eric@riverbankcomputing.com
> > https://www.riverbankcomputing.com/mailman/listinfo/eric

___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] getpass in Eric console

[Hans-Peter Jansen]

On Dienstag, 14. März 2017 15:12:45 Detlev Offenbach wrote:
> Hi Pete,
> 
> that was a good one. It was actually two issues in one. The fix will be in
> the next release. If it is needed urgently please get it from the source
> code repository.

Glad, you found it. 

I was able to work around and solve the issue, that I was hunting while 
stumbling over this...

Happily awaiting your next release.

Thank you,
Pete

> Detlev
> 
> Am Dienstag, 14. März 2017, 14:01:24 CET schrieb Hans-Peter Jansen:
> > Hi Detlev,
> > 
> > using Eric 17.03, something strange happens, when attempting to debug
> > scripts that use getpass.
> > 
> > Here's, what happened, when calling it manually (in console).
> > 
> > I've entered '12345' all three times:
> > >>> import getpass
> > >>> getpass.getpass("PW:")
> > 
> > PW:
> > '45'
> > 
> > >>> getpass.getpass("PW:")
> > 
> > PW:
> > '4512345'
> > 
> > >>> getpass.getpass("PW:")
> > 
> > PW:
> > '451234512345'
> > 
> > 
> > Cheers,
> > Pete
> > ___
> > Eric mailing list
> > Eric@riverbankcomputing.com
> > https://www.riverbankcomputing.com/mailman/listinfo/eric

___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

[Eric] getpass in Eric console

[Hans-Peter Jansen]

Hi Detlev,

using Eric 17.03, something strange happens, when attempting to debug scripts 
that use getpass.

Here's, what happened, when calling it manually (in console). 
I've entered '12345' all three times:

>>> import getpass
>>> getpass.getpass("PW:")
PW:
'45'
>>> getpass.getpass("PW:")
PW:
'4512345'
>>> getpass.getpass("PW:")
PW:
'451234512345'
>>> 

Cheers,
Pete
___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

[Eric] Wish: Debug option to move program counter

[Hans-Peter Jansen]

Hi Detlev,

me again :wink:

During a debug session today, the wish to modify the debug program counter 
arose (similar to pdbs j(ump) command). 

Do you think, this is something useful for Eric, or is it available already, 
and I just haven't figured out, how?

I attempted to add Eric to:
https://wiki.python.org/moin/PythonDebuggingTools
but the wiki don't let me authenticate at the moment. 

If somebody in the audience already has a functioning python wiki account, 
please add Eric there...

Cheers,
Pete
___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

[Eric] Debug modules, that are supposed to be executed with -m

[Hans-Peter Jansen]

Hi Detlev,

I ran into a funny issue related to Python3. 

How can I debug a script in Eric, if this is supposed to be executed as a 
module (e.g. called with python3 -m ...), which happens, if one want to use 
relative paths and arbitrary entry points...

The option to rewrite everything with absolute paths isn't always feasible.

Cheers,
Pete
___
Eric mailing list
Eric@riverbankcomputing.com
https://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] eric6 with git plugin

[Hans-Peter Jansen]

On Mittwoch, 8. April 2015 19:33:12 Detlev Offenbach wrote:
 Hi Hans-Peter,
 
 first of all, thanks for the nice words.
 
 Is anything put out in the dialog (output or error panes)? Sometimes
 programs seem to redirect such questions directly to the terminal bypassing
 stdout and stderr. In such cases I don't know of any way to intercept the
 output with a QProcess. Maybe somebody can help.

The ssh message is partly shown in the password dialog, where ssh is waiting 
for a yes or no answer..

Cheers,
Pete
 
 Regards,
 Detlev
 
 On Wednesday 08 April 2015, 19:02:08 Hans-Peter Jansen wrote:
  Hi Detlev,
  
  eric6, together with git 1.1.1 plugin is simply great.
  I never had a snapier eric experience before.
  
  Congrats.
  
  Just one note: the git plugin doesn't handle the usual ssh question, when
  connecting the first time to a certain host. Here's a session log:
  
  $ pit push --verbose --porcelain origin master:master
  Versende nach g...@bla.bla:xxx.git
  The authenticity of host 'bla.bla (x.x.x.x)' can't be established.
  RSA key fingerprint is xx:xx:xx:xx:xx... [MD5].
  Are you sure you want to continue connecting (yes/no)? yes
  Warning: Permanently added 'bla.bla' (RSA) to the list of known hosts.
  Zähle Objekte: 15, Fertig.
  Delta compression using up to 4 threads.
  Komprimiere Objekte: 100% (15/15), Fertig.
  Schreibe Objekte: 100% (15/15), 26.24 KiB | 0 bytes/s, Fertig.
  Total 15 (delta 7), reused 0 (delta 0)
  To g...@bla.bla:xxx.git
  
  refs/heads/master:refs/heads/master bae4b23..d37138c
  
  updating local tracking ref 'refs/remotes/origin/master'
  Done
  
  Sure, rare situation, and easily fixed once on command line.
  
  Cheers,
  
Pete
  
  ___
  Eric mailing list
  Eric@riverbankcomputing.com
  http://www.riverbankcomputing.com/mailman/listinfo/eric-- 
 *Detlev Offenbach*
 det...@die-offenbachs.de

___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

[Eric] eric6 with git plugin

[Hans-Peter Jansen]

Hi Detlev,

eric6, together with git 1.1.1 plugin is simply great.
I never had a snapier eric experience before.

Congrats.

Just one note: the git plugin doesn't handle the usual ssh question, when 
connecting the first time to a certain host. Here's a session log:

$ pit push --verbose --porcelain origin master:master
Versende nach g...@bla.bla:xxx.git
The authenticity of host 'bla.bla (x.x.x.x)' can't be established.
RSA key fingerprint is xx:xx:xx:xx:xx... [MD5].
Are you sure you want to continue connecting (yes/no)? yes
Warning: Permanently added 'bla.bla' (RSA) to the list of known hosts.
Zähle Objekte: 15, Fertig.
Delta compression using up to 4 threads.
Komprimiere Objekte: 100% (15/15), Fertig.
Schreibe Objekte: 100% (15/15), 26.24 KiB | 0 bytes/s, Fertig.
Total 15 (delta 7), reused 0 (delta 0)
To g...@bla.bla:xxx.git
refs/heads/master:refs/heads/master bae4b23..d37138c
updating local tracking ref 'refs/remotes/origin/master'
Done

Sure, rare situation, and easily fixed once on command line.

Cheers,
  Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

[Eric] New logo

[Hans-Peter Jansen]

Hi,

to whom it may concern, the new logo is: GREAT!

Detlev, you might remember, that I started the initiative of replacing the 
ugly gnome with the face of Eric Idle, because that's the joke, Phil was after 
naming it Eric vs. the Python delivered IDE: IDLE.. Be aware of your roots 
;)

Phil, correct me, if I'm wrong!

Anyway, the new logo is just nice.

Cheers,
  Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] [PyQt] Eric4 issue with splitted views

[Hans-Peter Jansen]

[Reposted on Detelv's request in order to address also the qscintilla ML ---
Note to Phil: I really think, that sip.setdestroyonexit is the right approach 
to fix this issue (class), the question is, why it doesn't..]

Hi Detlev,

you're right - splitting is a red herring in this case, just choosing a New 
view is sufficient to trigger the exact same traceback, so QScintilla is  
most likely to be in charge of is crash.

Thanks for the hint, Detlev.

Kind regards,
Pete

On Dienstag, 25. März 2014 18:22:20 Detlev Offenbach wrote:
 Hi Hans-Peter,
 
 Does this only happen when you SPLIT the view or when you create a new view
 without a new split as well? If so it might be a QScintilla issue because
 when a new view is created both views are accessing the *same* document,
 i.e. the very same instance.
 
 Regards,
 Detlev
 
 On Monday 24 March 2014, 15:01:34 Hans-Peter Jansen wrote:
  Dear Phil,
  
  I suffer from a long term Eric 4 problem, where eric crashes consistently
  on exit, when I split just one view.
  
  Prerequisites:
  Running on KDE4 with PyQt4 available
  Using tabbed view in Viewmanager
  
  Reproduce:
  Just running eric with a single file argument, right click into editor,
  choose New view (with new split) and exit reproduces this for me.
  
  Results in:
  Thread 1 (Thread 0x7fc162564700 (LWP 4712)):
  [KCrash Handler]
  #6  0x7fc15d707aca in QWidgetPrivate::deleteTLSysExtra
  (this=this@entry=0x44e45f0) at kernel/qwidget_x11.cpp:2885 #7
  0x7fc15d6bea85 in QWidgetPrivate::deleteExtra
  (this=this@entry=0x44e45f0) at kernel/qwidget.cpp:1832 #8
  0x7fc15d6becad in QWidgetPrivate::~QWidgetPrivate (this=0x44e45f0,
  __in_chrg=optimized out) at kernel/qwidget.cpp:365 #9 
  0x7fc15daa8e75
  in ~QMenuPrivate (this=0x44e45f0, __in_chrg=optimized out) at
  widgets/qmenu_p.h:171 #10 QMenuPrivate::~QMenuPrivate (this=0x44e45f0,
  __in_chrg=optimized out) at widgets/qmenu_p.h:183 #11 0x7fc15fc68c35
  in cleanup (pointer=optimized out) at
  ../../src/corelib/tools/qscopedpointer.h:62 #12 ~QScopedPointer
  (this=0x44e4588, __in_chrg=optimized out) at
  ../../src/corelib/tools/qscopedpointer.h:100 #13 QObject::~QObject
  (this=0x44e4580, __in_chrg=optimized out) at kernel/qobject.cpp:816 #14
  0x7fc15d6c76ac in QWidget::~QWidget (this=0x44e4580,
  __in_chrg=optimized out) at kernel/qwidget.cpp:1554 #15
  0x7fc15e5bd989 in sipQMenu::~sipQMenu (this=0x44e4580,
  __in_chrg=optimized out) at sipQtGuiQMenu.cpp:316 #16 0x7fc15e5bd9fe
  in release_QMenu (sipCppV=0x44e4580) at sipQtGuiQMenu.cpp:2785 #17
  0x7fc1604ba476 in forgetObject (sw=sw@entry=0x422c2b0) at
  siplib.c:10891 #18 0x7fc1604bb599 in sipWrapper_dealloc
  (self=0x422c2b0) at siplib.c:10443 #19 0x7fc16209aefa in
  subtype_dealloc (self=0x422c2b0) at Objects/typeobject.c:1030 #20
  0x7fc1620937e0 in dict_dealloc (mp=0x44f9d50) at
  Objects/dictobject.c:1010 #21 0x7fc1620937b9 in dict_dealloc
  (mp=0x3d30f20) at Objects/dictobject.c:1010 #22 0x7fc1604b86a7 in
  sipSimpleWrapper_clear (self=self@entry=0x3c11b00) at siplib.c:10016 #23
  0x7fc1604bb4c4 in sipWrapper_clear (self=0x3c11b00) at siplib.c:10402
  #24 0x7fc16207b232 in delete_garbage (old=0x7fc162370260
  generations+96, collectable=0x7fff271e2780) at Modules/gcmodule.c:820
  #25
  collect (generation=2) at Modules/gcmodule.c:984
  #26 0x7fc16207b710 in PyGC_Collect () at Modules/gcmodule.c:1440
  #27 0x7fc1620727b9 in Py_Finalize () at Python/pythonrun.c:449
  #28 0x7fc162072255 in Py_Exit (sts=0) at Python/pythonrun.c:1777
  #29 0x7fc1620723a4 in handle_system_exit () at Python/pythonrun.c:1151
  #30 0x7fc1620723d1 in PyErr_PrintEx
  (set_sys_last_vars=set_sys_last_vars@entry=1) at Python/pythonrun.c:1161
  #31 0x7fc162072691 in PyErr_Print () at Python/pythonrun.c:1064 #32
  0x7fc162072bf9 in PyRun_SimpleFileExFlags (fp=fp@entry=0x83e4e0,
  filename=optimized out, filename@entry=0x7fff271e4f42
  /usr/lib64/python2.7/site-packages/eric4/eric4.py,
  closeit=closeit@entry=1, flags=flags@entry=0x7fff271e29e0) at
  Python/pythonrun.c:952 #33 0x7fc16207312c in PyRun_AnyFileExFlags
  (fp=fp@entry=0x83e4e0, filename=filename@entry=0x7fff271e4f42
  /usr/lib64/python2.7/site-packages/eric4/eric4.py,
  closeit=closeit@entry=1, flags=flags@entry=0x7fff271e29e0) at
  Python/pythonrun.c:752 #34 0x7fc16207a81e in Py_Main (argc=optimized
  out, argv=0x7fff271e2b98) at Modules/main.c:640 #35 0x7fc161a31be5 in
  __libc_start_main () from /lib64/libc.so.6 #36 0x00400791 in
  _start
  ()
  
  
  As you can see, this crash is a typical case of dtor called after python
  exited. Of course, there isn't much, you can do about this. You even
  provide sip.setdestroyonexit() for such issues. But that's exactly my
  problem here.
  
  With this patch applied, shouldn't such an issue not happen anymore:
  
  --- eric4.py.orig   2014-01-07 07:46:19.0 +0100
  +++ eric4.py2014-03-24 10:07

[Eric] Eric4 issue with splitted views

[Hans-Peter Jansen]

Dear Phil,

I suffer from a long term Eric 4 problem, where eric crashes consistently on 
exit, when I split just one view. 

Prerequisites:
Running on KDE4 with PyQt4 available
Using tabbed view in Viewmanager

Reproduce:
Just running eric with a single file argument, right click into editor, choose 
New view (with new split) and exit reproduces this for me.

Results in:
Thread 1 (Thread 0x7fc162564700 (LWP 4712)):
[KCrash Handler]
#6  0x7fc15d707aca in QWidgetPrivate::deleteTLSysExtra 
(this=this@entry=0x44e45f0) at kernel/qwidget_x11.cpp:2885
#7  0x7fc15d6bea85 in QWidgetPrivate::deleteExtra 
(this=this@entry=0x44e45f0) at kernel/qwidget.cpp:1832
#8  0x7fc15d6becad in QWidgetPrivate::~QWidgetPrivate (this=0x44e45f0, 
__in_chrg=optimized out) at kernel/qwidget.cpp:365
#9  0x7fc15daa8e75 in ~QMenuPrivate (this=0x44e45f0, __in_chrg=optimized 
out) at widgets/qmenu_p.h:171
#10 QMenuPrivate::~QMenuPrivate (this=0x44e45f0, __in_chrg=optimized out) at 
widgets/qmenu_p.h:183
#11 0x7fc15fc68c35 in cleanup (pointer=optimized out) at 
../../src/corelib/tools/qscopedpointer.h:62
#12 ~QScopedPointer (this=0x44e4588, __in_chrg=optimized out) at 
../../src/corelib/tools/qscopedpointer.h:100
#13 QObject::~QObject (this=0x44e4580, __in_chrg=optimized out) at 
kernel/qobject.cpp:816
#14 0x7fc15d6c76ac in QWidget::~QWidget (this=0x44e4580, 
__in_chrg=optimized out) at kernel/qwidget.cpp:1554
#15 0x7fc15e5bd989 in sipQMenu::~sipQMenu (this=0x44e4580, 
__in_chrg=optimized out) at sipQtGuiQMenu.cpp:316
#16 0x7fc15e5bd9fe in release_QMenu (sipCppV=0x44e4580) at 
sipQtGuiQMenu.cpp:2785
#17 0x7fc1604ba476 in forgetObject (sw=sw@entry=0x422c2b0) at siplib.c:10891
#18 0x7fc1604bb599 in sipWrapper_dealloc (self=0x422c2b0) at siplib.c:10443
#19 0x7fc16209aefa in subtype_dealloc (self=0x422c2b0) at 
Objects/typeobject.c:1030
#20 0x7fc1620937e0 in dict_dealloc (mp=0x44f9d50) at 
Objects/dictobject.c:1010
#21 0x7fc1620937b9 in dict_dealloc (mp=0x3d30f20) at 
Objects/dictobject.c:1010
#22 0x7fc1604b86a7 in sipSimpleWrapper_clear (self=self@entry=0x3c11b00) at 
siplib.c:10016
#23 0x7fc1604bb4c4 in sipWrapper_clear (self=0x3c11b00) at siplib.c:10402
#24 0x7fc16207b232 in delete_garbage (old=0x7fc162370260 generations+96, 
collectable=0x7fff271e2780) at Modules/gcmodule.c:820
#25 collect (generation=2) at Modules/gcmodule.c:984
#26 0x7fc16207b710 in PyGC_Collect () at Modules/gcmodule.c:1440
#27 0x7fc1620727b9 in Py_Finalize () at Python/pythonrun.c:449
#28 0x7fc162072255 in Py_Exit (sts=0) at Python/pythonrun.c:1777
#29 0x7fc1620723a4 in handle_system_exit () at Python/pythonrun.c:1151
#30 0x7fc1620723d1 in PyErr_PrintEx 
(set_sys_last_vars=set_sys_last_vars@entry=1) at Python/pythonrun.c:1161
#31 0x7fc162072691 in PyErr_Print () at Python/pythonrun.c:1064
#32 0x7fc162072bf9 in PyRun_SimpleFileExFlags (fp=fp@entry=0x83e4e0, 
filename=optimized out, filename@entry=0x7fff271e4f42 
/usr/lib64/python2.7/site-packages/eric4/eric4.py, closeit=closeit@entry=1, 
flags=flags@entry=0x7fff271e29e0) at Python/pythonrun.c:952
#33 0x7fc16207312c in PyRun_AnyFileExFlags (fp=fp@entry=0x83e4e0, 
filename=filename@entry=0x7fff271e4f42 
/usr/lib64/python2.7/site-packages/eric4/eric4.py, closeit=closeit@entry=1, 
flags=flags@entry=0x7fff271e29e0) at Python/pythonrun.c:752
#34 0x7fc16207a81e in Py_Main (argc=optimized out, argv=0x7fff271e2b98) 
at Modules/main.c:640
#35 0x7fc161a31be5 in __libc_start_main () from /lib64/libc.so.6
#36 0x00400791 in _start ()


As you can see, this crash is a typical case of dtor called after python 
exited. Of course, there isn't much, you can do about this. You even provide
sip.setdestroyonexit() for such issues. But that's exactly my problem here.

With this patch applied, shouldn't such an issue not happen anymore:

--- eric4.py.orig   2014-01-07 07:46:19.0 +0100
+++ eric4.py2014-03-24 10:07:04.223139840 +0100
@@ -17,6 +17,7 @@ import traceback
 import cStringIO
 import time
 import logging
+import sip
 
 from PyQt4.QtCore import QTextCodec, SIGNAL, SLOT, qWarning, \
 QLibraryInfo, QTimer
@@ -162,7 +163,12 @@ def main():
 global args, mainWindow, splash, restartArgs
 
 sys.excepthook = excepthook
-
+
+try:
+sip.setdestroyonexit(True)
+except AttributeError:
+pass
+
 options = [\
 (--config=configDir, 
  use the given directory as the one containing the config files), 


I checked it also without exception handling, just in case, but problem
persists. Since I always use splitted views, and don't want to miss eric's 
KDE integration, I would like to see this fixed.

I've checked eric's source for suspect operations during the split, but 
couldn't spot any fishy operations. (ViewManager/ViewManager.py, 
Plugins/ViewManagerPlugins/Tabview/Tabview.py). 

Given the environmental conditions, I suspect some timing related issue, 
which might 

[Eric] eric4 and PyQt dtors

[Hans-Peter Jansen]

Hi Phil,

If I remember correctly, you talked about making a new termination scheme 
default, where you don't call any dtors from a certain point in the 
termination process. Is this already available as an option?

When leaving eric4 running for an extended period of time, I *always*
harvest belows seg fault on termination, with current sip/PyQt snapshots.

I guess, such an option would be easier then hunting and deleting the 
last forgotten menu item...

Cheers,
Pete


Application: eric4 (python2.7), signal: Segmentation fault
Using host libthread_db library /lib64/libthread_db.so.1.
[Current thread is 1 (Thread 0x7faa86595700 (LWP 5604))]

Thread 2 (Thread 0x7faa62a5b700 (LWP 9430)):
#0  0x7faa85b2a14f in __GI___poll (fds=optimized out, nfds=optimized 
out, timeout=optimized out) at ../sysdeps/unix/sysv/linux/poll.c:87
#1  0x7faa83991684 in g_main_context_poll (n_fds=1, fds=0x7faa5801d680, 
timeout=-1, context=0x7faa58002720, priority=optimized out) at gmain.c:3440
#2  g_main_context_iterate (context=context@entry=0x7faa58002720, 
block=block@entry=1, dispatch=dispatch@entry=1, self=optimized out) at 
gmain.c:3141
#3  0x7faa839917a4 in g_main_context_iteration (context=0x7faa58002720, 
may_block=1) at gmain.c:3207
#4  0x7faa84725166 in QEventDispatcherGlib::processEvents 
(this=0x7faa5801bee0, flags=...) at kernel/qeventdispatcher_glib.cpp:426
#5  0x7faa846f597f in QEventLoop::processEvents 
(this=this@entry=0x7faa62a5ae50, flags=...) at kernel/qeventloop.cpp:149
#6  0x7faa846f5c08 in QEventLoop::exec (this=0x7faa62a5ae50, flags=...) at 
kernel/qeventloop.cpp:204
#7  0x7faa845f80e0 in QThread::exec (this=optimized out) at 
thread/qthread.cpp:542
#8  0x7faa846d60ef in QInotifyFileSystemWatcherEngine::run (this=0x76477a0) 
at io/qfilesystemwatcher_inotify.cpp:256
#9  0x7faa845fb0bc in QThreadPrivate::start (arg=0x76477a0) at 
thread/qthread_unix.cpp:338
#10 0x7faa85dfae0e in start_thread (arg=0x7faa62a5b700) at 
pthread_create.c:305
#11 0x7faa85b322cd in clone () at 
../sysdeps/unix/sysv/linux/x86_64/clone.S:115

Thread 1 (Thread 0x7faa86595700 (LWP 5604)):
[KCrash Handler]
#6  0x7faa81fcf41a in QWidgetPrivate::deleteTLSysExtra (this=0x632e390) at 
kernel/qwidget_x11.cpp:2884
#7  0x7faa81f8711e in QWidgetPrivate::deleteExtra 
(this=this@entry=0x632e390) at kernel/qwidget.cpp:1834
#8  0x7faa81f8734d in QWidgetPrivate::~QWidgetPrivate (this=0x632e390, 
__in_chrg=optimized out) at kernel/qwidget.cpp:361
#9  0x7faa8237c7f9 in ~QMenuPrivate (this=0x632e390, __in_chrg=optimized 
out) at widgets/qmenu_p.h:171
#10 QMenuPrivate::~QMenuPrivate (this=0x632e390, __in_chrg=optimized out) at 
widgets/qmenu_p.h:183
#11 0x7faa8470d1cc in cleanup (pointer=optimized out) at 
../../src/corelib/tools/qscopedpointer.h:62
#12 ~QScopedPointer (this=0x632e328, __in_chrg=optimized out) at 
../../src/corelib/tools/qscopedpointer.h:100
#13 QObject::~QObject (this=0x632e320, __in_chrg=optimized out) at 
kernel/qobject.cpp:816
#14 0x7faa81f89440 in QWidget::~QWidget (this=0x632e320, 
__in_chrg=optimized out) at kernel/qwidget.cpp:1556
#15 0x7faa82e8c189 in sipQMenu::~sipQMenu (this=0x632e320, 
__in_chrg=optimized out) at sipQtGuiQMenu.cpp:316
#16 0x7faa82e8aeb6 in release_QMenu (sipCppV=0x632e320) at 
sipQtGuiQMenu.cpp:2785
#17 0x7faa834d5bee in forgetObject (sw=sw@entry=0x620) at siplib.c:10427
#18 0x7faa834d7549 in sipWrapper_dealloc (self=0x620) at siplib.c:9979
#19 0x7faa860b8c76 in subtype_dealloc (self=0x620) at 
Objects/typeobject.c:1014
#20 0x7faa860b2a78 in dict_dealloc (mp=0x631e2d0) at 
Objects/dictobject.c:985
#21 0x7faa860b2a50 in dict_dealloc (mp=0x7faa54261a80) at 
Objects/dictobject.c:985
#22 0x7faa834d5033 in sipSimpleWrapper_clear (self=self@entry=0x5fd3200) at 
siplib.c:9559
#23 0x7faa834d74c2 in sipWrapper_clear (self=0x5fd3200) at siplib.c:9940
#24 0x7faa8610e17d in delete_garbage (old=0x7faa86391f60 generations+96, 
collectable=0x7b1e86f0) at Modules/gcmodule.c:769
#25 collect (generation=2) at Modules/gcmodule.c:930
#26 0x7faa8610e661 in PyGC_Collect () at Modules/gcmodule.c:1386
#27 0x7faa86103724 in Py_Finalize () at Python/pythonrun.c:444
#28 0x7faa86103248 in Py_Exit (sts=0) at Python/pythonrun.c:1760
#29 0x7faa8610333f in handle_system_exit () at Python/pythonrun.c:1134
#30 0x7faa861033f5 in PyErr_PrintEx 
(set_sys_last_vars=set_sys_last_vars@entry=1) at Python/pythonrun.c:1144
#31 0x7faa8610369a in PyErr_Print () at Python/pythonrun.c:1047
#32 0x7faa86103a2b in PyRun_SimpleFileExFlags (fp=optimized out, 
fp@entry=0x2339450, filename=optimized out, filename@entry=0x7b1e9268 
/usr/lib64/python2.7/site-packages/eric4/eric4.py, closeit=closeit@entry=1, 
flags=flags@entry=0x7b1e8920) at Python/pythonrun.c:947
#33 0x7faa86103d33 in PyRun_AnyFileExFlags (fp=fp@entry=0x2339450, 
filename=filename@entry=0x7b1e9268 

[Eric] Unable to create initial pysvn repo with eric4

[Hans-Peter Jansen]

Dear Detlev,

unfortunately, I currently suffer from another regression in eric4:

eric4 doesn't allow me to initialize a new pysvn project anymore, hosted on a 
remote apache server. The empty project was created on the server with 
svnadmin create as usual..

~ svn info http://server/repos/diskbackup
Pfad: diskbackup
URL: http://server/repos/diskbackup
Basis des Projektarchivs: http://server/repos/diskbackup
UUID des Projektarchivs: 5a45a0ec-9180-11e2-b6a6-310c6d2ea48d
Revision: 0
Knotentyp: Verzeichnis
Letzte geänderte Rev: 0
Letztes Änderungsdatum: 2013-03-20 18:05:23 +0100 (Mi, 20. Mär 2013)


With pysvn, the next step was creating a new project with eric, and let it 
setting up the initial repo structure (trunk, branches, tags), but the SVN 
repository info dialog doesn't enable the OK button anymore, using values 
similar to:

Protocol: http://
URL: server/repos/name
Log: whatever
Create standard repo structure: yes

This happens with both: 4.5.9 and 4.5.10-svn, the former tested with PyQt 
latest release versions, the latter with latest snapshot versions, everything 
else is pretty current (pysvn is at 1.7.6), and publicly available from my OBS 
repos:

https://build.opensuse.org/project/monitor?project=home%3Afrispete%3APyQt
https://build.opensuse.org/project/monitor?project=home%3Afrispete%3APyQt-next
https://build.opensuse.org/project/monitor?project=home%3Afrispete%3Apython

A manual repo setup from the client was possible, of course, and is usable for 
eric4 thereafter.

While at it: a minor glitch happens with the file selector (open file mode): 
by default, you select batch files, which is a little contra productive for 
The Python IDE. Either use all files (*) filter, or python files, if called 
from a python project. (I'm using (Py)KDE 4.10 integration), which also 
suffers from non translated dialog items, BTW).

Cheers,
Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] [QScintilla] No monospaced fonts anymore...

[Hans-Peter Jansen]

Am Sonntag, 10. März 2013, 16:46:08 schrieb Detlev Offenbach:
 Can you test latest sources from source code repository?

They're automatically picked up tonight and build in public obs:

https://build.opensuse.org/project/monitor?project=home%3Afrispete%3APyQt-next

I hope to get around testing it tomorrow.

Will let you know about the outcome.

Thanks,
Pete

 On Sunday 10 March 2013, 11:14:06 Hans-Peter Jansen wrote:
  Dear Detlev,
  
  Am Samstag, 9. März 2013, 17:09:16 schrieb Detlev Offenbach:
   Hello,
   
   that issue will be fixed in the next release. Sorry that it took so
   long,
   but I was busy doing some eric5 stuff.
  
  Since an workaround exists, there's no reason to hurry. What I found out
  meanwhile: the editor in sql mode doesn't play well: no matter, what
  settings are given (including the trick from below), it renders documents
  in variable spaced fonts.
  
  Thanks,
  Pete
  
   Regards,
   Detlev
   
   On Sunday 17 February 2013, 23:36:26 Hans-Peter Jansen wrote:
Hi Detlev,

this turned out to be a plain eric issue:

Am Dienstag, 22. Januar 2013, 01:09:20 schrieb Hans-Peter Jansen:
 with current qscintilla snapshots and eric4. It might be related to
 any
 of
 sip, PyQt4 snapshots, too.
 
 I was in a hurry, hence had to revert to the latest release
 versions,
 and
 got my monospaced fonts back this way..

I lost monospaced fonts within eric, although I've set:

Editor-Style: Use monospaced as default

and

Highlighters-Styles: Python - Monospaced Font (Bitstream Vera Sans
Mono)

but get them back, if I *deselect* Use Monospaced Font in editors
context
menu (of a python file), and loose them the other way around. This
behavior
is reproducible here with:

python: 2.7.3
sip: 4.14.3
qt4: 4.8.4
pyqt4: 4.9.6
eric4: 4.5.8

BTW, shouldn't this option just toggle the related Editor-Style
setting?

Cheers,
Pete


___
QScintilla mailing list
qscinti...@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/qscintilla-- 
   
   *Detlev Offenbach*
   det...@die-offenbachs.de--
 
 *Detlev Offenbach*
 det...@die-offenbachs.de
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] [QScintilla] No monospaced fonts anymore...

[Hans-Peter Jansen]

Hi Detlev,

this turned out to be a plain eric issue:

Am Dienstag, 22. Januar 2013, 01:09:20 schrieb Hans-Peter Jansen:
 with current qscintilla snapshots and eric4. It might be related to any of
 sip, PyQt4 snapshots, too.
 
 I was in a hurry, hence had to revert to the latest release versions, and
 got my monospaced fonts back this way..

I lost monospaced fonts within eric, although I've set:

Editor-Style: Use monospaced as default

and 

Highlighters-Styles: Python - Monospaced Font (Bitstream Vera Sans Mono)

but get them back, if I *deselect* Use Monospaced Font in editors context 
menu (of a python file), and loose them the other way around. This behavior is 
reproducible here with:

python: 2.7.3
sip: 4.14.3
qt4: 4.8.4
pyqt4: 4.9.6
eric4: 4.5.8

BTW, shouldn't this option just toggle the related Editor-Style setting?

Cheers,
Pete


___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

[Eric] Strange eric4 start up issue

[Hans-Peter Jansen]

Hi Detlev,

after upgrading to openSUSE 12.2, eric4 4.5.6, and tidying up the usual 
fallout (svn upgrade was necessary this time), I suffer from a real strange 
start up issue of eric4.

In my project Makefiles, I usually use a target like this:

edit:
@$(ERIC) $(BASENAME).e4p 

Now, this takes ages to complete, and keeps a strange bash -i instance 
running in a loop.

Investigating that a bit further shows, that I'm not able to start eric from a 
shell in the background (eric4 ), as that leads to a stopped eric process.
Sigkilling that reveals the bash -i instance running wild.

I seems to happen when eric creates its toolbars. 

hp@xrated:~ eric4 --debug 
[1] 22082   
 
hp@xrated:~ DEBUG:root:Importing Preferences   
 
DEBUG:root:Importiere Packages...   
 
DEBUG:root:Erzeuge das Hauptfenster...  
 
DEBUG:root:Initialisiere Plugin Manager...  
 
DEBUG:root:Erzeuge das Hauptfenster...  
 
DEBUG:root:Creating Layout...   
 
DEBUG:root:Creating Viewmanager...  
 
DEBUG:root:Creating toolboxes...
 

If you cannot reproduce this, let me know.

Cheers,
Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] a few eric4 usability issues

[Hans-Peter Jansen]

On Thursday 17 February 2011, 19:38:05 Detlev Offenbach wrote:
 On Sonntag, 30. Januar 2011, Hans-Peter Jansen wrote:
   * An action to remove/restore missing items would be nice

 This functionality will be in the next eric 4.5 snapshot (it is in
 the repository already).

Detlev, you're so cool, thanks. 

And sorry for the quietness on my side, but I'm deep inside a complex 
project, that take my full attention and doesn't allow any setup 
changes for the moment..

Will test the new stuff in the next days. 

Thanks again,
Pete

___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] a few eric4 usability issues

[Hans-Peter Jansen]

On Sunday 30 January 2011, 12:28:25 Detlev Offenbach wrote:
 On Donnerstag, 27. Januar 2011, Hans-Peter Jansen wrote:
  On Thursday 27 January 2011, 19:20:17 Detlev Offenbach wrote:
   On Mittwoch, 26. Januar 2011, Hans-Peter Jansen wrote:
Hi Detlev,
   
I would like to discuss some usability aspects of eric4:
   
   
Check in: I tend to prefer checking in larger changes on the
command line (to svn), since that gives me the hint, what files
are checked in beforehand. Ideally, I would want to check the
diff of some files' changes before proceeding. I'm dreaming of
a check in facility, that offers an interface similar to svn
ci in the shell, but with the possibility to check the diffs
of certain files, or a full diff. You offer all of these
functions within the version control context menu, but the
workflow is pretty uncomfortable.
  
   That would be a dialog like kdesvn commit dialog, right?
 
  Yes, something similar, although that dialog suffers from its
  attempt to manage new files by default..

 How about extending the Status dialog. It already has all the
 functionality except the diff. I could move the context menu entries
 to buttons and place them to the right of the list.

Hey, yes, that would be great. Btw, I never noticed the context menu in 
that dialog (probably, because its functionality wasn't immediately 
obvious). It might be worth thinking about adding some actions as 
buttons below the list on the left (including the diff). That way 
(given that the list is in multi selection mode) one is able to call 
the diff/check in/revert for the currently selected items in one go. A 
checkbox in the list view items could visualize active items, e.g the 
ones that the button actions operate on. All items should be set on by 
default. Then this dialog deserves a global shortcut, and would be the 
one stop shop for all major version control operations.

What do you think?

Pete

P.S.: I'm away for the rest of the day.
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] a few eric4 usability issues

[Hans-Peter Jansen]

On Sunday 30 January 2011, 12:28:25 Detlev Offenbach wrote:
 On Donnerstag, 27. Januar 2011, Hans-Peter Jansen wrote:
  On Thursday 27 January 2011, 19:20:17 Detlev Offenbach wrote:
   On Mittwoch, 26. Januar 2011, Hans-Peter Jansen wrote:
Hi Detlev,
   
I would like to discuss some usability aspects of eric4:
   
   
Check in: I tend to prefer checking in larger changes on the
command line (to svn), since that gives me the hint, what files
are checked in beforehand. Ideally, I would want to check the
diff of some files' changes before proceeding. I'm dreaming of
a check in facility, that offers an interface similar to svn
ci in the shell, but with the possibility to check the diffs
of certain files, or a full diff. You offer all of these
functions within the version control context menu, but the
workflow is pretty uncomfortable.
  
   That would be a dialog like kdesvn commit dialog, right?
 
  Yes, something similar, although that dialog suffers from its
  attempt to manage new files by default..

 How about extending the Status dialog. It already has all the
 functionality except the diff. I could move the context menu entries
 to buttons and place them to the right of the list.

A few more notes: 
 * The column resizing seems not to work properly for me at least
 * It would be nice to filter items based on their status
 * An action to remove/restore missing items would be nice

Pete

python: 2.6
sip: 4.12
qt4: 4.6.3
pyqt4: 4.8.2

___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] Fwd: Segmentation fault in eric-4.4.11 when opening File open dialog

[Hans-Peter Jansen]

On Wednesday 26 January 2011, 20:12:00 Juan Manuel Santos wrote:
 Forwarding to list so we can keep it in sync. Sorry for the hassle.

 -- Forwarded message --
 From: Juan Manuel Santos vicariou...@gmail.com
 Date: Wed, Jan 26, 2011 at 4:09 PM
 Subject: Re: [Eric] Segmentation fault in eric-4.4.11 when opening
 File open dialog
 To: Detlev Offenbach det...@die-offenbachs.de




 On Wed, Jan 26, 2011 at 3:55 PM, Detlev Offenbach

 det...@die-offenbachs.dewrote:
  I don't know. Did you update anything recently? Due to the fact,
  that eric doesn't seg fault all around the world, I suspect
  something else is not working right on your system.
 
  Detlev
  --
  Detlev Offenbach
  det...@die-offenbachs.de

 I just tested Eric4 (same version) on my work laptop, which has
 KDE-4.4 and it did not segfault. I'm guessing this has something to
 do with either with KDE-4.6 or with KDE-4.6-rc2, which is what i'im
 using on the desktop.

 On another note, I am not experiencing the QScintilla keyboard layout
 reset problem on my laptop either, and am using the same QScintilla
 versions as on the desktop. Could this be related to KDE also? In
 both KDEs my layout is set to en_US international variant.

Juan, do you use kdebindings-python, python-kde or whatever your distro 
names these packages? If yes, try to deinstall them, if not, most 
probably, your Qt/QScintilla/sip/PyQt installation is inconsistent. 
Play with the PyQt examples and see, if you can reproduce those 
crashes..

Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

[Eric] a few eric4 usability issues

[Hans-Peter Jansen]

Hi Detlev,

I would like to discuss some usability aspects of eric4:

Autocompletion: I would like to have autocompletion working for 
translated ui files, but it seems, that no combination of settings 
allows me to do so: 
 * Autocompletion is enabled, case sensitive, replacing words with
   threshold 2
 * Eric AC is enabled from API files, document, and project
   (but doesn't seem to work)
 * QScintilla AC is enabled bot showing singles, not using fillup chars
   with document and API files sources

The latter seems to work fine, but isn't able to use any project members 
apart from the current document.

I've disabled Rope AC, since it operates in a too intrusive way. 

Do I really need to manually create api files on any UI change and add 
an API file for every project?


Other files handling: usually, one selects some changed files in the 
project viewer in order to check them in. Loading the associated app 
isn't the best thing to do, when selecting them, especially, if you 
click on them with some accelerator key pressed (to select many or 
regions of files). It's simply not funny, that eric triggers a dozen 
oowriter instances in that case. Better provide that as a context menu 
item. Same goes for UI files.


Check in: I tend to prefer checking in larger changes on the command 
line (to svn), since that gives me the hint, what files are checked in 
beforehand. Ideally, I would want to check the diff of some files' 
changes before proceeding. I'm dreaming of a check in facility, that 
offers an interface similar to svn ci in the shell, but with the 
possibility to check the diffs of certain files, or a full diff. You 
offer all of these functions within the version control context menu, 
but the workflow is pretty uncomfortable.


It's a real pity, that you dropped Qt3 support a long time ago. Since I 
still have a host of such projects running, I feel discriminated 
compared to, say, wxpython users (which I classify much more legacy 
then Qt3..).


Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

[Eric] eric4-4.4.10-0.3872 snapshot evaluates a project member module on startup

[Hans-Peter Jansen]

Hi Detlev,

I see a strange behavior with current eric4:

a projects module is called browsermainwindow.py, which is in broken 
state, as I'm on the way to convert it from C++ to python. When I 
start eric with the project, it gets into an endless loop without any 
UI responses. The only thing I can do is to kill eric, then. The same 
happens, if run eric and open that project. It does not happen, if I 
rename the browsermainwindow.py to browsermainwindow.p, it opens the 
project with many other similar broken modules and just warns about 
the missing member browsermainwindow.py.

Stracing the startup reveals something strange (comments inlined):

# check, if module exists: yes
[pid 19196] stat64(.../browser/browsermainwindow.py, {st_mode=S_IFREG|0644, 
st_size=35825, ...})
# check, if directory has __init__.py fails:
[pid 19196] stat64(.../browser/__init__.py, 0xad5e2f48) = -1 ENOENT (No such 
file or directory
# now for something really strange: eric tries to execute the module
# in some way, otherwise this won't happen:
[pid 19196] stat64(.../browser/browsermainwindow, 0xad5dfe94) = -1 ENOENT (No 
such file or directory)
[pid 19196] open(.../browser/browsermainwindow.so, O_RDONLY|O_LARGEFILE) = -1 
ENOENT (No such file or directory)
[pid 19196] open(.../browser/browsermainwindowmodule.so, 
O_RDONLY|O_LARGEFILE) = -1 ENOENT (No such file or directory)
[pid 19196] open(.../browser/browsermainwindow.py, O_RDONLY|O_LARGEFILE) = 28
[pid 19196] fstat64(28, {st_mode=S_IFREG|0644, st_size=35825, ...}) = 0
[pid 19196] fstat64(28, {st_mode=S_IFREG|0644, st_size=35825, ...}) = 0
[pid 19196] _llseek(28, 0, [0], SEEK_CUR) = 0
[pid 19196] fstat64(28, {st_mode=S_IFREG|0644, st_size=35825, ...}) = 0
[pid 19196] mmap2(NULL, 1048576, PROT_READ|PROT_WRITE, 
MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xacc61000
[pid 19196] _llseek(28, 0, [0], SEEK_CUR) = 0
[pid 19196] mprotect(0xae7f9000, 28672, PROT_READ|PROT_WRITE) = 0
[pid 19196] futex(0x8064280, FUTEX_WAKE_PRIVATE, 1) = 1
[pid 19196] read(28, ..., 1048576) = 35825

Then, after a couple of mprotect, mremap, mmap2 calls, eric stalls.

Obviously, the evaluation ends in some endless looping, but the big
question is, why does eric try to eval the module in the first place.
This can lead do really silly things to happen..

python -v seems to be of no help, as it doesn't show this module loading
at all (probably you're forking on startup, and -v traces the wrong track).

The last message is:
import QScintilla.TypingCompleters.CompleterBase # precompiled 
from 
/usr/lib/python2.6/site-packages/eric4/QScintilla/TypingCompleters/CompleterBase.pyc

Executing the module directly leads to:

  File browsermainwindow.py, line 15
from bookmarks import ..
  ^
SyntaxError: invalid syntax

as expected. Let me know, if I can do more for this to resolve.

Pete


browsermainwindow.py
Description: application/python
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] eric4 4.3.9 crashes with Segmentation fault about 5 minutes after starting [SEC=UNCLASSIFIED]

[Hans-Peter Jansen]

Dear Matthew,

please look up my reply to Arndt Meiers issue:
[Eric] segmentation fault with Eric4-4.3.8 on SUSE 11.1_x64
in this list. Funny enough, he's also member of gov.au, but they're using 
a much less scary disclaimer. Since I've no idea, what the section 70 of 
the Crimes Act 1914 determines, I hope, I'm not getting delinquent with my 
reply below ;-).

While that answer targeted to a sane openSUSE debug environment, the basic 
idea remains. Build/install debug info of glibc, libstdc++, qt4, python, 
sip (./configure.py --debug) and PyQt, and create a proper backtrace. 

BTW, instead of reinventing the wheel, err, recreating everything from 
scratch with a non reproducible outcome, I would start thinking about some 
virtual machine setup on the SAN with a _deterministic_ setup. That way, a 
debug environment would have been a few finger snips away, could be kept 
current easily, and you would be able to get a full featured version of 
eric (just by using the right repo, i.e. mine). On your current route, 
that's going to take you a awful lot of continuous effort.

Good luck,
Pete

On Thursday 10 December 2009, 07:21:26 Thyer, Matthew wrote:
 UNCLASSIFIED

 Hello,

 I have built eric4 4.3.9 on RHEL 4.4 x86_64 in another prefix
 /usr/local/san and with a non-default compiler for that system.
 I've had to install another compiler as the system comes with GCC 3.4.6
 but Python 2.6.4 had to be built with GCC 4.1 for the NumPy Python
 module as it required newer FORTRAN support for updated LAPACK and BLAS
 libraries.

 eric4 seems to run fine but crashes with Segmentation fault about 5
 minutes after starting.

 Any ideas ?
 Or at least any pointers on how to build debug versions of everything to
 work out why it's crashing ?
 I suppose I'll have to build a debug Python 2.6.4 too.


 The build procedure was as follows:


 Qt 4.5.3 (use the open source edition):

 export CC=gcc4

 export CXX=g++4
 ./configure -prefix /usr/local/san
 make
 make install


 SIP 4.9.3:

 export CC=gcc4
 export CXX=g++4
 /usr/local/san/bin/python configure.py
 make
 make install


 QScintilla-gpl 2.4:

 cd QScintilla-gpl-2.4/Qt4
 export CC=gcc4

 export CXX=g++4
 export LINK=g++4
 /usr/local/san/bin/qmake qscintilla.pro
 find . -name Makefile -exec sed -r -i '/^(CC|CXX|LINK) *=/ s/^/#/' {} \;
 make
 make install


 PyQt4 4.6.2:

 export CC=gcc4

 export CXX=g++4
 export LINK=g++4
 /usr/local/san/bin/python configure.py -q /usr/local/san/bin/qmake
 find . -name Makefile -exec sed -r -i '/^(CC|CXX|LINK) *=/ s/^/#/' {} \;
 make
 make install


 QScintilla-gpl 2.4 Python bindings:

 cd QScintilla-gpl-2.4/Python
 /usr/local/san/bin/python configure.py
 find . -name Makefile -exec sed -r -i '/^(CC|CXX|LINK) *=/ s/^/#/' {} \;
 make
 make install


 eric4 4.3.9:

 export CC=gcc4

 export CXX=g++4
 export LINK=g++4
 /usr/local/san/bin/python install.py -b /usr/local/san/bin


  Matthew Thyer Phone:  +61 8 8259 7249
  Science Corporate Information Systems Fax:+61 8 8259 5537
  Defence Science and Technology Organisation, Edinburgh
  P.O. Box 1500 EDINBURGH South Australia 5111



 IMPORTANT: This email remains the property of the Australian Defence
 Organisation and is subject to the jurisdiction of section 70 of the
 Crimes Act 1914. If you have received this email in error, you are
 requested to contact the sender and delete the email.


___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] PyLint plugin not producing any output

[Hans-Peter Jansen]

On Sunday 06 December 2009, 13:11:21 detlev wrote:
 Hi,

 what I forgot to mention is, that a single module of a project may be
 checked via the project sources viewer context menu. Shall an entry be
 added to each editor context menu as well?

It would fit nicely to the Check sub menu. 

 Regards,
 Detlev

 On Samstag, 5. Dezember 2009, Dan Bullok wrote:
  On Saturday 05 December 2009 11:46:47 you wrote:
   On Dienstag, 1. Dezember 2009, Dan Bullok wrote:
On advice from Mikhail Terekhov, I put an empty __init__.py file in
the same directory as my project.  PyLint works properly now.  I
don't really understand this requirement, as I run PyLint on single
files from the command line all the time.  And if it is a
requirement, then it should definitely be stated clearly in the
PyLint dialog that it only works on packages, and suggest that the
user put an empty __init__.py file in the project directory if they
are not already working on a package. -Dan
  
   The PyLint plugin is used to check the project (that's why it is
   accessible via the project menu). pylint -h gives you the following
   output.
  
   Usage:  pylint [options] module_or_package
  
   That clearly indicates, that pylint only works on modules and
   packages. Consequently the project's main directory must be a
   package. Should eric4 generate an empty __init__.py file, if it
   doesn't discover one in the project dir?
 
  Hi Detlev.
 
  I agree that it's absolutely logical that PyLint works this way. Since
  it's invoked from the project menu, it makes sense that it acts on the
  entire project.  However, I think that this can be confusing from the
  user's perspective, especially for someone who is used to using PyLint
  from the command line (In my own use, I almost always run it on one
  file at a time). My thought was that if I click on Check-PyLint, the
  file that I'm currently editing should be checked.  Now, in retrospect,
  since all of the functions on the project menu act on the entire
  project, my assumption is obviously not a good one.  However, I think
  it may be a common misunderstanding.
 
  In any case, if the user runs pylint, and there is no __init__.py file,
   they get an empty result dialog, which is very frustrating.  An error
   message would be extremely helpful to figure out WHY it isn't working.
   Something like PyLint can only be run on a package.  If you wish to
  run PyLint, your project's main directory must contain an __init__.py
  file. An empty file will suffice.

Do it the other way around. If called from the menu, check, if __init__.py 
exists, and if not, present a simple dialog, stating that pylint for the 
whole project needs an empty __init__.py module, and ask the user, if eric 
should fix this problem by creating the file and adding it to the project.

A hint on how to check single modules would be nice in that dialog, too.

  Also, I think it would be useful to be able to check ONLY the file
  being edited.  If you agree, I'd be happy to create patches to do both
  of these things.

Cheers,
Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] Re: eric for openSUSE WAS: pysvn versus svn

[Hans-Peter Jansen]

On Wednesday 04 November 2009, 17:10:41 Mikhail wrote:
 Hans-Peter Jansen h...@... writes:
  For those, who are interested in the fallout:

 http://download.opensuse.org/repositories/home:/frispete:/branches:/KDE:/
KDE4:/Factory:/Desktop

  Please _do_ provide _feedback_, if you test/use these packages.
  Especially but not limited to the pylint issue..

 Peter,

 I've tried to install 4.3.8 from your repository and it doesn't work for
 me. It looks like python-pyqt4 and python-sip packages are not
 compatible:

 ~ /usr/bin/eric
 Traceback (most recent call last):
   File /usr/lib64/python2.5/site-packages/eric4/eric4.py, line 47, in
 module from KdeQt.KQApplication import KQApplication
   File /usr/lib64/python2.5/site-packages/eric4/KdeQt/__init__.py, line
 20, in module
 import Preferences
   File
 /usr/lib64/python2.5/site-packages/eric4/Preferences/__init__.py, line
 26, in module
 from PyQt4 import Qsci
 RuntimeError: the sip module implements API v6.0 but the PyQt4.Qsci
 module requires API v5.0

You missed the python-qscintilla package from my repo. Mind looking into my 
mail regarding segmentation fault with Eric4--4.3.8 on SUSE 11.1_x64, 
where I described simple procedures to add repos to your system, and keep 
your installation consistent? Most packages in that repo are useful add-ons 
to get the most out of eric..

 BTW the rpms are:
 ~ rpm -qa | grep -i python-qt4
 python-qt4-4.6.1-53.2
 ~ rpm -qa | grep -i python-sip
 python-sip-4.9.1-54.1

Good luck,
Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] segmentation fault with Eric4-4.3.8 on SUSE 11.1_x64

[Hans-Peter Jansen]

On Thursday 29 October 2009, 01:23:06 Arndt Meier wrote:
 Hi!

 Eric4 starts its IDE and all is well. I get a message eric4 has not been 
 configured yet. The configuration dialog will be started. 
OK. Whether or not I press OK, after some 30 seconds eric4 just crashes with 
no error message in the bash shell other than segfault:
  eric4

 Warning: translation file 'qt_en_GB'could not be loaded.
 Using default.
 Warning: translation file 'eric4_en_GB'could not be loaded.
 Using default.
 Warning: translation file 'qscintilla_en_GB'could not be loaded.
 Using default.
 Segmentation fault (core dumped)

 ls -lrt  shows
 -rw---  1 meiera users 91054080 2009-10-29 10:42 core.32758

 And even stranger, gdb does not understand the core dump
 ~ gdb core.32758 /usr/local/bin/eric4
 GNU gdb (GDB; openSUSE 11.1) 6.8.50.20081120-cvs
 Copyright (C) 2008 Free Software Foundation, Inc.
 License GPLv3+: GNU GPL version 3 or later
 http://gnu.org/licenses/gpl.html This is free software: you are free to
 change and redistribute it. There is NO WARRANTY, to the extent permitted
 by law.  Type show copying and show warranty for details.
 This GDB was configured as x86_64-suse-linux.
 For bug reporting instructions, please see:
 http://bugs.opensuse.org/...
 /home/meiera/core.32758: not in executable format: File format not
 recognized /usr/local/bin/eric4 is not a core dump: File format not
 recognized (gdb) bt
 No stack.

 Starting a minimalistic eric:
 eric4 --nosplash --noopen  --nokde
 and not doing/touching anything, again after some 30 seconds eric4
 crashes and again there is a pseudo core dump -rw---  1 meiera users
 76963840 2009-10-29 11:10 core.402
 that gdb again cannot read (as above).

 I am running:
 openSUSE 11.1_x64
 Qt 4.5.2
 QScintilla-gpl-2.4
 PyQt-x11-gpl-4.6.1
 sip-4.9.1
 python 2.6.2
 python-kde4 4.3.1-2.3

You seem to compile stuff from tarballs (and might be using 
rpm --force | --nodeps for any of these or related ones, e.g. kdelibs4).

That tends to produce unreproducible effects, and any of the advices below
might be void. Better start with a full manual cleanup (remove everything 
from /usr/local apart from directories), and _reinstall_ the packages, that 
you installed manually, with yast2.

 Any help in resolving this error is greatly appreciated.

Then you might try my packages from:

http://download.opensuse.org/repositories/home:/frispete:/branches:/KDE:/KDE4:/Factory:/Desktop/

They contain full debugging infos, btw.

In the openSUSE_11.1_KDE_Qt tree, I provide packages based on Qt 4.5.3
from http://download.opensuse.org/repositories/KDE:/Qt/openSUSE_11.1
while openSUSE_Factory is based on current KDE 4.3.1 from:
http://download.opensuse.org/repositories/KDE:/KDE4:/Factory:/Desktop/openSUSE_11.1
This should allow you to rebuild your environment fully rpm based in any
flavor you want.

Further debugging is better done with these packages installed:

gcc43-debuginfo-4.3.3_20081022-9.5
gcc43-debugsource-4.3.3_20081022-9.5
glibc-debuginfo-2.9-2.12.1
glibc-debugsource-2.9-2.12.1
kdelibs4-debuginfo-4.1.3-4.10.4
kdelibs4-debugsource-4.1.3-4.10.4
libqt4-debuginfo-4.5.3-70.1
libqt4-debugsource-4.5.3-70.1
python-base-debuginfo-2.6.0-2.22.1
python-base-debugsource-2.6.0-2.22.1
python-debuginfo-2.6.0-2.21.1
python-debugsource-2.6.0-2.21.1
python-qt4-debuginfo-4.6.1-53.1
python-qt4-debugsource-4.6.1-53.1
python-sip-debuginfo-4.9.1-54.1
python-sip-debugsource-4.9.1-54.1

available from:
http://download.opensuse.org/debug/distribution/11.1/repo/oss/ and
http://download.opensuse.org/debug/update/11.1/

Minors might differ..

All these URLs contain repo files, that is easiest fetched with the attached
script. 

Make your choice of repos, and run zypper -v dup then. It might be necessary
to force installation of the right eric with zypper -v in eric-4.3.8-40.5.

Lookup the available minors with zypper -v se -s eric.

Then run eric:
gdb python -ex set args /usr/lib/python2.6/site-packages/eric4/eric4.py -ex 
run

This should give you a usable backtrace (bt full).

Hope, that gets you started ;-)..

Pete


wget_repo.sh
Description: application/shellscript
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] navigator pane

[Hans-Peter Jansen]

On Monday 19 October 2009, 21:21:47 Kamil Páral wrote:
 Dne 19.10.2009 19:50, detlev napsal(a):
  Sure, just use the file browser. Opening (expanding) a Python file will
  show the internal structure.
 
  Detlev

 I must say the the navigator in geany/spyder/eclipse is a little more
 handy. 

What's missing - could you elaborate?

Pete

___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] Segmentation fault

[Hans-Peter Jansen]

Am Mittwoch, 26. August 2009 schrieb Star Glider:
 Hi,

 I'm using eric4 all day and never exit the application, but today I need
 to perform a maintenance in my computer and when get
 to work with eric4 it says: segmentation fault.
 I uninstall and after I run the installer it says:
 [zo...@canil eric4-4.4-snapshot-20090821]$ python install.py
 Checking dependencies
 Python Version: 2.6.0
 Segmentation fault

 I'm using Fedora Linux 11/KDE in a amd 64 bit with the last release of
 eric4

Almost certainly, your sip/PyQt/qscintilla setup is damaged...

A lot less probably, your settings in ~/.eric4 are.

A gdb bt would help.

Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] Segmentation fault

[Hans-Peter Jansen]

[Posted to ML also to be a daunting example of consequences, when mixing 
installations from tarballs with those from package management]

Am Donnerstag, 27. August 2009 schrieben Sie:
 PyQt4 was not there!
 Maybe because I install a program called Camelt!

 but now that I have PyQt4 I get :

 [r...@canil eric4-4.4-snapshot-20090821]# python install.py
 Checking dependencies
 Python Version: 2.6.0
 Found PyQt
 Segmentation fault


Looks like you trashed your installation - usually by installation of 
arbitrary tarballs, while those packages are installed by rpm.

This is the best recipe to produce such a disaster, you're looking at.

Now you know the reason, why you should _always_ install rpms and never ever 
install tarballs in a sane system. I don't. If you need to update to a 
newer version, roll your own _RPM_. Learn to build rpms, and your life's 
getting much easier under linux. While at it, be careful with rpm --nodeps 
and rpm --force. There's a strong reason, why yum doesn't provide those.

OTOH, rpm -e  rpm -Uhv is fine..

Note, that package management is one of the greatest advantages of linux 
over other OS. The computer equivalent of the difference between childhood 
and being grown-up.

Come back, when you cleaned /usr/local (basically, remove _all_ _files_ 
there), reinstalled everything below /usr/lib64/python2.6 from rpm and when 
output of rpm -Va doesn't contain any lines with 5 in the third column 
which aren't below /etc.

If you managed to provide a --prefix=/usr argument to a connfigure script, 
wipe your platter, reinstall, and follow my words in the first paragraphs.

Pete


 2009/8/27 Hans-Peter Jansen h...@urpla.net

  Am Donnerstag, 27. August 2009 schrieben Sie:
   Allways learning [?]
  
   the output :
   (gdb) set args /usr/lib64/python2.6/site-packages/eric4/eric4.py
   (gdb) run
   Starting program: /usr/bin/python
   /usr/lib64/python2.6/site-packages/eric4/eric4.py
   [Thread debugging using libthread_db enabled]
   /usr/bin/python: can't open file
   '/usr/lib64/python2.6/site-packages/eric4/eric4.py': [Errno 2] No
   such file or directory
  
   Program exited with code 02.
  
   the dir /usr/lib64/python2.6/site-packages/eric4 doesn't exist
   probalby because after the first segmentation fault I uninstall and
   eric4 and now
   I can't install again.
 
  Does
 
  python -c from PyQt4 import QtCore, QtGui
 
  work? If not:
 
 
  $ gdb python
  (no debugging symbols found)
  (gdb) set args -c from PyQt4 import QtCore, QtGui
  (gdb) run
  segfault
  (gdb) bt
 
  Pete
 
   Thanks.
  
   Jorge
  
  
  
  
   2009/8/27 Hans-Peter Jansen h...@urpla.net
  
Am Donnerstag, 27. August 2009 schrieben Sie:
 Thanks Pete, but whats a gdb bt?
   
a gnu debugger command: backtrace
$ gdb python
(no debugging symbols found)
(gdb) set args /usr/lib/python2.6/site-packages/eric4/eric4.py
(gdb) run
segfault
(gdb) bt
...
   
Paste this into  mail.
   
Pete
   
 jorge

 2009/8/27 Hans-Peter Jansen h...@urpla.net

  Am Mittwoch, 26. August 2009 schrieb Star Glider:
   Hi,
  
   I'm using eric4 all day and never exit the application, but
   today I need to perform a maintenance in my computer and when
   get to work with eric4 it says: segmentation fault.
   I uninstall and after I run the installer it says:
   [zo...@canil eric4-4.4-snapshot-20090821]$ python install.py
   Checking dependencies
   Python Version: 2.6.0
   Segmentation fault
  
   I'm using Fedora Linux 11/KDE in a amd 64 bit with the last
   release of eric4
 
  Almost certainly, your sip/PyQt/qscintilla setup is damaged...
 
  A lot less probably, your settings in ~/.eric4 are.
 
  A gdb bt would help.
 
  Pete
  ___
  Eric mailing list
  Eric@riverbankcomputing.com
  http://www.riverbankcomputing.com/mailman/listinfo/eric


___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: [Eric] Re: eric for openSUSE

[Hans-Peter Jansen]

Am Montag, 10. August 2009 schrieb Mikhail:
 Hans-Peter Jansen h...@... writes:
  Would you be so kind and check the pylint functionality after your next
  update?

 It installed fine this time, but when I run a check in eric I got the
 following exception:

 Traceback (most recent call last):
   File /usr/bin/pylint, line 4, in module
 lint.Run(sys.argv[1:])
   File /usr/lib64/python2.5/site-packages/pylint/lint.py, line 881, in
 __init__ linter.check(args)
   File /usr/lib64/python2.5/site-packages/pylint/lint.py, line 500, in
 check self.check_astng_module(astng, checkers)
   File /usr/lib64/python2.5/site-packages/pylint/lint.py, line 577, in
 check_astng_module
 if implements(checker, IASTNGChecker)])
   File /usr/lib64/python2.5/site-packages/pylint/lint.py, line 594, in
 astng_events
 self.astng_events(child, checkers, _reversed_checkers)
   File /usr/lib64/python2.5/site-packages/pylint/lint.py, line 594, in
 astng_events
 self.astng_events(child, checkers, _reversed_checkers)
   File /usr/lib64/python2.5/site-packages/pylint/lint.py, line 594, in
 astng_events
 self.astng_events(child, checkers, _reversed_checkers)
   File /usr/lib64/python2.5/site-packages/pylint/lint.py, line 594, in
 astng_events
 self.astng_events(child, checkers, _reversed_checkers)
   File /usr/lib64/python2.5/site-packages/pylint/lint.py, line 591, in
 astng_events
 checker.visit(astng)
   File /usr/lib64/python2.5/site-packages/logilab/astng/utils.py, line
 320, in visit
 method(node)
   File /usr/lib64/python2.5/site-packages/pylint/checkers/format.py,
 line 267, in visit_default
 prev_sibl = node.previous_sibling()
   File /usr/lib64/python2.5/site-packages/logilab/astng/nodes.py, line
 251, in previous_sibling
 stmts = self.parent.child_sequence(self)
   File /usr/lib64/python2.5/site-packages/logilab/astng/nodes.py, line
 216, in child_sequence
 raise ASTNGError(msg % (repr(child), repr(self)))
 logilab.astng._exceptions.ASTNGError: Could not found _ast.Yield object
 at 0xf39f90 in _ast.Assign object at 0xf39f10's children

Not here, could you give me a list of all concerning packages, and could you 
try to pylint something simple?

Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

[Eric] Re: eric for openSUSE

[Hans-Peter Jansen]

Am Freitag, 7. August 2009 schrieb Mikhail:
  You're right, and I've checked in the dependencies into the eric
  package a few moments ago (that way, they're in one place, ad can get
  easily removed for the distribution).
  Unfortunately, the build system is under heavy factory build stress,
  thus it will take some time... Did you install python-sip, python-qt4,
  and the qscintilla packages as well from there (please note the -devel
  split)?

 Yes, I've installed python-sip, python-qt4, python-rope and qscintilla
 packages from there.

 It looks like there is a problem with python-rope rpm. When I try to
 update eric rpm from 4.3.6-19.1 to 4.3.6-20.3 YaST shows a warning:
 nothing provides python-rope = 0.92 needed by eric-4.3.6-20.3.x86_64.

Sure, there's no 0.92 indeed, only a 0.9.2. Fixed. Should be ready in a few 
hours :-(...

Thanks for testing, Mikhail. 

Would you be so kind and check the pylint functionality after your next 
update?

Thanks  regards,
Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

eric for openSUSE WAS: [Eric] pysvn versus svn

[Hans-Peter Jansen]

Am Mittwoch, 5. August 2009 schrieb Hans-Peter Jansen:

 I'm in the process of preparing full featured eric packages via openSUSE
 build service (http://build.opensuse.org). Eg.: all of erics dependencies
 are fairly current: pylint (logilab-common, logilab-astng), rope, besides
 the usual suspects ;-)

I forgot to note, python-sip contains a backported fix from Phil for the 
pylint issue. 

For those, who are interested in the fallout:
http://download.opensuse.org/repositories/home:/frispete:/branches:/KDE:/KDE4:/Factory:/Desktop

Please _do_ provide _feedback_, if you test/use these packages. Especially 
but not limited to the pylint issue..

I'm out for a day, will be back tonight.

Thanks,
Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

[Eric] pysvn versus svn

[Hans-Peter Jansen]

Dear Detlev,

I'm in the process of preparing full featured eric packages via openSUSE 
build service (http://build.opensuse.org). Eg.: all of erics dependencies 
are fairly current: pylint (logilab-common, logilab-astng), rope, besides 
the usual suspects ;-)

Then I remembered, that I noticed pysvn, when eric asked me the last time, 
which version control system I want to use. Since I do use svn, and I've an 
issue since day one with it and eric: when I'm connecting to a local svn 
server via http for the first time, I have to provide an password, which is 
fetched from the console: yes, right, I need to run eric from an xterm, 
setup the project, and enter the pw in the xterm, when erics svn setup 
dialog asks me for it...

Now I ask myself, is that fixed, when using pysvn, or more general, could 
you briefly outline the difference of both implementations (from users 
perspective). The only other issue, I have with svn is, that it doesn't 
list the tags after eric's termination anymore.

BTW, if you create an Novell account, you can watch this effort here:
home:frispete:branches:KDE:KDE4:Factory:Desktop

Hopefully, I find time to contribute to eric codewise, too. I've a few ideas 
here and there (e.g. a smart dialog for check-in with the diff in the 
advanced section, since I nearly always look at the diff before I commit).

BTW, do you have an idea, who's in change of governing, that Linux-users 
don't want shortcuts for dialog buttons - that's just ridiculous. 

E.g., take svn diff/check-in dialogs in a typical work flow: look at the 
diff, enter some log text, and then it takes 4 tabs to get to the Safe 
button. Another dialog opens with the commit message, which has Close 
and Cancel, while the still opened diff has Save and Close. Notice 
the while the Close button swapped, which is unfortunate, the 
commits Cancel is disabled, thus forcing the user to push Close left 
(in commit) - Close right (in diff). Inconvenient. 

Probably, I'm going to create a petition against using Qt's DialogButtonBox 
for these reasons.

Sorry for being so gabby today.

Kind regards,
Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

[Eric] PyQt3 api files

[Hans-Peter Jansen]

Hi Detlev,

since PyQt3 will still come along for quite some time, is there a way to 
generate PyQt3 api files, since eric4-api is intended for python based 
modules only?

Apart from a few exceptions popping up on project load, eric4 feels nice 
(after I managed to finally switch over...)

Thanks,
Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric

Re: *** GMX Spamverdacht *** [Eric] Seg fault at startup

[Hans-Peter Jansen]

Am Donnerstag, 25. Januar 2007 10:22 schrieb Andreas Pakulat:

 The strace is of 0 use when trying to debug a segfault. You have to run
 the python session in gdb, i.e.


 python --args python eric3

better use:
gdb --args python eric3


 And when the app crashed produce a backtrace by issuing bt.


Pete
___
Eric mailing list
Eric@riverbankcomputing.com
http://www.riverbankcomputing.com/mailman/listinfo/eric