Regarding the latest discussion about the Mac version.
There is an application to build python .apps:
http://www.python.org/cgi-bin/moinmoin/BundleBuilder
So if Kent would look into it, it would be much appreciated, both from us developers and I guess from the HT-Mac-community.
It seems that we are right in assuming that the ComboBoxes caused the problem. Posts on the wxPython-list names ComboBoxes as one of the most error pruned functions in wxPython for OS X.
/Oscar
Ill take a look at it later tonight. About the ComboBoxes, is there ways around the problems?
-- kent
On 2004-01-12, at 08.24, Oscar Gustafsson wrote:
Regarding the latest discussion about the Mac version.
There is an application to build python .apps:
http://www.python.org/cgi-bin/moinmoin/BundleBuilder
So if Kent would look into it, it would be much appreciated, both from us developers and I guess from the HT-Mac-community.
It seems that we are right in assuming that the ComboBoxes caused the problem. Posts on the wxPython-list names ComboBoxes as one of the most error pruned functions in wxPython for OS X.
/Oscar _______________________________________________ Mattricks mailing list Mattricks@lists.lysator.liu.se http://lists.lysator.liu.se/mailman/listinfo/mattricks
Either changing to wxSelection (normal drop-down without the ability to type) or maybe reading the follwing:
http://lists.wxwindows.org/cgi-bin/ezmlm-cgi?11:sss:24612:200312:klbchmkfgdl...
The first post is: --- I figured this out from one of my users. If you create a combobox without specifying a value and then call GetValue(), wx coredumps. It works fine on Windows and Linux.
cb=wx.ComboBox(panel, -1, style=wx.CB_DROPDOWN, choices=["1", "2", "3"] print cb.GetValue()
The above coredumps. Adding "" as a parameter after -1 fixes it. ---
You could try adding the following row after row 78 (in the CVS version) in Mattricks/MatchPrediction.py (just after the initialization of the wxComboBox):
cb.SetSelection(3)
this may help, if I read the post correctly.
/Oscar
I tried this in the 0.4 release code and it worked perfectly (I needed a break where stuff works, this came perfect) =)
Is there any plans on a match archive, btw?
-- kent
On 2004-01-12, at 11.14, Oscar Gustafsson wrote:
You could try adding the following row after row 78 (in the CVS version) in Mattricks/MatchPrediction.py (just after the initialization of the wxComboBox):
cb.SetSelection(3)
this may help, if I read the post correctly.
/Oscar _______________________________________________ Mattricks mailing list Mattricks@lists.lysator.liu.se http://lists.lysator.liu.se/mailman/listinfo/mattricks
On Mon, 12 Jan 2004, Kent Karlsson wrote:
I tried this in the 0.4 release code and it worked perfectly (I needed a break where stuff works, this came perfect) =)
That's good news!
It seems like there will be a working Mac version, finally.
Is there any plans on a match archive, btw?
Yes, there is one (or a start of one, depending on what you wants) in the current CVS version. If you would like to try it out, it would be much appreciated. Especially since it is using a bsddb database, which says it has support for Unix and Windows only, although googling may give you the impression that it should work on OS X as well (which makes sense as it is a unix). This should be able to do without dowloading any extra libraries (right Ragnar?).
/Oscar
On Mon, 12 Jan 2004, Oscar Gustafsson wrote:
On Mon, 12 Jan 2004, Kent Karlsson wrote:
I tried this in the 0.4 release code and it worked perfectly (I needed a break where stuff works, this came perfect) =)
That's good news!
Indeed good news!
It seems like there will be a working Mac version, finally.
Is there any plans on a match archive, btw?
Yes, there is one (or a start of one, depending on what you wants) in the current CVS version. If you would like to try it out, it would be much appreciated. Especially since it is using a bsddb database, which says it has support for Unix and Windows only, although googling may give you the impression that it should work on OS X as well (which makes sense as it is a unix). This should be able to do without dowloading any extra libraries (right Ragnar?).
If it indeed does work on Mac OS X, then it is should already be in the python-2.3 distribution. Try starting python and do "import bsddb" to see if it is included.
/Ragnar
On 2004-01-12, at 15.46, Oscar Gustafsson wrote:
It seems like there will be a working Mac version, finally.
Yeah, but there are some issues.. (more below)
Is there any plans on a match archive, btw?
Yes, there is one (or a start of one, depending on what you wants) in the current CVS version. If you would like to try it out, it would be much appreciated. Especially since it is using a bsddb database, which says it has support for Unix and Windows only, although googling may give you the impression that it should work on OS X as well (which makes sense as it is a unix). This should be able to do without dowloading any extra libraries (right Ragnar?).
I checked out the latest cvs and tried it out with the confusing error that bsddb wasnt there. It took me a while to figure out that it was the lib that was missing.
If anyone want bsddb to work on MacOS X 10.3, do this (preceding every cmd with a # due to the linebreaks):
# curl http://undefined.org/python/pimp/bsddb-4.1.6.darwin-6.6- Power_Macintosh.tar.gz > bsddb.tgz # tar xzvf bsddb.tgz # mv ./Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/site- packages/_bsddb.so . # install_name_tool -change /Library/Frameworks/Python.framework/Versions/2.3/Python /System/Library/Frameworks/Python.framework/Versions/2.3/Python _bsddb.so # mv _bsddb.so /System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/
Found the idea here: http://divmod.org/users/mailman.twistd/pipermail/divmod-dev/2003- November/000150.html
Is the fetching of the fixtures supposed to work? I get this:
Traceback (most recent call last): File "/Users/kentk/Projects/OpenSource/Mattricks/Mattricks/LeaguePage.py", line 195, in OnDataUpdate self.Update(True) File "/Users/kentk/Projects/OpenSource/Mattricks/Mattricks/LeaguePage.py", line 276, in Update self.fixtures = Fixtures.Fixtures(current_league, current_season, force_update) File "/Users/kentk/Projects/OpenSource/Mattricks/Mattricks/Fixtures.py", line 35, in __init__ self.__extract_data(llu_id, season, force_update) File "/Users/kentk/Projects/OpenSource/Mattricks/Mattricks/Fixtures.py", line 39, in __extract_data data = XMLParse.get(["league fixtures", llu_id, season], force_update = force_update) File "/Users/kentk/Projects/OpenSource/Mattricks/Mattricks/XMLParse.py", line 362, in get return get(identifier, paths, normal, display, userid, True, False) File "/Users/kentk/Projects/OpenSource/Mattricks/Mattricks/XMLParse.py", line 373, in get values = databases.get_all(nuserid, xml_info[2], idstr) File "/Users/kentk/Projects/OpenSource/Mattricks/Mattricks/Database.py", line 72, in get_all data[key] = db.get(key) AttributeError: _DBWithCursor instance has no attribute 'get'
On Mon, 12 Jan 2004, Kent Karlsson wrote:
I checked out the latest cvs and tried it out with the confusing error that bsddb wasnt there. It took me a while to figure out that it was the lib that was missing.
If anyone want bsddb to work on MacOS X 10.3, do this (preceding every cmd with a # due to the linebreaks):
# curl http://undefined.org/python/pimp/bsddb-4.1.6.darwin-6.6- Power_Macintosh.tar.gz > bsddb.tgz # tar xzvf bsddb.tgz # mv ./Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/site- packages/_bsddb.so . # install_name_tool -change /Library/Frameworks/Python.framework/Versions/2.3/Python /System/Library/Frameworks/Python.framework/Versions/2.3/Python _bsddb.so # mv _bsddb.so /System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/
Found the idea here: http://divmod.org/users/mailman.twistd/pipermail/divmod-dev/2003- November/000150.html
Seems like Apple forgot the _bsddb.so in their python-2.3 dist.
Is the fetching of the fixtures supposed to work? I get this:
Traceback (most recent call last): [..snip..] "/Users/kentk/Projects/OpenSource/Mattricks/Mattricks/Database.py", line 72, in get_all data[key] = db.get(key) AttributeError: _DBWithCursor instance has no attribute 'get'
This should work, but since it is unnecessary to get the value again you should change that row (72) in Database.py to read "data[key] = value".
/Ragnar
On 2004-01-12, at 16.11, Ragnar Ouchterlony wrote:
Traceback (most recent call last): [..snip..] "/Users/kentk/Projects/OpenSource/Mattricks/Mattricks/Database.py", line 72, in get_all data[key] = db.get(key) AttributeError: _DBWithCursor instance has no attribute 'get'
This should work, but since it is unnecessary to get the value again you should change that row (72) in Database.py to read "data[key] = value".
Works perfectly. =) Will someone commit these changes and in MatchPrediction.py?
Btw, when I fetch all the matches, it seem to nest a lot of windows. I get a added shadow for every downloading/updating window and pretty fast its solid black.
-- kent
On Mon, 12 Jan 2004, Kent Karlsson wrote:
Works perfectly. =) Will someone commit these changes and in MatchPrediction.py?
I will commit that change when I do my next commit spree. :-) I can commit the changes to MatchPrediction.py too if you give me a diff on what you did.
Btw, when I fetch all the matches, it seem to nest a lot of windows. I get a added shadow for every downloading/updating window and pretty fast its solid black.
There shouldn't be more than two progress dialogs at the same time. The current code is a little clumsy there. There is one progress dialog for the whole process of downloading all match details and match lineups. Also there will be one progress dialog for the process of each file (since this is done automatically). The second one should probably be switched off when downloading all files at the same time. Something for you to do Oscar?
/Ragnar
On 2004-01-12, at 16.25, Ragnar Ouchterlony wrote:
On Mon, 12 Jan 2004, Kent Karlsson wrote:
I will commit that change when I do my next commit spree. :-) I can commit the changes to MatchPrediction.py too if you give me a diff on what you did.
Here you are.
Btw, I noticed that in the League table, the slider has Match 1 in the bottom and match 14 in the top, but the lowered numbered match will always be displayed on top.
-- kent
On Mon, 12 Jan 2004, Kent Karlsson wrote:
Here you are.
Thanks!
Btw, I noticed that in the League table, the slider has Match 1 in the bottom and match 14 in the top, but the lowered numbered match will always be displayed on top.
That's strange. It does not show up that way for me. I have match 1 at the top and 14 at the bottom. Either that is some strangeness with the MacOSX version of wxPython or I have done something strange. :)
/Ragnar
On 2004-01-12, at 16.42, Ragnar Ouchterlony wrote:
That's strange. It does not show up that way for me. I have match 1 at the top and 14 at the bottom. Either that is some strangeness with the MacOSX version of wxPython or I have done something strange. :)
Hmm, I think it would be a interesting test to have the linux and MacOSX version side by side and check how they compare. =)
In the Lineup Selection, am I supposed to be able to change the lineup myself, or is it only to view the computer generated lineup? Also, is there some easy way to change the formation? Yeah, and computer is spelled wrong on the button.
-- kent
On Mon, 12 Jan 2004, Kent Karlsson wrote:
Hmm, I think it would be a interesting test to have the linux and MacOSX version side by side and check how they compare. =)
It would, I don't have a Mac to try on, so I can't do that. You can install linux on a Mac too. :-)
In the Lineup Selection, am I supposed to be able to change the lineup myself, or is it only to view the computer generated lineup?
It's only to get a computer generated lineup.
Also, is there some easy way to change the formation?
No, not yet. You're not the first one to ask. I *will* do this before next version. I keep getting the wrong formation too. It's unfortunately not good enough when selecting the formation itself (for now at least).
Yeah, and computer is spelled wrong on the button.
That one I fixed at once.
/Ragnar
On 2004-01-12, at 17.10, Ragnar Ouchterlony wrote:
It would, I don't have a Mac to try on, so I can't do that. You can install linux on a Mac too. :-)
I don't think that will happen, but I might install wxPython on a linuxmachine at work and try it out over X11.
No, not yet. You're not the first one to ask. I *will* do this before next version. I keep getting the wrong formation too. It's unfortunately not good enough when selecting the formation itself (for now at least).
I tried to read the code, but I didn't really get what was happening, partly due to Python but not entirely. =)
Have you seen this: http://www.databased.at/hattrick/x-ray/index.asp? I also lack a way to go directly from a unplayed match to the MatchPrediction using the info from both teams last match. =)
-- kent
On Mon, 12 Jan 2004, Kent Karlsson wrote:
On 2004-01-12, at 17.10, Ragnar Ouchterlony wrote:
It would, I don't have a Mac to try on, so I can't do that. You can install linux on a Mac too. :-)
I don't think that will happen, but I might install wxPython on a linuxmachine at work and try it out over X11.
I am glad as long as there is people with access to Macs testing the program there.
I tried to read the code, but I didn't really get what was happening, partly due to Python but not entirely. =)
The TeamSelecting code is rather a mess unfortunately (which might be why I'm reluctant to add that code).
Have you seen this: http://www.databased.at/hattrick/x-ray/index.asp?
Yes, a weatherforecast might be good. I don't know where you get that information out of Hattrick though.
I also lack a way to go directly from a unplayed match to the MatchPrediction using the info from both teams last match. =)
I have just been working on something like that. I just checked in it too. On the match detail page you can now copy the ratings to the match prediction page. Not as good as your variant, but good enough for now.
/Ragnar
On Mon, 12 Jan 2004, Ragnar Ouchterlony wrote:
There shouldn't be more than two progress dialogs at the same time. The current code is a little clumsy there. There is one progress dialog for the whole process of downloading all match details and match lineups. Also there will be one progress dialog for the process of each file (since this is done automatically). The second one should probably be switched off when downloading all files at the same time. Something for you to do Oscar?
Now checked in!
/Oscar
I changed my slackware Linux in for a LFS system and I wanted to start Mattricks again (after installing it), I got the following error
No module named _bsddb
I tried every solution, but couldn't find one
greets bart
Oscar Gustafsson wrote:
On Tue, 13 Jan 2004, bart van der hall wrote:
No module named _bsddb
As far as I know, bsddb should be included in the standard python distribution. Which version are you using? If < 2.3 I would suggest upgrading as a first thing to do.
/Oscar _______________________________________________ Mattricks mailing list Mattricks@lists.lysator.liu.se http://lists.lysator.liu.se/mailman/listinfo/mattricks
I'm using 2.3.3
On Tue, 13 Jan 2004, bart van der hall wrote:
Oscar Gustafsson wrote:
As far as I know, bsddb should be included in the standard python distribution. Which version are you using? If < 2.3 I would suggest upgrading as a first thing to do.
I'm using 2.3.3
I don't know what a LFS system is, but the python package in your system doesn't include all that should be included. Is there any python2.3-bsddb package for you? Do you have any db4.1 packages that need to be installed that isn't?
/Ragnar
Ragnar Ouchterlony wrote:
On Tue, 13 Jan 2004, bart van der hall wrote:
Oscar Gustafsson wrote:
As far as I know, bsddb should be included in the standard python distribution. Which version are you using? If < 2.3 I would suggest upgrading as a first thing to do.
I'm using 2.3.3
I don't know what a LFS system is, but the python package in your system doesn't include all that should be included. Is there any python2.3-bsddb package for you? Do you have any db4.1 packages that need to be installed that isn't?
/Ragnar _______________________________________________ Mattricks mailing list Mattricks@lists.lysator.liu.se http://lists.lysator.liu.se/mailman/listinfo/mattricks
got it working again, I forgot to add some things
On Tue, 2004-01-13 at 15:04, bart van der hall wrote:
Ragnar Ouchterlony wrote:
On Tue, 13 Jan 2004, bart van der hall wrote:
Oscar Gustafsson wrote:
As far as I know, bsddb should be included in the standard python distribution. Which version are you using? If < 2.3 I would suggest upgrading as a first thing to do.
I'm using 2.3.3
I don't know what a LFS system is, but the python package in your system doesn't include all that should be included. Is there any python2.3-bsddb package for you? Do you have any db4.1 packages that need to be installed that isn't?
/Ragnar _______________________________________________ Mattricks mailing list Mattricks@lists.lysator.liu.se http://lists.lysator.liu.se/mailman/listinfo/mattricks
got it working again, I forgot to add some things
Mattricks mailing list Mattricks@lists.lysator.liu.se http://lists.lysator.liu.se/mailman/listinfo/mattricks
which things are that, because i cant get the cvs version to work (even with an update of python to the newest version because of the bsddb module)
mattricks@lists.lysator.liu.se
-
bart van der hall -
Kent Karlsson -
Kent Karlsson -
Oscar Gustafsson -
Ragnar Ouchterlony -
Zazell Kroeske