Re: pyZui - anyone know about this?
Hi, Yes, the toolkit used is PyQt. The ZUI is implemented using a simple QPainter, and employs pyramidal tiling for efficiency (I haven't used any Qt/KDE voodoo in this regard). I'm using Gnome at the moment, but it should work just as well on KDE. Web pages are rendered using QtWebKit, and PDF with the pdftoppm utility. The project is opensource (GPLv2), but just hasn't been published yet :) . I'll try to make a release over the next few days, and I'll post a link here when I do. -- David Roberts http://da.vidr.cc/ On Dec 15, 10:33 am, Donn wrote: > On Tuesday 15 December 2009 01:43:52 David Boddie wrote:> I managed to catch > his address and sent him a message saying that people > > were discussing PyZUI in this thread. > > Oooh. Sits,fidgets and waits. I want my socks back! (OP) :D > > \d > -- > \/\/ave: donn.in...@googlewave.com > home:http://otherwise.relics.co.za/ > 2D vector animation :https://savannah.nongnu.org/projects/things/ > Font manager :https://savannah.nongnu.org/projects/fontypython/ -- http://mail.python.org/mailman/listinfo/python-list
Re: pyZui - anyone know about this?
> > and employs pyramidal tiling for efficiency > > \me ... time to hit Wikipedia :) It involves scaling an image to various resolutions, and partitioning them into fixed-size tiles. It's roughly the same technique used by Google Maps/Earth. > It is very cool, but I would inject a note of caution here: I'd a hate a zui > to become a case of "hunt-the-zoom." A link is a link. They already work very > well, click and it goes to the page. > I find the notion of minute "hot" areas to be a little obscure -- Quick! Zoom > into the last full-stop, it's a whole word in there! > What I would enjoy is when you click a link - it zooms into the sub-page so > you get a feeling of traversal. Back buttons would zoom out again. Add to that > a kind of birds'-eye view of one's history (like a thumbnails node-graph of > some kind) and it would be perfect! Sure, it was just a quick mockup of a potential application. A proper implementation would probably have more sophisticated features such as that. > This aspect reminds me of the Red Dwarf episode "Back to Reality", in > which Rimmer is criticised for not finding information contained in a > microdot hidden in the dot on the 'i' of his name on a swimming > certificate. Haha, true. > ZUIs are useful for particular types of data - images & mapping > especially - but I'd hate to have to navigate my desktop using its > approach. Obviously there will be some applications that suit more traditional GUIs better than ZUIs, just like there's plenty of applications more suited to the command-line than a GUI. After all, things such as the web and the desktop metaphor came into being long before ZUIs. On Dec 16, 1:09 pm, alex23 wrote: > Donn wrote: > > I find the notion of minute "hot" areas to be a little obscure -- Quick! > > Zoom > > into the last full-stop, it's a whole word in there! > > This aspect reminds me of the Red Dwarf episode "Back to Reality", in > which Rimmer is criticised for not finding information contained in a > microdot hidden in the dot on the 'i' of his name on a swimming > certificate. > > ZUIs are useful for particular types of data - images & mapping > especially - but I'd hate to have to navigate my desktop using its > approach. -- http://mail.python.org/mailman/listinfo/python-list
Re: pyZui - anyone know about this?
PyZUI 0.1 has been released: http://da.vidr.cc/projects/pyzui/ On Dec 15, 12:29 pm, David Roberts wrote: > Hi, > > Yes, the toolkit used is PyQt. The ZUI is implemented using a simple > QPainter, and employs pyramidal tiling for efficiency (I haven't used > any Qt/KDE voodoo in this regard). I'm using Gnome at the moment, but > it should work just as well on KDE. Web pages are rendered using > QtWebKit, and PDF with the pdftoppm utility. > > The project is opensource (GPLv2), but just hasn't been published > yet :) . I'll try to make a release over the next few days, and I'll > post a link here when I do. > > -- > David Robertshttp://da.vidr.cc/ > > On Dec 15, 10:33 am, Donn wrote: > > > > > On Tuesday 15 December 2009 01:43:52 David Boddie wrote:> I managed to > > catch his address and sent him a message saying that people > > > were discussing PyZUI in this thread. > > > Oooh. Sits,fidgets and waits. I want my socks back! (OP) :D > > > \d > > -- > > \/\/ave: donn.in...@googlewave.com > > home:http://otherwise.relics.co.za/ > > 2D vector animation :https://savannah.nongnu.org/projects/things/ > > Font manager :https://savannah.nongnu.org/projects/fontypython/ -- http://mail.python.org/mailman/listinfo/python-list
Re: pyZui - anyone know about this?
> /home/fetchinson/pyzui/pyzui/tilestore.py:22: DeprecationWarning: the > sha module is deprecated; use the hashlib module instead > import sha Yeah, I'd noticed that. It's fixed in the repository now. On Dec 16, 10:55 pm, Daniel Fetchinson wrote: > > PyZUI 0.1 has been released: > > >http://da.vidr.cc/projects/pyzui/ > > Cool, thanks very much! > > I'm using python 2.6 these days and noticed that you use the sha > module which makes py2.6 spit out a deprecation warning: > > /home/fetchinson/pyzui/pyzui/tilestore.py:22: DeprecationWarning: the > sha module is deprecated; use the hashlib module instead > import sha > > It's no big deal but if you want to be future proof maybe you can > switch to hashlib for py2.6 and stay with sha for py2.5 and before (a > try/except block would suffice). > > Cheers, > Daniel > > -- > Psss, psss, put it down! -http://www.cafepress.com/putitdown -- http://mail.python.org/mailman/listinfo/python-list
Re: pyZui - anyone know about this?
> Personally I see a merging of normal app windows and a zui: some kind of new > window manager. Have you seen Eagle Mode[1]? [1] http://eaglemode.sourceforge.net/ On Dec 17, 5:14 pm, Donn wrote: > On Wednesday 16 December 2009 07:03:19 David Roberts wrote:> It involves > scaling an image to various resolutions, and partitioning > > them into fixed-size tiles. It's roughly the same technique used by > > Google Maps/Earth. > > Thanks. That gives me something to go on. Wikipedia didn't like my search > terms. > > > > ZUIs are useful for particular types of data - images & mapping > > > especially - but I'd hate to have to navigate my desktop using its > > > approach. > > Ever since Corel Draw in the 90's zoomed into my life I have been in love with > the idea of an endless canvas that makes me feel like a satellite on a bungee > cord. I think it would fit the desktop very well. > > Personally I see a merging of normal app windows and a zui: some kind of new > window manager. > If I planned it out it would look something like this: > Your apps all run as they do now*, but they live on this endless plain. > Perhaps it can be divided up into 'zones' or 'galaxies' or something. I would > have a 'hyperspace' or 'hyperlink' or 'jump' facility (like alt-tab, I guess) > to make transits from one custom-defined area to another quick. > > I would have a home position for the view -- like Inkscape does in terms of > show all, zoom to selected, zoom to last, etc. > > I would have rules about traversing. Things like file-managers need some kind > of static display - like the bread crumbs and up, back, home etc. > > Each app would only be active when 'locked-in', beyond that it's a bitmap of > the last paint. You could drag apps around when you zoom out, and you can > resize them at any time too. > (Just imagine OOCalc in a zui! Super/Capslock and mouse wheel for scroll/pan) > > The other cool idea I had was to (handwavium here) graphically convey the > notion of pipes and import/export between apps. Also between any nodes across > the Universe of the zui. Perhaps a special 'node view' that overlays and shows > all the conduits between them -- sharp where your mouse is, faded away from > that so the whole thing is not too complex. > Imagine the flow from Inkscape to Gimp and back. Instead of File -> Export and > then File -> Import, you connect pipes along the side of each app. > Inkscape, [save selected as png (properties preset)] goes to Gimp [import to > layers by names (a script perhaps)] Now as you work in Inkscape and hit a > hotkey, all your selected vectors are sent to Gimp which reacts as if you were > there and places the new pngs into layers. > This can work both ways and between multiple programs. Mix-in Blender and > Scribus and Lyx and some grep and a loop or two and some imagemagick... > > Ah, I better stop. I can ramble on sometimes :) > > *I have many issues with the endless variety of re-invented wheels afa gui > toolkits go. This is another whole can of shai-Hulud... > > I wrote some stuff about this a while back, if anyone wants to be put to > sleep:http://otherwise.relics.co.za/wiki/Particles/DreamDesignApp/ > :) > > \d > > -- > \/\/ave: donn.in...@googlewave.com > home:http://otherwise.relics.co.za/ > 2D vector animation :https://savannah.nongnu.org/projects/things/ > Font manager :https://savannah.nongnu.org/projects/fontypython/ -- http://mail.python.org/mailman/listinfo/python-list
OverflowError in RLock.acquire()
Hi, I'm trying to port a Python application to Windows, and I'm getting the following error (which did not occur when running on Linux): Exception in thread Thread-4: Traceback (most recent call last): File "C:\Python26\lib\threading.py", line 525, in __bootstrap_inner self.run() File "C:\Documents and Settings\David\My Documents\pyzui\pyzui\tileprovider.py", line 97, in run self.__tilecache[tile_id] = Tile(tile) File "C:\Documents and Settings\David\My Documents\pyzui\pyzui\tilecache.py", line 165, in __setitem__ with self.__lock: File "C:\Python26\lib\threading.py", line 115, in acquire me = current_thread() File "C:\Python26\lib\threading.py", line 803, in currentThread return _active[_get_ident()] OverflowError: can't convert negative value to unsigned long Where __lock is an RLock object. The error only occurs for a single class (which is derived from the TileProvider class in tileprovider.py, which in turn is derived from threading.Thread), which would lead me to believe that there's an error in my code, but the traceback doesn't help much, and I haven't been able to find any similar problems with google. Any help would be appreciated. -- David Roberts http://da.vidr.cc/ -- http://mail.python.org/mailman/listinfo/python-list
Re: OverflowError in RLock.acquire()
I forgot to mention, Python version is 2.6.2. -- David Roberts http://da.vidr.cc/ On Fri, Jul 24, 2009 at 14:27, David Roberts wrote: > Hi, > > I'm trying to port a Python application to Windows, and I'm getting > the following error (which did not occur when running on Linux): > > Exception in thread Thread-4: > Traceback (most recent call last): > File "C:\Python26\lib\threading.py", line 525, in __bootstrap_inner > self.run() > File "C:\Documents and Settings\David\My > Documents\pyzui\pyzui\tileprovider.py", line 97, in run > self.__tilecache[tile_id] = Tile(tile) > File "C:\Documents and Settings\David\My > Documents\pyzui\pyzui\tilecache.py", line 165, in __setitem__ > with self.__lock: > File "C:\Python26\lib\threading.py", line 115, in acquire > me = current_thread() > File "C:\Python26\lib\threading.py", line 803, in currentThread > return _active[_get_ident()] > OverflowError: can't convert negative value to unsigned long > > Where __lock is an RLock object. > > The error only occurs for a single class (which is derived from the > TileProvider class in tileprovider.py, which in turn is derived from > threading.Thread), which would lead me to believe that there's an > error in my code, but the traceback doesn't help much, and I haven't > been able to find any similar problems with google. > > Any help would be appreciated. > > -- > David Roberts > http://da.vidr.cc/ > -- http://mail.python.org/mailman/listinfo/python-list
Re: OverflowError in RLock.acquire()
Done: http://bugs.python.org/issue6562 -- David Roberts http://da.vidr.cc/ On Fri, Jul 24, 2009 at 20:24, Gabriel Genellina wrote: > En Fri, 24 Jul 2009 01:27:10 -0300, David Roberts escribió: > >> I'm trying to port a Python application to Windows, and I'm getting >> the following error (which did not occur when running on Linux): >> >> Exception in thread Thread-4: >> File "C:\Python26\lib\threading.py", line 803, in currentThread >> return _active[_get_ident()] >> OverflowError: can't convert negative value to unsigned long > > Looks like a bug in the thread module - you should report it at > http://bugs.python.org > > -- > Gabriel Genellina > > -- > http://mail.python.org/mailman/listinfo/python-list > -- http://mail.python.org/mailman/listinfo/python-list
Re: idiom for list looping
To the best of my knowledge the second way is more pythonic - the first is a little too reminiscent of C. A couple of notes: - you don't need the parentheses around "i, e" - if you were going to use the first way it's better to use xrange instead of range for iteration -- David Roberts http://da.vidr.cc/ On Wed, Jul 29, 2009 at 19:02, superpollo wrote: > hi clp. > > i want to get from here: > > nomi = ["one", "two", "three"] > > to here: > > 0 - one > 1 - two > 2 - three > > i found two ways: > > first way: > > for i in range(len(nomi)): > print i, "-", nomi[i] > > or second way: > > for (i, e) in enumerate(nomi): > print i, "-", e > > which one is "better"? is there a more pythonic way to do it? > > bye > > ps: > > m...@192.168.1.102:~/test$ python -V > Python 2.3.4 > m...@192.168.1.102:~/test$ uname -a > Linux fisso 2.4.24 #1 Thu Feb 12 19:49:02 CET 2004 i686 GNU/Linux > m...@192.168.1.102:~/test$ > -- > http://mail.python.org/mailman/listinfo/python-list > -- http://mail.python.org/mailman/listinfo/python-list