You are not logged in.
Here you go.
Thanks, I'm rewriting almost the whole program at this point (It will no longer share the majority of it's code with GtkPacman), and I'll be using ConfigParser to load the pacman config file now, instead of a line by line setup. I'm also considering including a seperate Wakka config file to let you override your default pacman settings, but that's a last ditch if-you-can't-fix'it-hack-it option. I'd obviously rather have it working.
So now, my todo list:
- Seperate the Gui and backend
- A new WxPython Gui
- Config Dialog
- Remove VTE dependency, use terminal with -e
- AUR support via... something. Still working on it.
#Edit: I didn't realize how much gui code was jumbled up with the backend functions... This could take awhile, but I'll make sure to post when I'm done. Thanks for the support folks!
Last edited by minasmorath (2011-03-13 23:00:21)
Offline
You might even be better off starting from scratch, and rather than wrapping pacman, build off of pyalpmm.
coder formally known as EnvoyRising
Offline
Hey everyone,
Sorry I haven't posted in awhile, I've been in the process of a move, but I'm still hard at work on the next release (0.5 - New Gui, cleaner code, config, possibly pyalpmm after I do some more research on it). It's going to take a bit longer than the original flurry of updates, but fret not, I haven't lost sight of the project, I'm just a little overwhelmed everywhere else at the moment. So, until the next release, happy travels, and as always, bug reports are much appreciated.
Mitch
Offline
Grats on the move and take your time. I couldn't get pyalpmm to build myself, but like the direction it was heading.
coder formally known as EnvoyRising
Offline
So nice, I preferred to install from terminal 'cause it was faster but, with this nice program, I can see what a package is when there is an update for it, so I have not to check it on this website.
Two requests:
Support for AUR?
File list for non installed packages?
Anyway, good job indeed!
Offline
Just tried it!
Support for AUR would be nice indeed but I think it already looks and feels very nice.
Offline
Just tried it!
Support for AUR would be nice indeed but I think it already looks and feels very nice.
Yes it is but it would make it from perfect to plusperfect
Offline
Hey there Archers,
Still hard at work on the GUI rewrite, and figured I would keep you all updated appropriately. I'm moving everything over to WxPython and doing away with VTE, so the default way to execute pacman commands from wakka will be 'xterm -e' from 0.5 on up, but you will be able to use your favorite terminal via config file.
AUR support will be provided via a wrapper, I've been digging into the JSON interface and even took a peek at AUR2, but it's not something I have the time or expertise to implement right now, so I'm just going to get it working as opposed to promising perfect features in the future. I'm almost settled on forking my own version of aurora for the distant future, but I'm still open to persuasion there. For now I think packer will do what I need, as it's tiny, fast, and easy to install.
File list for non installed packages?
This is actually quite tricky. I'm pulling package information from /var/lib/pacman/sync/$repo/$package-$ver/desc which contains only rudimentary package information for things that aren't installed, it doesn't know about the files in the package until it's expanded into the system.
I preferred to install from terminal 'cause it was faster
Actually, me too. It's kinda sad that I don't even use my own program unless I'm testing for bugs or doing some routine stress testing on new code, but hey, I'm havin a ton of fun developing it and apparently people are using it, so that's good enough for me.
Until next time (Which will hopefully be the 0.5 release), happy travels!
Mitch
Offline
Until next time (Which will hopefully be the 0.5 release), happy travels!
Mitch
have a nice day!
We really appreciate your work, and in my opinion, your program reduced a little my nerdness: you know, terminal-addiction is somewhat alienating
Offline
Hey gang,
Well, I seem to have a slight problem. The latest pacman update has changed (slightly) the database format yet again, so I will be working to get a bugfix up that brings Wakka into this new and wonderful world of changes. So, yet again, the wxWakka build goes on the backburner. Until I figure out all the changes I have to make, You can still use the program, but the summary and files panel won't display information for local packages.
Thanks for hanging in there,
Mitch
Offline
It's a nice project, but i prefer the command line.
Offline
Posted a patch but my router is somehow preventing particular connections. You can check the issues page though. Doesn't add any features, but makes wakka more pythonic in a few areas, specifically packaging (python setup.py vs ./install.sh)
coder formally known as EnvoyRising
Offline
Thank you,
Between two jobs and University I'm really lacking the time I need to keep Wakka up to date, but as we move into summer I'll have more time to work on it. The wxWidgets gui is pretty much done, and the backend rehaul is somewhere around 70% last time I checked (last weekend I tinkered for a couple hours), and I'll work on integrating your patches ASAP. The main reason I ditched setup.py as the install method is my noobishness. I don't 100% understand its intricacies, and bash was just easier to get the project off the ground. In the long run though I think being as pythonic as possible will have the most benefit.
Thank you everyone for being patient, and thanks a million for the patch(es),
Mitch
PS - Aur is still as complex as ever, but it's coming along.
Offline
Hey there Archers!
It took almost 2 weeks, but I _finally_ got a couple hours today to get Wakka back up to date, and to keep this short (I unfortunately still have a real life... blah) it's alive! I've updated the google code page and provided a new pkgbuild in the AUR. So as always, if you find any errors, don't hesitate to report them.
Until next time,
Mitch
EDIT: I should note that there are some small issues with the way updates are listed, but it still hits them all in the final part of the update routine.
Last edited by minasmorath (2011-04-23 20:37:00)
Offline
Hello Forum,
I am flagging Wakka out of date. Until further notice it is deprecated software. I've been working on, and am still working on, a new release that works with the repo.db files now utilised by libalpm, but I've hit a brick wall. Anyone with information please email me or post in this thread.
Thank you for all your support,
Mitch
Offline
If you can give details of the issues you are having, I might be able to make suggestions on how to fix them.
Offline
Hey Allan,
Basically, there used to be a directory hierarchy in '/var/lib/pacman/sync' that had a bunch of packages ala the ones in '/var/lib/pacman/local' but now there are just a few files, each one named after a repo in the pacman.conf file. I personally have core.db, community.db, and extra.db to work with. Problem for me is that they're not in plain text, which is what I've been reading thus far. I have no idea how to grab the data from them.
Thanks mate,
Mitch
Offline
They are plain tarballs.
Offline
Python's gzip module should make short work of this:
http://www.doughellmann.com/PyMOTW/gzip/index.html
Or if you prefer the official docs:
http://docs.python.org/library/gzip.html
In short:
import gzip
with gzip.open('community.db', 'rb') as gzip_file:
gzip_file.read()
Obviously you can use os.listdir to get the different database names and do something more useful inside the context manager.
edit:
BTW, if you aren't sure what kind of file something is, you can use the file command:
[edwin@enigma sync]$ file community.db
community.db: gzip compressed data, from Unix, last modified: Fri Jul 22 14:56:39 2011
[edwin@enigma sync]$
Last edited by aspidites (2011-07-24 06:37:15)
coder formally known as EnvoyRising
Offline
Holy crap I got a bugfix done! Wakka is going to be slow to start up after database updates but should run just fine afterwards! Muwahahaha!
Python's gzip module should make short work of this:
http://www.doughellmann.com/PyMOTW/gzip/index.htmlOr if you prefer the official docs:
http://docs.python.org/library/gzip.htmlIn short:
import gzip with gzip.open('community.db', 'rb') as gzip_file: gzip_file.read()
Obviously you can use os.listdir to get the different database names and do something more useful inside the context manager.
edit:
BTW, if you aren't sure what kind of file something is, you can use the file command:[edwin@enigma sync]$ file community.db community.db: gzip compressed data, from Unix, last modified: Fri Jul 22 14:56:39 2011 [edwin@enigma sync]$
Thanks for this, I had given up on finding a more elegant solution than what I did in the bugfix (extract the db files before program startup... bleh) and had basically given up hope, especially after the project got put on the 'abandoned' list on the wiki. But nonetheless, I'm still chugging along, and although the full, proper fix is going to take awhile, I'm determined to get it out the door so I can have some peace of mind.
Thanks for hanging in there guys!
Mitch
Offline
Hows Wakka getting on? No posts for some time.
I read through the forum posts and then decided to give it a try – great stuff!
The only problem so far (I admit I'm more admiring it than using it in the past 5 minutes since install) was the terminal output -
(wakka-package-manager:2612): libglade-WARNING **: unknown attribute `swapped' for <signal>.
Is this anything to worry about?
Offline