now used mv. tested and works for me.

[Mar2zz]

3rd time lol. need to get grips with git, sorry. I deleted my repo for it haha

also added a warning to create a settings.py if it is not found (see database.py)

[mrkipling]

Looks good - will have a proper look as soon as I can, hopefully over the next couple of days, and see about getting this merged in.

[Mar2zz]

Ok. Think it will give some trouble merging it in other branches because of the renaming of Maraschino.py.

My other pullrequest, about the commandline stuff, will finish it when you have decided about this one (or it will conflict again).

Also, on xbmc.org people requested about the database being created by default in the maraschino's root. That's an idea?

Also, port 5000 is conflicting with Couchpotato and with Synology DSM. Could you consider changing the default ports in settings_example.py?

[gugahoi]

I also believe creating at root by default is best. Dont know if you guys realise this but we can use relative paths in the DATABASE variable. So by default I think we should have './maraschino.db' and eliminate the need to creating a settings.py. The settings should be overwritten with command line args if desired (--dbpath --port --cherrypyport --user --pw are just some ideas)

[N3MIS15]

Tested on unraid vm without dependencies installed and worked flawlessly

[mrkipling]

Mar2zz: that's just the development port (for us developers, and people testing it out quickly before committing to a proper install. The CherryPy method uses 7000 by default (although you can change this) and if you configure Apache then you can put it on whatever port you want.

[mrkipling]

Okay, so I've just reviewing this pull request and it all appears to work fine. Good job!

I think that including all of these dependencies is the way to go, so I've pushed this straight to master. Let's see how people find it, and if they have any problems we should make fixing these high priority. Worst case scenario where we've borked everything, we can just revert the single commit where I merged it in and do so some more work on it.

I'll post on of the forums explaining the benefits, and that people need to update their WSGI files (or however they're serving Maraschino) to point to Maraschino.py instead of maraschino.py.

Will also update the official website with new installation instructions.

Thanks for the awesome contribution, Mar2zz!

[mrkipling]

Would just like to add that I haven't tested it super-thoroughly, but the only thing I needed to do was pull, make the simple change in my WSGI file from maraschino to Maraschino, restart Apache, and bam - everything working perfectly. I also uninstalled the dependencies from my server beforehand.

As mentioned above, let's listen to our users and see what issues they find (if any). I'm pretty sure that it will be mostly fine though.

[Mar2zz]

Op 05-01-12 21:21, Bradley Abrahams schreef:

Would just like to add that I haven't tested it super-thoroughly, but the only thing I needed to do was pull, make the simple change in my WSGI file from maraschino to Maraschino, restart Apache, and bam - everything working perfectly. I also uninstalled the dependencies from my server beforehand.

As mentioned above, let's listen to our users and see what issues they find (if any). I'm pretty sure that it will be mostly fine though.

---

Reply to this email directly or view it on GitHub:
#32 (comment)
Great. This lowers install-anxiety a lot for newcomers!

About the cherrypy, will show tomorrow a new commandline pull request,
which enables cherrypy inside Maraschino.py. This makes choosing a
servertype easy. Start with Maraschino.py --quiet or something and start
cherrypy-server, or start with --develop and start flask-server. Dunno
which best. For users not coding, the best behaviour is perhaps to
always start the cherrypy by default, and for dev's with the --develop
parameter to start flask? An idea to go by? What's your opinion on this?
Using opt-parse it's very easy to write, opt-parse is a default
python-package so it doesn't need to be in lib. CouchPotato uses it too
(see CouchPotato.py for examples). Sickbeard uses own written argparser
version with a for loop, can be used too (see SickBeard.py for examples).

[mrkipling]

@Mar2zz: I look forward to trying out your command line pull request, sounds intriguing. I haven't seen any mention of Apache in your comment, so my only 2 cents would be to make sure that it doesn't break Apache (I can test this if you like as I prefer Apache over CherryPy).

[Mar2zz]

I can test apache too. I have two apache-servers running, so no problem.