why duplicate pastebinit?

[Rolf Leggewie]

I don't quite understand why you guys felt the need to reinvent the wheel known as pastebinit that's been a part of Ubuntu since at least since hardy.

Wouldn't it be better to collaborate?

[Rolf Leggewie (r0lf)]

If there is anything in pastebinit that you find lacking, please open a ticket. Let's cooperate.

http://packages.ubuntu.com/search?keywords=pastebinit

[Rolf Leggewie (r0lf)]

opening task in Ubuntu so that I will find this ticket next time I erroneously think I haven't opened it, yet ;-)

[Alessio Treglia (quadrispro)]

Rolf, thanks!

Making nautilus-pastebin relies on pastebinit would be great, but I think we need a shared and interface independent python module: what do you think of that split? Plus, some pastebin services now provide both the HTTP and XML-rpc methods (the latter seem more safe and secure), so maybe we should provide different objects (as per Strategy pattern) to select the proper method at runtime.

nautilus-pastebin also tries to associate a MIME type to the pastebin-specific format, what do think about merging our efforts and looking for a GUI-independent way to do that?

[Rolf Leggewie (r0lf)]

Alessio, thank you for your comment. I'm all for cooperation and creating a better experience for everyone, both devs and users.

I'm merely the packager and an interested party. Stéphane would be the one to talk to about making changes upstream and adding you to the pastebinit devs. You opened an upstream pastebinit task which I think is invalid but beneficial nonetheless ;-) to get it on his radar so he has a chance to comment. I wouldn't be surprised if the answer was "patches welcome".

I did open bug 580848 because I felt that was something that nautilus-pastebin was offering and pastebinit currently doesn't. I'm also all for the separation of backend and GUI/nautilus integration. I haven't had a look, but I was hoping that nautilus-pastebin was modular enough so that we could just shamelessly steal that part of your code and use it with pastebinit ;-)

I'd be happy to test any changes/hacks you may consider interesting. If they work well, I'd also consider to include a patch in Debian irrespective of progress upstream (although I would think that Stéphane is usually quicker to react than me).

[Rolf Leggewie (r0lf)]

I would for this to be come a single upstream project, packaging on Debian would create two interdependent debs, one for the CLI, one for the GUI/nautilus integration.

[Alessio Treglia (quadrispro)]

Subscribing Stèphan, I'd like to know what he thinks about that.

[Stéphane Graber (stgraber)]

Hi Alessio, sorry for the delay.
I think it might indeed be interesting to move most of the current code of pastebinit to a python module, then have pastebinit be the CLI frontend for that. Then we can have nautilus-pastebin and others to use that same module and avoid code duplication.

Sadly, I only allocate a very very small part of my free time to pastebinit so I doubt I'd have the time to do it myself (most of my free time is spent on Edubuntu, LTSP, LTSP-Cluster, WebLive and sandbox nowadays) but I'd be happy to review any proposed branch/patch for it.

[MestreLion (mestrelion)]

May I add WebBoard to the list? Its an "yet another pastebin app" that could be merged with both nautilus-pastebin and pastebinit

IMHO, this would be the ideal solution for Pastebin integration into desktop:
- CLI (from pastebinit)
- Nautilus integration (from nautilus-pastebin)
- Panel Applet (from WebBoard, really useful to pastebin things that are already in clipboard, like terminal dumps)
- GUI Settings (merged from nautilus-pastebin and WebBoard, so username could be configurable too)
- All tools sharing same backend (same settings, features and list of supported sites)

Maybe settings would have to move from gconf to a text file in ~/.pastebin so CLI would work in a DE-agnostic way?

Last but not least, a gedit plugin would also be awesome (currently I do it using a combination of Tools -> Manage External Tools ans pastebinit)

[Rolf Leggewie (r0lf)]

Thank you, Mestrelion for your useful comments. I absolutely agree. Let's hope one of the devs can allocate some time for this soonish.

[Rolf Leggewie (r0lf)]

There is also libapp-nopaste-perl, but I wonder if that would be too much integration. Is it feasible/desirable to have enough standardization so as to make the separation between GUI and backend work across perl/python? In other words, any of the front-ends works with any of the backends.

[Rolf Leggewie (r0lf)]

FWIW, webboard development looks fairly stale, though.

[Alessio Treglia (quadrispro)]

Fix released in Nautilus Pastebin.

[Alessio Treglia (quadrispro)]

This bug was fixed in the package nautilus-pastebin - 0.7.1-1

---------------
nautilus-pastebin (0.7.1-1) unstable; urgency=low

  * New upstream release:
    - Changes in nautilus-pastebin 0.7.1:
      + Regression fix: the configuration didn't reflect changes to
        'author', 'username' and 'password'.
      + Remove "askconfirmation" feature, needs more work.
      + Fix libnotify integration.
    - Changes in nautilus-pastebin 0.7.0:
      + Port to GObject Introspection. (Closes: #644692)
      + Migrate from GConf to GSettings.
      + Port configurator's UI from libglade to GtkBuilder.
      + Now depends on pastebinit at runtime;
      + Fix Debian bug#570131, desktop file missed the "Exec" key.
  * Refresh 01-default_pastebin.diff.
  * Drop 02-desktop_file.patch, applied upstream.
  * Drop python-pastebin, not provided anymore.
  * Drop python-support, build with dh_python2.
  * Refresh Depends,Recommends to reflect pytgtk->gobject-introspection
    migration.
  * Update debian/copyright.
  * Bump Standards.

 -- Alessio Treglia <email address hidden> Wed, 01 Feb 2012 20:48:14 +0100

[Rolf Leggewie (r0lf)]

Awesome, thanks Alessio.

[Rolf Leggewie (r0lf)]

The webboard package has been dropped from Debian some time last year.

I guess there's nothing left to be done for pastebinit. Cool.