Re: Download manager and proxies

[Henry Baldursson <henry baldursson gmail com>]

On Thu, 02 Dec 2004 14:13:02 +1100, Peter Harvey <pah06 uow edu au> wrote:
> While the file broker idea sounds very nice, what of the more advanced
> aspects of HTTP than just 'GET' requests? How will we handle cookies? Or
> authentication?

Good point. I'm not so much worried about the authentication, that
would be the problem of whichever HTTP library got chosen. Cookies,
however, would require more conversation between the broker and the
browser. So the browser's acquired cookies could be used to
authenticate the browser as a legitimate client for receiving a file
(i.e. pron :-). So some callback registering on the part of the caller
would be necessary. Bonobo should be equip for that.

At any rate, I have started working on this. I decided this would be
as good a chance as any to get more familiar woth Gnome/GTK+ and GUI
programming in general, since I've never done that before.  So I fired
up Glade yesterday and threw together the API as I intend it to be for
an initial release. A simple gnome window with the menu and toolbar
removed, a notebook style container, scrollable windows with vertical
at automatic and horizontal off, and a GtkTreeView for the main list.
The notebook has 3 tabs, Downloads, Uploads, and Completed. The
Uploads is there for things like FTP puts, and the Torrent kickbacks.
The name of the project is gnome-file-broker.

I´m not sure from the usability standpoint which is better, to keep
the notebook containers or to have just a single window with pixmaps
and text style differentiating between the down/up/complete entries.
Thoughts?

I must say, Glade is an amazing tool. I'm attempting to implement this
in C++. Going to fish around for some nice objects that can do
HTTP/1.1 and SSL and then try to fit it nicely into an extensible
hereditary object model to which adding new protocols and special file
handlers (the HTTP getting a .torrent, f.ex.) would be relatively
easy. I'm cautiously optimistic, but I do have a good background in C,
and some background in C++ and working with CORBA's IDLs, so I'm
hoping Bonobo won't give me any difficulties for things like drag&drop
and such. There surely will be implementation problems, but hopefully
they can all be overcome.

I'm also going to try to do as little work as possible, and rely
mostly on existing code mainly because I am lazy. Biggest problem I'm
expecting to have is implementing torrent support, so I think I'll
focus off that in the beginning.