Thanks for the comment. Please note that adding new dependencies will only be allowed given the original performance of LXDE is preserved and the resource usage remains reasonable. Using more libraries doesn't necessarily mean that we'll be slower. That depends on how we use them. Besides, I mentioned in previous mail. If we're going to use gio, we may create a gvfs fork to kill gnome dependencies. The problem here is, in practice, we're not able to make all things working without any gnome dependency anymore.
Many mainstream modern linux distros are deeply "infected" by gnome already. PolicyKit is now extensively used, which may requires gnome stuff sometimes. HAL is now bound to PolicyKit. Besides, some important system tools more or less requires gnome stuff. At the same time, gnome developers are trying to remove the dependencies on gnome-specific libs in gnome 3.0. Like it or not, some gnome stuff will go into gtk+ later. So, gnome deps is now less a critical issue than it was before, especially when we can fork their programs to remove gnome dependencies. On Thu, May 28, 2009 at 4:16 PM, rafael <[email protected]> wrote: > New comment on your post #352 "Please vote for the future of PCManFM." > Author : rafael (IP: 10.0.0.8 , 10.0.0.8) > E-mail : [email protected] > URL : > Whois : http://ws.arin.net/cgi-bin/whois.pl?queryinput=10.0.0.8 > Comment: > Please look at the lxde design guidelines: > > "- If only several simple APIs of another big library are needed, try to > extract them and add them to your program instead of depending on the whole > library whenever possible. (Beware of license) > > - Only use libraries from other desktops when they are small or efficient > enough and have few dependencies. > > - Only create a daemon if there is a really good reason. > > - Try to keep maximal compatibility with lower gtk+ versions (gtk+ 2.6 is > preferred). Try to make features requiring higher gtk+ versions optional with > proper conditional compilation and compatibility macros." > > In this sentiment, pcmanfm should have less dependencies, not more. The added > funtionality of handling remote volumes isn't worth sacrificing speed, or > having few dependencies (thus occupying less space on the harddrive), or the > modularity which defined lxde in the first place. > > My take: Fork thunar-vfs (or whatever code base that suits your purpose > best). Fork it not as part of the pcmanfm codebase, but as an standalone > project. Give it a new name, make it as lightweight possible providing just > the funcionality that you need. Look out for developers to work on this and > keep it as a separated project (but coordinating efforts) from the > development of pcmanfm. > > Please, keep lxde lightweight. > > Regards, > > You can see all comments on this post here: > http://blog.lxde.org/?p=352#comments > > Delete it: http://blog.lxde.org/wp-admin/comment.php?action=cdc&c=1021 > Spam it: http://blog.lxde.org/wp-admin/comment.php?action=cdc&dt=spam&c=1021 > > ------------------------------------------------------------------------------ Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT is a gathering of tech-side developers & brand creativity professionals. Meet the minds behind Google Creative Lab, Visual Complexity, Processing, & iPhoneDevCamp as they present alongside digital heavyweights like Barbarian Group, R/GA, & Big Spaceship. http://p.sf.net/sfu/creativitycat-com _______________________________________________ Pcmanfm-develop mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/pcmanfm-develop
