New and Improved List-O-Things That Gnome Needs
Posted on 2005-12-29 at 10:44
A while back, I wrote down a list of things I wanted to see Linux get. It's been a year and a half since I wrote that list and I wanted to revisit it to see what has been done and what needs doing now.
My first list in review:
- Money Management Engine: Not only is this not done, it's not even begun or discussed. I still think it's a good idea, but it looks like unless I'm willing to write it myself it ain't happening...unless Gnucash decides to add it to their 2.0 version and have kept that feature hidden and secret this whole time. I'll hold out hope for that.
- Music Engine: There hasn't been much movement on this front, though the discussions of Gnome 3.0 (codenamed Topaz) suggest that such a thing might be in the works for that version. In the interim, Muine and Rhythmbox, for instance, still insist on using their own back ends instead of agreeing on a standard.
- Supplyline integration with the entertainment industry: This was pie-in-the-sky when I first mentioned it. Since I put it on my list that hasn't changed, though at least now there exists iTunes support in Gnome, which is a step in the right direction.
- Better SDL/OpenGL APIs: A complete no-go.
- Better Hardware Integration: Finally a success! Since I wrote this list, hardware integration in Linux has grown greatly. Most things that I've used are simply plug-n-play (but for real, unlike what windows calls Plug-n-Play). More work needs to be done, but I can only be pleased with how smoothly hardware is being supported in Linux nowadays.
- First-Timer Wizards: I've heard it mentioned elsewhere, but so far I've not seen anyone implement such a thing. In retrospect, I'm not sure it's a must-have item as much as a would-be-nice item.
- Hot chicks: Linux still lacks the hot chicks. I hold out hope that it'll change soon, though, and us Linux geeks will be the new rock stars with our stretch limos and groupies and trashed hotel rooms.
So what about now? What do I think Linux and Gnome need now? Let me tell you:
- Money Management Engine: This idea is still good enough that I want to see it happen. Here is a quick run down of what I mean by this.
- Unified API: Linux, or at least Gnome, needs to clarify and unify the API as I describe here. This will help to bring in outside developers, which Gnome and Linux both sorely need. It's time to open this thing up to the legion of Windows-centric developers out there who would write for Linux if the API were more approachable.
- Administrative simplicity: Right now, if you use any of the mainstream linux distros, you can use linux without difficulty, but administering the box is still difficult. Gnome needs to HIGify (yes that's a word, google "Gnome HIG" to see what I mean!) the admin processes. That means, for instance, installation needs to be cleaned up (Ubuntu and Apt-Get do a lot for this but the problem still exists). But it also means we need to start seeing good, HIG-compliant admin interfaces for at least the most common of the services, like apache, bind, mailman, postfix, proftpd, and others. Sure I can install webmin and do it that way, but Webmin is hardly HIGified. I mean something as simple to use as Microsoft's Enterprise Manager to administer PostgreSQL or MySQL. I mean something as simple as Microsoft's IIS admin tool to administer Apache. We've taken care of the end users, not let's show a little love for the linux sys and network admins.
- User-loadable drivers: Let's face it. Linux can supply decent drivers for our hardware, but things like graphics cards are always going to perform better when using the closed-source drivers from their own manufacturers until we can change their minds on that issue. In the interim, how about something as simple as a directory like ~/.drivers where users can place proprietary drivers for use by the system as needed. I mean a simple schema like this: System boots up. System finds built in drivers for all hardware. System shows log in screen. When user logs in, system scans ~/.drivers for any drivers that purport to be for hardware that is currently being handled by a system driver. If a driver is found, the system driver resets with the user driver. If not, the system driver continues to work. This is brain dead simple, allows a user to decide for themselves if they want a proprietary driver running on their system, and makes it easy to make such a thing a user preference. Of course, it would be appropriate to have a way to turn this option on and off and the system level so that corporate users can't play with drivers without the admin's permission. Easy enough to add that option to an /etc/userdrivers.conf file. Each distro can decide for itself based on its target market what the defautl setting will be. Red Hat, being corp user centric might defalt to "No", Ubuntu being home user centric might default to "Yes". How hard is that? I know there are some logistical issues with swapping drivers during login, but to the best of my knowledge none of them are real showstoppers. Seems like something that would take a kernel hacker a day or two to throw together. And yes, I do understand the stability ramifications of this idea. I still think it's better than the current solution!
- Mac Menus: Yes, it's been discussed ad nauseum, but I have to side with the pro-Mac menu guys on this one. It'll free up real estate and make for a cleaner interface in general. It's a major change (in that it'll require retooling of a lot of extant apps, but the end result will be worth it. For the record, it'd be much easier to do the retooling if Gnome had a unified API already! ;-)
- Hot chicks: I'm gonna have to reinclude this item. I mean seriously, what's it all about if not being the idol of millions of screaming girl-fans?