So what can you do with SUM once it's installed? According to the Web site, plenty. You can change:
Grub timeout, default boot title, number of kernels in bootloader menu, enable/disable boot option for memtest86, enable/disable boot option for "rescue mode", if the default boot option should be automatically updated, boot up resolution and color depth, grub menu colors and background, and usplash theme.
But wait, there's more. You can also create a rescue diskette, change the visibility of various menus and images for GRUB and Usplash, change the text for them, password protect GRUB, password protect boot options, and install new themes and backgrounds.
Installing SUM from the command line
I followed the instructions for a command-line install. After downloading startupmanager_1.03.-1_all.deb from the project site, I first entered:
sudo dpkg -i startupmanager_1.0.3-1_all.deb
Just as the instructions predicted, I got an error message, which I resolved by typing:
sudo apt-get install -f
That correcting a missing dependency by installing the ImageMagick graphics suite, then configured SUM.
Next I added new GRUB splash images with apt-get:
sudo apt-get install grub-splashimages kubuntu-grub-splashimages
Finally, I added alternative Usplash images, but this time I went with the GUI, using Synaptic and searching on Usplash, then installing the alternative images it found.
Putting SUM to use
I tweaked and booted several times, playing with various options for GRUB and the bootloader menu. Everything was fine until I started tweaking Usplash. Changes I made in StartUp Manager didn't always "stick."
Not finding much in the way of documentation for SUM proper, I read deeper into the README file that comes with the Fingerprint Usplash theme I had downloaded from SourceForge.net. The additional information and commands there seemed to make a difference in whether various themes would work.
For one thing, I learned that I needed to reset the screen resolution for the boot time to 1024x768. I also found the command that rewrites the boot screen -- sudo update-initramfs -u -- and that helped me get the Ichthux theme working. Perhaps it was this mixing and matching of instructions from installing the theme and installing SUM that led to my problems.
I had changed splash screens, but Ichthux wasn't cool enough for me; I wanted Fingerprint. Alas, that proved to be more of a stretch than I could make. After installing the Fingerprint theme, then running the update-initramfs -u, I ran into serious problems.
The first problem was that after the Fingerprint splash screen appeared, and I pressed ALT-F1 in order to get text scrolling down the screen as the system booted, the system locked up. I tried powering down and rebooting, this time without the ALT-F1. Same thing -- locked up tight.
I recovered by choosing a recovery option from the GRUB menu at my next boot, then running apt-get remove startupmanager as root. That allowed me to get past the lockup, see the Fingerprint splash screen, and then proceed as normal. I decided that was about as cool as I needed to be.
Conclusion
I found StartUp Manager to have some real utility, primarily in providing a clean, easy-to-use GUI which I can use to tweak GRUB boot settings. But I think the main thrust of the project is for fun, to allow advanced users to tweak the default splash settings and make their Ubuntu installation unique. Whether or not the coolness achieved is worth the effort required to install StartUp Manager and develop and maintain alternative splash themes is strictly up to the individual and his skill level. I went for cool, but ended up a fool. Your mileage may vary.
Note: Comments are owned by the poster. We are not responsible for their content.
There are about a zillion things in GNU/Linux that really need fixing. Why waste time on the boot loader and splash screens? Lilo works. Grub works. Please, developers, if you have time to spend on GNU/Linux, fix something that does not work, or works badly. How often do you reboot your machine anyway? uptime tells me I last rebooted mine 18 days ago... and usually it is even longer than that between reboots.
"Lalalalalalalalal, I can't hear you!"<nobr> <wbr></nobr>;)
I wish more FOSS devs would invest their talents in that last 5% of polish and refinement, instead of continually inventing new projects that never get fully-baked.
The first 95% of the project takes 95% of the time and effort, and the last 5% of the project takes 95% of the time and effort...
Seriously, getting something up and running is fun; the "last 5% of polish and refinement" is work, and takes the kind of attitude that separates professionals from amateurs.
There are a lot of fundamental features in Linux that should be bulletproof by now, instead of continual exercises in frustration. I wouldn't call this project a waste of time. But I wish more FOSS devs would invest their talents in that last 5% of polish and refinement, instead of continually inventing new projects that never get fully-baked.
But this project is part of that 5%, it provides GRUB an easy to use facility to add on screen graphics. Basically the "bling" that people keep moaning about that supposedly isn't present in Linux.
Bruce S.
This app does not replace grub or lilo, it makes it easier to maintain grub.
It's needed. It's not redundant.
My apologies to both of you for getting that wrong. I'll change it in the story. Thanks for the comments and the correction.
A shiny new alternative to lilo and grub. A step backwards for GNU/Linux.
As someone already said, Lilo and Grub already work pretty well, so nobody really needs this. But what's the harm in offering another choice?
Here's the harm.
Suppose a newbie asks you, a "Linux expert", to help him/her with a problem. You go to the newbie's machine and find that the problem is with the boot process. You know Grub inside out and you know a bit about Lilo<nobr> <wbr></nobr>... but the newbie is using the Mickey bootloader or the Froboz boot loader<nobr> <wbr></nobr>... "Sorry, can't help"<nobr> <wbr></nobr>... and another frustrated newbie goes back to Windows.
GNU/Linux needs fewer redundant applications, and many of the applications it has need to be properly finished to a professional standard. But it's much more fun to start a new project than to help finish someone else's. It's a big weakness, folks, and I think the way to deal with it is to greet every redundant new project with a Bronx cheer.
Giving a lot more kudos and recognition to people who help with existing projects would help, too, if we could figure out a way to do that. A start would be for Linux.com to stop publishing articles about a "new" XYZ and start publishing articles like "Joe Smith contributes bug fixes to ABC project", telling us that ABC is now a lot more usable because Joe has fixed bugs that the original author couldn't.
People who rarely boot
Posted by: Anonymous Coward on April 21, 2007 07:19 PM#