GEEK ALERT: This article is about nerd stuff!
A good friend of mine has, for a long time, stated “free software is only free if your time has no value.” I was reminded of this today while reading reviews for commercial implementations of Wine (this is a geek piece so if you don’t know what something is follow the link). In the comments on these pieces the usual suspects were in attendance: “Why pay for something you can get for free”, “They are taking mind-share away from Wine so they suck”, “You should never run anything that’s not completely Free”, “Running windows software on Linux is stupid” and so on and so forth. But one comment caught my eye. It was in reference to Bordeaux, which is a commercial implementation of Wine that runs about 20.00 USD. I am paraphrasing but the gist of the comment was “It’s only $20.00 and the time I would spend configuring wine is worth a lot more than that to me so it’s an easy decision.” I completely agree with the commenter on this point. My time is valuable and if you offer me something at a price point that makes me realize that the time to do it for free is more valuable I’lll buy it.
This, of course, means I am not an OSS purist but then again I never claimed to be one in the first place. I take the Linux Torvalds view on that little piece of OSS piety: “anybody who tells me I can’t use a program because it’s not open source, go suck on rms (Richard M. Stallman). I’m not interested. 99% of that I run tends to be open source, but that’s my choice, dammit.” I use Ubuntu on my desktops and laptops for a majority of the things I do because it’s simply more stable, less prone to malware, and faster. Everyone else’s mileage may vary. The only time I even try to get other people to use Linux is when they need a server or they call me way too often for tech support. I figure if I am their line of support then they need to be running what I know how to support. Outside of that I couldn’t care less what my friends run on their systems. I will admit to feeling mildly superior to those running windows but that could merely be an artifact of me feeling mildly superior to most of humanity most of the time. But I digress…
Back to the point at hand. My company is looking at replacing an aging hardware/network monitoring system and the first place we looked, being a Linux based company, was the OSS solutions. All of the solutions that are completely free could do what we want, eventually, if we hacked it, and maintained it, and spent countless hours tweaking it and therein lies the rub. While the decision is not final it looks like we have chosen and OSS solution with a commercial/enterprise edition, for which we will pay, to get all the features and the workflow that we need. The upfront cost will be close to 40,000 USD and dropping to half of that the following years. While the difference between 40k and free seems like a lot, especially for a small company like ours, the fact of the matter is that it would cost well more than 40k a year to take a fully OSS monitoring system and make it do what we need it to do. The initial cost of development coupled with the cost of maintaining the solution over years is extremely high. I would have loved to have found a system that does what we need it to do and is fully OSS but it didn’t exist.
This is a problem with the OSS community as a whole. I see articles every week opining on things whether there is too much choice available in Linux or whether the KDE vs Gnome wars are keeping Linux off the desktop and so on ad nauseum. At some point those debates may actually be on point but right now they aren’t. For the average used Windows is still the easiest choice and until the folks that do the coding realize that then OSS will remain an anemic sibling to Windows. Ubuntu works, out of the box, on every system I have dropped in on the past five years. It boots, it runs, and on a wired connection it hits the intertubes without a hitch however my desktop distro of choice doesn’t let me play MP3s without having to either search for the packages to support it or manually adding some GStreamer codec pack and honestly that’s a complete deal-breaker for the average user. Furthermore there is absolutely no functionality for hardware profiles like there is in Windows. You want to see frustration in a geek do the following: Hand him a laptop with the latest Ubuntu, a docking station with two monitors attached, and ask him to setup a workspace. Come back two hours later and ask him to do a presentation on the overhead monitor in the the conference room. Hint: On Windows you plug the laptop into the docking station, configure the monitors, un-dock it, plug in the projector and configure it and never have to do it again. On our venerable OSS OS you have to configure each of these things each time you use it. And don’t get me started on trying to run three monitors. I am nearing the rant phase at this point so I will wrap this up. For anyone who has cure suggestions to cure my hardware profile woes: I don’t want a package like whereami that will copy Xorg configurations into place based on some network/arbitrary hardware detection script that I have to write by hand. If I wanted to deal with that I’d code it myself. I want standard things to work. I have it down to a science and the scenario described above doesn’t much phase me however the average user isn’t me.
The point to this whole article is that I can do everything with OSS that I can do with closed source software (I am referring to myself and my job functions not some specialty software package required for a job a I don’t do and never will) but it’s simply not worth my time in many cases. I even run a VirtualBox VM so I can use Visio because the OSS alternatives suck. If someone coded a working display profiler that would save hardware profiles like Windows does, and has for years and years, and charged 20 or so bucks for it they’d get my money. Sure I can write a completely free thing that kind of does the same thing and would work for just me but it’s frankly not worth my time. Just like it’s not worth my time to learn a new set of scripts, like whereami, to accomplish something so simple that Windows users can do it without even clicking a damn button! Linux isn’t making inroads in the desktop arena because it can’t compete with the basic feature set available in the competition. I am not referring to Visio or anything else software package related, that’s another rant completely, I am referring to things Windows can do out-of-the-box with nothing else even installed that no Linux distribution can do. Until that’s fixed the Linux desktop will be relegated to nerds like me. I love Linux and I will never stop running it but I am not going to kid myself and say it’s a viable alternative for the every day user because it’s just not. And don’t get me started on the state of e-mail clients in the OSS world…
I was bumming around on last.fm today and started to wonderin’ about a couple issues involved in organizing large music collections and I thought I’d pose some questions to my three readers. I am up in the air about these three questions so following in each question is my current thoughts and dilemma.
1. How do you handle tagging multi-artist albums?
For instance I have a bunch of split albums and the artists don’t necessarily collaborate on the songs. So I think I should put which artist did the recording in the Artist tag but then sorting by artist ends up with a bunch of weird half albums but I am not sure that matters. And then there’s the question of songs that have other artists on them like Billy Joe Shaver’s “Get Thee Behind Me Satan” with John Michael Montgomery. If I add in John Michael’s name then I have a new “Artist” with one song and not an accurate Artist to boot. So I am thinking that the Comments field might be the place for this sort of information but I don’t know. And then there’s compilation albums. Maybe the answer to my question is to sort by Album but so many music players don’t allow a tag for compilation anywhere and I end up with seventeen albums with the same name and one song each.
2. How anal retentive are you about genre tags?
Another one I am torn on. Do I care about the difference between Red Dirt, Texas Country, Outlaw Country, Country, Classic Country, Alt Country, and probably a few more that I haven’t even thought to include? I do use the Genre tag in sorting and sometimes in smart playlists but how granular should I be? And are genres arbitrary or is it easier than I am making it? Is Tim Barry Alt Country, Americana, Folk, Rock, or some other sub-genre I missed. And what is the difference between Americana and Folk, and can a Canadian artist be Americana?
3. If you use Linux what do you use for your music player/organizer?
I am currently using gmusicbrowser and miss Amarok but the new version sucks balls. No player since has matched the smart capability, the options, and the sheer usability of the old Amarok. I think gmusicbrowser could catch up as it’s a young project and the maintainer is active and considers suggestions but it lacks some of my favorite features. So if you have a suggestion that’s not Amarok, Banshee, or Rhythmbox then please let me know. On a side note I think I’ll review gmusicbrowser this week, like with screenshots and everything. Oh and if you use Windows or Mac and get an urge to suggest iTunes for organizing a large music collection then I reserve the right to mock you for iTunes and for using Windows.
So there you have it. If any of the three you want to weigh in I would appreciate it. Meanwhile I’ll suffer through inaccurate tag information as best I can.
I am a dyed in a wool Ubuntu user as far as desktop systems go but I haven’t always been. I used to use Mandriva, and before that SuSE, and before that Slackware but I happened onto Ubuntu and my “I may be a geek but I am a lazy geek” streak kicked in. I have been very satisfied with apt and the Synaptic interface. In face other than for my love of tweaking my OS I would have the most stable desktop/laptop of anyone I know…after an Ubuntu install of course. Now don’t get me wrong Ubuntu took getting used to. I was a KDE fanatic for the longest time and KUbuntu is the red-headed stepchild of the Ubuntu family. But then the KDE folks went and crapped out the 4.x line and killed my love for KDE so Ubuntu was just about perfect. Not was – is. I still love it. Which is why I cannot answer the question: Why did I spend hours trying to get OpenSuSE 11.2 running on a Dell M4400 over the past couple of days?
First off everything didn’t work out of the box. The resolution for the install isn’t supported all the way through so a black screen and multiple reboots didn’t help me fall in love. Well I got past that and since I am a Linux nerd I won’t let a little install weirdness stop me. Then I went to set up my development environment and the version of Eclipse in the repos is Ganymede…ARGH! I won’t bore you with the details but I need Galileo for what I am doing. A manual install worked of course but for the repo to be that out of date on a brand new release is upsetting. Now I went with the defaults during install in hopes that a properly integrated KDE 4.x wouldn’t suck. I was wrong. The first time it tried to update it was a morass of pop-ups telling me the update widget had, for all intents and purposes, dropped a deuce all over itself. Nothing I did would drag the desktop out of what looked a pornado minus the porn. Pop-up after pop-up after pop-up after pop-up…well you get the picture. Still soldiering on I rebooted and did the initial update via YAST which worked. Of course none of this takes into account the fact I have to allow KWallet to be opened twice, even with remember password checked, just to connect to my passwordless WIFI which is an annoyance I don’t feel like troubleshooting at this point.
Now all of this would be alright if the KDE 4.x widgets didn’t randomly crash and I do mean randomly. Then today after all of that I was home sick from the office. I wanted to surf the web just a little bit, maybe watch the video of Barney Frank mouth off about changing the Senate rules so they can cram health care down our throats, but that wasn’t in the cards. Flash simply refused to work at all. I googled and found various solutions but none that made sense or addressed the issue of flash, from the official OpenSuSE blessed install source, not freaking working out of the box on a fully updated system. I went to install my favorite terminal program, Terminator, to try and fix the problem and it wasn’t in the repos at all. Well after so many straws a camel’s back will give in and so I gave up.
If you want to know why Ubuntu is the apparent flagship of the Linux world, why it has so many fanboys running amok, why end users and nerds alike use it without shame then you have your answer. It works. Period. Of course there are quirks but the issues you could run into are no more than getting Windows 7 to the point where you like it. Sure the answers are different than Windows answers but the problems are no more numerous. If you want to build a distro that can compete in the desktop space, kernel scheduler aside, then build one that works. So as of hitting submit on this post, burning the iso that just finished downloading, I am done with OpenSuSE. Period. I want an OS that works thank you very much.
Tune in tomorrow to watch my head explode trying to explain modules and classes in Ruby….
… the Module class of module is the superclass of the Class class of class…
Yeah so…be seeing you from an actual function OS shortly.
You may remember my post on Ubuntu Netbook Remix and how to get it running on an Acer Aspire One with the now infamous GMA500 graphics adapter. It isn’t hard but isn’t easy either and it’s not for the beginner. With the price of GMA500 netbooks dropping there will be more people picking them up. From experience I can safely say that the Windows experience on the lower end netbooks is not that fun. The performance with Linux is much better and since the UNR interface is intuitive I can see folks wanting to run it but not wanting to go through the pain in the arse that it is to get UNR running. Top that off with the issues with the sound that haven’t been worked out yet UNR won’t be the choice of the masses no matter how good the interface is. In comes Jolicloud… The install was dead easy. After getting my invite I followed the instructions on the website and about an hour from start of the download to reboot after install I had a running system. I didn’t play with any drivers and have native resolution, didn’t have to reboot to get any weird sound issues sorted out, didn’t have to add any PPAs or download anything in order to get anything working. Out of the box Jolicloud installed on the Acer Aspire One 751h and just plain worked. The install consisted of less questions than a Windows install and never gave an error. It was so easy it was almost decadent. Under the hood Jolicloud is some sort of social/cloud/netbook OS based on Ubuntu. You can follow other Jolicloud users and get suggestions for applications. I don’t really get the social aspect just yet but I also haven’t really tried it out. The Jolicloud Dashboard has the best end user interface for installing applications that I have seen in a Linux based system at present. Most of the apps are Prism based which makes sense for a cloud based OS. For those that don’t know Prism is a system that allows web based apps to run on your desktop but isn’t a browser. I know it doesn’t make a lot of sense but it works pretty damn well. Dropbox is included for online storage and all the Google services your little heart desires are available along with a ton of web based apps I have never heard of. They even include Adobe Air in their list of Applications so getting Tweetdeck up and running was as easy as installing Air and running over to the Tweetdeck website. Everything was really that easy. Even Boxee is available out of the box and while it doesn’t work wGMA500 it worked without any weird steps involved. This is an OS I could give to my mom without worrying about tech support calls until the end of time. Now being who I am I did do some playing. I installed Ubuntu One which isn’t included in the apps list. I also installed terminator because it’s my favorite terminal application. Being based on Ubuntu you can get to Synaptic but it’s not in the menu. Jolicloud would prefer you used their applications list and rightfully so since they can vet the apps and make sure they work properly as well as offering more cloud based apps than local apps. But if you really want to play, and I always do, you can hit alt-f2 and run ‘gksudo synaptic’ and get the old familiar Ubuntu package installer. Once you are there anything you install does show up in the menu so it’s pretty easy to fiddle around with your netbook and the fact that everything works, out of the box, on what they are referring to as a pre-beta release makes this a pretty freaking nice choice for geeks and end users alike. My recommendation is to skip futzing around with UNR and get yourself a Jolicloud invite. Even if you don’t go in for the social aspect of the OS, and I still haven’t, it’s the real deal and it works very nicely at the moment. I want to play with Chrome OS if there is ever GMA500 support but trying to get me to wipe out Jolicloud on this box will be a wrestling match for sure.
This guide is adapted from here for the initial rooting and here for the app removal. If you follow this guide and break your phone I can’t help you. All the following actions should be taken with the knowledge that it is at your own risk and will void your warranty. For that matter Sprint may send an army of ninjas after you. Don’t say you weren’t warned….
This guide assumes a base install of Ubuntu 9.10 but should work for previous versions. If you know enough to have changed your download directory then you know enough to modify the following commands.
- Download the Android SDK
- Download asroot2
- Download the recovery image
- Extract the SDK into your home directory
tar -xzvf ~/Downloads/android-sdk_r3-linux.tgz -C ~
- Copy the other two files into the newly created android-sdk directory
cp ~/Downloads/asroot2 ~/android-sdk-linux/
cp ~/Downloads/recovery-RA-heroc-v1.2.3.img ~/android-sdk-linux/
- Turn on USB debugging on your phone: Menu -> Settings -> Applications -> Development -> USB Debugging
- Connect your phone to your computer with a USB cable
Now in normal circumstances I would never recommend you “become” root as it’s generally not necessary but my attempts to root the phone did not work with sudo so for the next steps you will need to actually be the root user.
Rooting the phone:
- Become root
sudo su -
- cd to your the directory created when you un-tarred the SDK
- Copy asroot2 to your phone, change the permissions to make it executable, execute asroot2
./adb push ../asroot2 /data/local/
./adb shell chmod 0755 /data/local/asroot2
./adb shell /data/local/asroot2 /system/bin/sh
- Now we will use the access asroot2 has given up to create a su binary for root access
mount -o remount,rw -t yaffs2 /dev/block/mtdblock3 /system
cat sh > su
chmod 4755 su
If all has gone well you should now have a root prompt “#” instead of “$”. If you don’t then ask over at the first link in this post for help. I didn’t do any of the development on this I am merely giving the steps that worked for me. Now it’s time to flash a new recovery image so we can take a backup of the phone before we start screwing around. These steps are from the android-sdk-linux/tools directory you were in before. If you already exited your android shell bring it back up with ./adb shell and type su to get your root prompt.
- Reboot your phone. Yes this is a necessary step. The recovery image won’t flash if you don’t so this.
- Copy the recovery image to your phone once it has booted
./adb push ../recovery-RA-heroc-v1.2.3.img /sdcard/
- Connect to the shell on your phone and flash the recovery image
flash_image recovery /sdcard/recovery-RA-heroc-v1.2.3.img
- Reboot into recovery mode and take a Nandroid backup before you do anything else
- Once the recovery screen has come up take a backup, make sure you have enough free space on your sd card before you do this, so that when you screw up your phone you will be able to restore it. Once you get bored in the recovery image select the top option to boot the phone into your regular HTC Sense environment.
I received no output from the command to flash the recovery image and it took less than two seconds once I had rebooted the phone. If you do not reboot the phone prior to the above steps you will most likely see Out of memory errors. If you see them even after a reboot then uninstall all your silly little app market playthings and try again. If you play around and screw up your phone without a backup then you’ll have to track down a recovery image or single apps and so on. Once again I WILL NOT HELP YOU WITH THIS.
Removing the stock Sprint Apps:
Simply connect to your phone with the SDK, become root, and remove the files from /system/app
mount -o rw,remount -t yaffs2 /dev/block/mtdblock3 /system
At this point you will have a list of all the stock crap on your screen. If you go about removing stuff you may screw up your phone. That’s what the backup we took earlier was for. Don’t think you made a backup with app store app so you are alright. Do the nandroid backup. Don’t be an idiot about this.
To remove an app simply use the rm command. The following commands remove Sprint NFL, Sprint Nascar, Peep, Footprints, and Sprint Stock, the stock music program (if you are using something else), and Quick Office
There’s a list of apps and some more description here on XDA Developers.
Like I said: This is NOT my work. I merely typed up all of the steps and put them in a single place. All thanks should go to the folks who posted the threads and risked bricking their devices to make this possible!