It was a very nice day out today. All the pups enjoyed their walk.

I’d like to take a few minutes to remind people of something that I think they all know, but might have not had in the front of their mind recently:

Different types of communication have different amounts of information bandwidth.

• If you are talking in person, face to face with someone, you have a lot of bandwith. You can hear the tone in their voice, see if they are smiling or frowning. Hear if they are yelling. See if they are slouching or waving their hands a lot. See if they blink, look away or frown at you. You can get a lot of information from an in person conversation, and most of us have been doing it all our lives, so we pick up on subtle cues without thinking about it.
• If you are talking to someone on the phone you have less bandwith. You can now not use any of the visual ques. You can still however listen to tone of voice, volume and laughing/crying/sobbing. Most of us have spent time on the phone and pick up on these clues pretty easily as well.
• If you are exchanging emails with a list or persons, you now have even less bandwith. You can use things like or and you do have time to think and be more deliberate with what words you use (if you choose to), but you have no tone of voice or visual clues. Just the words.
• Finally, if you are communicating with someone via irc or other IM/chat you have even less bandwith. Here people are expecting you to just blurt out your thoughts without much editing or consideration (unlike email). You have no sight or sound to help you, as well as no time to carefully consider your response.

To quote the freenode channel guidelines:

Look for the best in people. If you assume people have no self-control, they’ll confirm your belief. If you look for personal responsibility, and ask for personal responsibility, most people will respond well.

Set a good example. Be what you want other people to be. If you want them to be calm, be calm. If you want them to be courteous and friendly, be courteous and friendy. The habitual behavior of people on a channel is the most powerful influence on newbies arriving on the channel.

Often over in the #fedora IRC channel, we have people who come in with some issue or problem who only spend a minute or two gathering information on it before asking: “Should I just re-install?”. Sometimes they mean re-installing the application/package, and sometimes they mean just reinstalling the entire OS! I fear this is a ingraned reaction from the closed source software world where there is really sometimes no way to tell what the application/os is doing and returning it to a clean state is the only choice. Thats not (usually) how things work in the open source world though.

1. First ask yourself: Did this thing work before? It doesn’t now? What changed? Check /var/log/yum.log or other history and find out exactly what changed. Did you change a setting? Did you upgrade a package? You can think look for how to revert that specific thing that caused the problem. No need to re-install.
2. Do you have reason to think the package or application was messed up somehow on your disk? ie, do you have disk problems, or did you run a script as root that might have messed up binaries? Sure, then a ‘yum reinstall package’ might be in order.
3. Does the problem/issue occur with another newly created user? If not, then that points to a user specific setting. Reinstalling the package won’t do any good, because the setting is tied to your user. Instead look for what that setting is or how to revert it.
4. Is the problem/issue causing you to not be able to boot? Instead of re-installing, look to a rescue media. You can often fix the issue by booting ‘linux rescue’. No need for a re-install.

About the only time you should absolutely re-install your OS is when your machine has been compromised. Otherwise, it’s a lot better to fix the real issue, no only because it’s often easier, but because you will learn something in the process!

Made a batch of espresso Porter tonight…yummy.

In response to John Poelstra’s recent post on Knowing Vs Doing:

I personally really like the “Knowing” side of things. I absorb a scary amount of things every day from the net. I read mailing lists, I am on a dizzing array of IRC channels, I read bugs people post about or refer to, I have a ton of RSS feeds and blogs, facebook feeds, etc. That said, I think it’s a balance for each of us, and sometimes it’s self correcting. If you find yourself slogging through posts on a mailing list, and it continues to be that way over time, you are more likely to just unsubscribe. If you find the traffic on an IRC channel holds little interest over time you are likely to /part it.

I think the best way to handle the overload is to ask yourself at the end of the day when you’re heading off to bed: What things did I get lots out of today? What things were wasting my time/interest? Perhaps if over time you keep thinking a particular thing was not interesting/useful, drop it, or switch to just ’skimming’ it for people you know who interest you.

Also it is possible to change a information channel some of the time, at least talking about the Open Source world. If you find a IRC channel for example doesn’t have topics that interest you and you thought it would, then bring up those topics. Start others talking about them and joining the channel. Perhaps you can make the information more of interest and more containing a signal that not only interests you, but others.

We all have only so many hours in the day and so much we can absorb. Follow the things that provide you value. If you need information from a channel that doesn’t, perhaps ask someone who does follow it to summarize to you whats going on in there.

Finally got around to some brewing again. Made a Scottish ale batch tonight.

I know some folks pointed this out a while back, but I only now just got around to watching it.

It’s 55min, but well worth it. See if you don’t recognize some types of people they mention from your mailing lists, irc, or other open source interaction.

Just thought I would note a few things about the Fedora nightly live composes I have been doing for a while:

1. They aren’t really ‘nightly’. I guess they should be ‘daily’. I start them composing as soon as the rawhide (or now, the Fedora 13 branched) tree finishes composing and lands. F13 branched has been landing around 11am-noon MST (18-19UTC).
2. They are currently composing from the F13 branched tree. This is the tree that will become F13. They do NOT use the updates-testing repo. Only the base repo. Updates-testing items are things that need to be tested and promoted to the base repo.
3. When all of them compose, it usually takes between 8-12 hours to compose. This is 13 “spins”: aos broffice.org design-suite desktop education electronic-lab games kde lxde security soas xfce, in both i686 and x86_64 versions. This means that they all usually get synced over late in the evening here (8pm-9pm MST).
4. They can be found, along with the script I use to make them at: http://alt.fedoraproject.org/pub/alt/nightly-composes/
5. None of them have composed for the last few days due to anaconda having a broken dependency with a python-urlgrabber update that has not yet been moved to the base repo.
6. If you need a particular one in a hurry on some day, ping me on irc and I can often sync over just that one spin as soon as it’s done.
7. If you would like to see something more from these, drop me an email, comment or hit me on IRC.

Some folks have noted recently that none of their Fedora updates have ever gotten any karma (either positive or negative) in the updates system. The default settings have a requirement of +3 karma to move an update from testing to stable and -3 to unpush the update entirely (This of course can be changed by the mainainer).

I run my laptop here with updates-testing installed, and while it’s true I almost never add positive karma to an update, I have very actively added negative karma to get a broken update unpushed. Some examples:

http://admin.fedoraproject.org/updates/F12/FEDORA-2009-13159 was a ppp update that was pushed out of sync with NetworkManager, causing my EVDO card to stop connecting.

http://admin.fedoraproject.org/updates/F12/FEDORA-2009-12432 was a dracut update that caused my encrypted root laptop to no longer boot.

http://admin.fedoraproject.org/updates/F12/FEDORA-2010-0685 was a nss update that had a setting that caused my fedora-cert to get wiped out when I tried to get a new one, so I couldn’t do any builds.

There are others. On the other side of the coin, according to ‘yum –disablerepo updates-testing list extras | grep updates-testing | wc -l’ I have 191 packages from updates-testing installed. There’s no way, given the time I have that I could look at each of those updates and add positive karma. Clearly more automation, less updates, or more testers are needed in order to provide positive karma, but even the chance to add negative is useful.

I finally got around today to editing the introduction to virtulization page on the Fedora wiki:

https://fedoraproject.org/wiki/Getting_started_with_virtualization

I had been pointing users to this for a while, and it was in bad need of a cleanup. Fedora has not supported Xen Dom0 (host) since Fedora 8, yet the above page was listing all kinds of xen info, very much confusing people new to Fedora and virt. There was also a long section on losetup and kpartx to edit guest disks, but we have guestfish now. Many of the libvirt defaults (like a default URI if you are talking to a local libvirt) had changed, making commands easier. Some tools like virt-viewer were not even listed.

So, I took a chainsaw to it and cleaned it up. If any Fedora using, libvirt savvy folks have a chance, please do go over this page again and clean up anything I missed or messed up on. I think Fedora is a great virt platform, but we need to have easy to follow docs to point our users at.

Perhaps people could consider doing one wiki edit a week… if everyone pitched in we would have the wiki cleaned up in no time.