For my trip to flock earlier this year, I picked up a external battery pack, since I wasn't sure when and where I would have power and wanted to make sure my laptop and phone and such could last for the long plane flight from us to europe. I looked around and ended up getting this one from amazon: http://www.amazon.com/gp/product/B00DN0KBXU/ref=oh_aui_detailpage_o08_s00?ie=UTF8&psc=1 It's pretty big and heavy, but it packs a lot of power as well. It's got 2 usb ports (one with higher Amperage) and 1 DC out that can take a variety of tips to connect to a variety of laptops and devices. I got a extra tip for my yoga 2 pro, but it turns out that was also one that was included. In addition to having power in places where you have no power it's been very handy on regular trips: Hotels often provide few power outlets or they are anoyingly placed. With this device I can plug it in, then plug usb devices (phone, etc) and laptop into it. No need to carry the usb to AC adapters and find them and plugs for them, or get out my laptop's power brick. It does have one anoying quirk in that if _anything_ is plugged in it's on and draining, so you can't leave just a usb cable attached without it running (slowly) down. Despite the weight I am going to continue to carry this on trips and would recommend it to anyone with lots of power hungy devices.

I see people arriving on lists, irc channels or conferences all the time with a clear peeve or agenda, but it's clear they don't understand how open source projects work and just get themselves and others frustrated and upset. In the end many go away angry with their plan/idea/bug unimplemented or unfixed. There is by no means a sure way to get whatever changes you want implemented in any open source project, but over the many years I have been involved with them there's definitely things that improve your chances. I thought I would share a list here and see if others had additional ideas:

• Be open about how your bug/idea/plan/goal gets implemented. It may be that the community of the project comes up with a better or different way to solve your problem. You should be happy, because it is a solution. Focus on the end goal more than nitpicking about some details on the way.
• If you are just joining a community to open a dialog about your bug/idea/plan/goal, be polite and ask for feedback or thoughts from the community. Never demand or try and force a specific detailed plan on others.
• Be involved in the discussion, but don't try and monopolize it. Force yourself to NOT reply to each and every reply to your initial interaction, only the ones where there are questions of you, or where you are adding new information.
• Have an open mind. Allow the community to explain to you why they think the way they do about your bug/idea/plan/goal. Perhaps they will convince you, perhaps you will convince them. Keep your eyes on the end goal of what you want to accomplish. Perhaps you are already there, but didn't realize it.
• Offer to do whatever of the work you can. Even if you aren't able to fix the bug, make the plan, have the idea, you can still offer to test solutions, write documentation, perform other functions the community needs but doesn't have people doing. If you start working with the community your bug/idea/plan/goal might be something you can more easily discuss, since you are now a member of the community.
• Have patience. Make your case, answer questions, then wait. If your pet bug/idea/plan/goal isn't something that high priority for the project it might be a while. Nagging people daily when they might have higher priorities is a poor approach. An occasional polite "Any news here? Happy to provide any information to move it along" is fine.
• "If you think you have a simple solution to a complex problem, perhaps you have not fully understood the problem". Don't be afraid to suggest some elegant solution to a complex issue, but realize that if it was that simple it would have been solved a while back.
• Know when to let go. Sometimes your bug/idea/plan/goal gets rejected by the community. Be graceful and move on. Bringing up your bug/idea/plan/goal after it's been rejected is likely to only decrease any desire the community has to work with you.

I'm sure there's a bunch more tips and likely the folks that need them the most won't see this post, but perhaps some of you will find these tips useful in getting your way in a open source project.

I grew up in largely in southern New Mexico and thus I assert that the worlds best vegetable is: The green Chilie I was back down there recently and of course had to hit Nellies Caffe (Twice) for some awesome food: The combo plate

I've seen a fair number of folks recently asking questions like: "Is $foo installed by default in Fedora?" or "How much space does a default Fedora install take?" or "If I do a default install of Fedora can I then... " or "I've heard rsyslog or sendmail or $foo is no longer installed by default, but I see it here, why is that?" The problem with all of the above questions is that the term 'default' isn't well, defined. The Fedora Project doesn't produce one single thing that is "Fedora" that has a specific known set of packages installed. With Fedora 21 there will be 3 'products': cloud, server and workstation. However there will still be all the various "spins", network installs, upgrades and the like. So, the lesson here is that almost always you need to be more specific when you talk about "default" when it comes to Linux installs.

I spent some of last week and this week getting all the Fedora koji builders updated/re-installed. There's quite a stable of them these days:

• 2 bkernel servers for building only kernel, grub2, pesign, and any secure boot needs.
• 27 buildvm servers. These are Fedora 20 kvm guests on 3 large virt servers (9 per). They do general i686/x86_64/noarch builds.
• 12 buildhw servers. These are Fedora 20 servers on hardware. They do general i686/x86_64/noarch builds and also image builds that need hardware virt support
• 2 buildppc servers. These are RHEL6 servers on ppc lpars. They do epel ppc64 builds.
• 48 arm-builder servers. These are highbank arm SOC's. (2 chassis of 24 each). They do general armv7 and noarch builds

Thats 91 servers to update/reboot/re-install. Happily we have ansible, so it's not so bad at all. I can reinstall all the buildvm's in about 15minutes with ansible, and updating itself is pretty simple as well, with most of the time being waiting until particular builders aren't building so there's no outage. All of them (except buildppc) are now up to date on Fedora 20 updates and are running the 3.17.2 kernel.

I did a bit more playing around with running Fedora with no firmware. Josh Boyer pointed out that the microcode_ctl package also had firmware in it. This is the 'errata' for cpus that fixes various bugs after they were released. Also, in my previous tests I didn't regenerate the initramfs without the firmware being available. So, this could have meant firmware was getting loaded in initramfs before the Linux kernel booted up. So, I removed all the firmware packages again, along with microcode_ctl and updated to the latest rawhide kernel. Amusingly I forgot that the kernel has a Requires(post) on linux-firmware, so the install pulled that in and used it when it generated the initramfs. I had to remove it again and re-run dracut to make a new initramfs to be sure it was not being used. A reboot later and... everything seemed to come up fine (again except for the iwl wireless that needed it's firmware package to work). This time I didn't notice any lags in X/input. So, that could have been just a bug in that previous rawhide kernel, or some interaction with firmware loading and debug kernels? Hard to say. In any case things seem to work fine here. I'm not really going to keep running like this, as I don't mind firmware that makes things better for me, but it's nice to know if you can run without some or all of them if you so choose.

From wikipedia: firmware: the combination of persistent memory and program code and data stored in it. Fedora ships firmware as long as it meets some critera: https://fedoraproject.org/wiki/Licensing:Main#Binary_Firmware Firmware is in a weird place. Some folks consider any use of loadable firmware to be bad and non free, but yet are fine with non loadable firmware in the devices they use. Consider two identical network cards: One with loadable firmware and one without. In both cases you can't examine the code or make changes, but in the case of the device with loadable firmware you can at least have the option for bugfixes or improvements, or use of the device at all in some cases. Is using the one without loadable firmware "ok" while the one that simply allows you to load firmware is "wrong" ? In any case, Fedora puts all the firmware in packages with 'firmware' in the name (most of them in a 'linux-firmware' package), so I wondered: what would happen if you removed all those. How well could a modern laptop work? So, a quick download of all the firmware rpms (so I could easily reinstall them), 'yum remove \*firmware\*' and reboot and... The laptop booted up fine. As expected my intel iwl wifi wireless adapter didn't work at all. It needs a firmware load in order to function. The second thing proved impossible to track down, but X and my touchpad in particular was very slow responding. I couldn't figure out what firmware was causing this, but reinstalling linux-firmware and rebooting got it back to normal. It's an intel video card, and I couldn't find any firmware loading failures in the journal anywhere. ;( If anyone has ideas on how to track that down or wants to try themselves and see... send me a pingback or comment. ;) So, it is indeed possible to run Fedora without the firmware packages.

For the last few years our dogs have been blood donors at our local animal hospital/blood bank. Yesterday was the last donation for our older hound Burner. He is going to turn 9 before his next donation and they want dogs to be younger than 9 to donate. He absolutely loved going to give blood. Tons of attention and pets from the staff, car rides, treats and all kinds of fun. They all signed a 'blood donor' bandana and gave him a really nice dog bed as retirement gifts. :) They also said he could come back anytime to visit. We loved it because not only were we doing good, but he got free yearly checkups, free heartworm meds, and discounts on procedures. I know the blood our pups have donated have saved lives of other dogs. If you have a dog, do consider applying for them to be a blood donor. Burner knows how to relax

I read/skim/follow a bunch of rss feeds. It's a handy way to follow something without the overhead of joining a mailing list or the like. That said there's some feeds that do things that are downright annoying. Sometimes to the extent that I unsubscribe. If you run a RSS feed, try NOT to do these things:

• Feeds that are just links to posts. If I wanted that, I would have just visited your site. If there's no information or context, just a link you are wasting the features of rss and I (and many others) will likely just unsubscribe.
• Feeds that are cut off at a arbitrary limit. I understand that some sites prefer you read the posts/articles on their site with their style, etc, so they cut off the information in the feed to try and 'hook' you in to reading the rest on the site. Sadly, pretty much all of these I have seen have been set to some arbitrary limit, which often makes the 'hook' make no sense or provide not enough information to make me interested. Just include all the content. :)
• Constant edits. We all have been there where you need to go back and edit a post to make something important clear. However, if your feed constantly reedits posts it means that I will constantly re-see them. Don't cry wolf, but use edits very very rarely.
• Posts that are entirely a video or audio file. Sure you can point to or show a video or audio file, but if there's no text for context or reason why I would want to take time to watch or listen to it, I am not going to.
• Have the link for your post/entry not actually go to that post/entry. I have had feeds that redirect all links to the top level of the blog or redirect you to some stupid sign up page, but yet the feed has the content. This is very annoying if you are trying to go comment on something.

I've resisted making a twitter account for a long long time, mostly because I didn't want another thing to follow and there wasn't a compelling reason to create an account. Recently I read an interview with the founder of untappd (the social beer gamification application) on opensource.com: http://opensource.com/life/14/8/interview-greg-avola-untappd Untappd is very much built on open source and wouldn't exist without it. It's awesome to see the tools and frameworks we work on shipping in distributions like Fedora being used in the real world by real people to provide real services. So, I figured to support a open source friendly platform, and to make a friend who's been bugging me to sign up for a long while now (Hi Xavier!), I decided to make a untappd account. I got signed up and installed, but then realized I needed to connect it to a social media to 'check in' when I am drinking a new beer. There's no connection for g+ (Thanks google, for STILL not providing a api to connect and post to g+) or diaspora (Ok, I can understand that since there's so few people on diaspora, but it would have been nice :). The options are: facebook (no thanks, I treat them as a read-only medium. I have no desire to ever share anything ever with facebook), foursquare (I have heard bad things about them, and didn't need more games, just untappd), and twitter. So, long story short, I am now @nirik99 on twitter. Not sure if I will post anything there aside from untappd checkins, but I might decide to. Feel free to follow me or whatever. I'm also 'nirik' on untappd. Happy to exchange good beers there. :)