things I didn’t know until today
okay, so I guess I kinda knew them but had never tried them.
1. generate rsa key using openssl or what not
2. have the private key on disk in a .pem file
3. run ssh-keygen -y -f myprivatekey.pem > ~/.ssh/authorized_keys
4. ssh -i myprivatekey.pem user@host_with_pub_key_on_it
simple as pie.
interesting
in my process for looking up info on android phones I found:
Which lets me run an android instance under kvm if I use the vesa driver.
That’s pretty hot.
phones
Shopping for a cell phone is hard.
really really hard.
JUST FEDORA
For the love of all that’s good, people. Edited doesn’t mean any modification to anyone’s post.
It doesn’t mean censorship.
it’s not a ‘star chamber’.
it’s nothing more than a subplanet/rss feed for people who JUST WANT TO READ ABOUT FEDORA RELATED ITEMS FROM FEDORA CONTRIBUTORS.
That’s it.
Here let me help you a bit more – in your mind just do: s/edited/justfedora/
It’s really not controversial nor disruptive.
holy crap has there been a lot of noise about this for no good reason.;
There’s no change, There’s no drama, it’s just a frelling focused planet collection.
You know what benefit you get out of this feed? If you follow planet ‘just fedora’ then you don’t have to read stupid crap from me about bicycles or about energy issues or that stuff.
and given some of the snarky-ass comments I’ve gotten at least some of you would be happy about that.
fedora planet – edited
Hi folks,
Andrea Veri (averi) came up the idea of having a planet instance that is fairly strictly controlled as to what content gets posted. More or less only fedora-relevant posts.
He wrote up guidelines and a wiki page ,etc.
http://fedoraproject.org/wiki/Planet_Edited
fudcon day 2 and 3 and 4
sunday:
Woke up late and just managed to make it to James’ and Florian’s talk on packaging things about 5 minutes late.
James’ talk was “10 things you don’t know about yum” which, in fact, I did know. A good talk more or less all the obscure items we’ve added that really need more docs written about them.
Florian was talking about appstream. I had a good idea what he was going to say – there’s a lot to sift there but the gist is this is a metadata layer on top of packages to help people find the thing they want to install. It’s not really sysadmin-oriented mostly end-user oriented. I’m still not sure who these end-users are who are not on systems managed by someone else but this is harm-free technology. If it isn’t used we’re out nothing. If it is used then we benefit.
Ended up sitting and talking to some folks about packaging some more following this talk and skipping the 10:00 talks.
Went to the next big fedora engineering project. The pitches were interesting but I suspect they are going to be much more time and labor intensive than was originally estimated. Toshio, Smooge and I gave our pitch for the Fedora Infrastructure project of ‘clean up our stuff’. Toshio was kind enough to type up our infrastructure goals from the night before. We gave our pitch, it wasn’t badly received but we probably didn’t stir up a ruckus, either.
Back outside for more time comfortably sitting in the sun and warmth and talking about package mgmt and yum/rpm goals. Tiring but man was it nice to be in the sun and be warm.
Lunch and then back inside for the lightning talks. I pitched my idea which was met with deafening disinterest. So onto the scrap heap of bad ideas it goes. 
It’s nice to be able to fail quickly and with relatively little damage.
Andrew Overholt’s lighting talk showing off the advanced features of the spec file editing/fedpkg integration of eclipse was impressive. It was almost impressive enough to get me to try out eclipse again. Andrew does a good job of making a fun set of slides. His pitch was like an infomercial for ginsu knives or the sham-wow, clever.
Met Major Hayden from rackspace, finally. Got a few questions in and a promise of more info before I had to run. I think there is much value in the conversation with rackspace for fedora infrastructure’s future.
My afternoon of hackfests was still more packaging/future-goals discussions. This afternoon’s talks were a bit more involved and we got into a couple of disagreements, mostly, ultimately on terminologies. Got to meet Jay Greguske and he helped flesh out important distinctions about how rpm and yum are being used in some of our virtualization tools.
Eventually get back to hotel and take a break for a few minutes before catching up with the other vegetarians and similar people and finding a FANTASTIC thai place to eat. Nice comfortable meal and a thai ice tea, too.
Back to the hotel for more lobby-ing and a bit too much ponderous, depressing discussion. Worthy of note, when I start talking about massive distances and the heat death of the universe, it’s time to go to bed.
monday:
woke up late but managed a shower and manual dexterity. Got a lift to the venue from herlo and managed to be ‘together’ for things.
More packaging talks which we finally broke from and got a bit of time to talk about other things. Oddly, ended up talking about repositories with Martin Langhoff and Florian Festi for a while. Martin has an idea of how we can store the primary/filelists repodata that could work and could be a win for deltas but definitely needs some research, especially in terms of how fast it is to query.
Lots of folks getting the hell out of dodge at this point b/c the snow attacking the middle and north east of the country.
Eventually back to the hotel, saw an egypt protest (or support? I’m not really sure whose side they were on).
Downstairs in the lobby for a while talking to all sorts of folks. Got to meet Christoph Wickert after all these years. He mentioned my “cogito, ergo I ride I steel bicycle” shirt which made me smile. If there is any justice in the world we will have a summer fudcon in Munster, Germany where where he lives b/c it sounds lovely.
Waiting around a bit for mel’s brilliant suggestion of going to la bocca for dinner. Things to note about this:
1. fabulous salad
2. fabulous pizza
3. fabulous cookie
4. movies about tragically lost citrus fruit.
5. Xavier ate a calzone significantly bigger than his head.
Came back to the hotel and sat and talked for a while and then called it an early night b/c I was just zonked. Off to bed and hopefully to a flight in the AM.
Tuesday:
Get a phone call at 3am from my airline telling me they’ve cancelled my flight. Spend the next hour rebooking and sorting it out. I’m now on a flight 3 hours earlier – 7am instead of 10am. No time to go back to sleep so I shower, finish packing up and get on the shuttle. On the way to the airport I get another call. They’ve cancelled my 7am flight and have rebooked me on a 9:25am flight. Yes, that’s right.
I check in and go to the gate. Get online using the handy-dandy mifi and get some info from eunice that the airport I’m flying through (DFW) has been closed due to ice. So I talk to the people at the gate who give me a not-so-helpful answer. I get online, book a ticket through southwest that flies through nashville back to durham. Hop the shuttle down one terminal and settle in at southwest’s gate.
From here on out I have the must uneventful travel trip ever. Calm flights, I had a whole row of 3 seats to myself on both legs of the trip. It was nice. I get into durham a full 3 hours ahead of when I was supposed to get here originally. I take a shower, a nap and am able to slide into a seat with eunice at carrburritos.
Thus closes my fudcon 2011 adventure.
A couple of people to mention here:
1. rbergeron for not being able to talk to anyone herself but facilitating hundreds and thousands of valuable discussions.
2. ianweller for those rather stunning booklets he put together. Well designed and thought out.
3. luke macken for entertaining Dennis and I, much, much later after fudpub.
ipv4 addresses
I’ll get back to fudcon recap in a bit – but I wanted to ask a silly question about ipv4 addresses:
Why are people worried about address exhaustion in ipv4? If you take a look through the class-a allocations you’ll notice that 6 of the class As are allocated to the US department of Defense in one form or another.
So I’ve got to ask: why does the military need to have 96 million public, routable, ip addresses?
96 million seems like a largish number to me.
It is, in fact, 1 ip for roughly every 3 americans.
If they are giving an ip to every device they have, I could sorta understand it – but I really don’t want them giving public, routable, ips to everyone of them, especially the weapons.
in the same vein – someone could explain to me why mit needs 16 million addresses? What do they have 1000 public devices per student/prof/staff? Somehow I doubt that.
boggles my mind.
fudcon travel time + day 1
- Flight from DFW sat on the tarmac for just under 2hours getting the AC fixed. Mainly amusing b/c the guy in the same row as me spent a good portion of time convinced that the real problem was b/c of the technician fixing it was in a union. Deeply odd chain of logic
- Shuttle to hotel stops running at 10pm. My flight would have been in at 9:00 but now instead in at 10:40. Taxi to airport not that expensive so
- Mostly dead to the world for much of the first night but sat down and talked with ricky and ian and others in the lobby for a little bit. Inadvertently suggested evil things which ricky and ian immediately did and I’m still not sure Will knows about it yet.
- Got to look at the fudcon flyers/etc which are just ridiculously well thought out – including the ‘leave blank space so we can add a sticker with last minute updates’
Day 1:
- get to the big room. Meet up with James and finally meet Florian Festi. Spend a good deal of time talking about yum and depsolving with Florian.
- Talks get sorted out but we end up going for an impromptu meeting of Fedora Infrastructure – mainly so Smooge can get coffee but also so we can talk about what we are NOT doing and what we need to do.
- Lunch and a fabulous day sitting in the sun.
- Following lunch another impromptu meeting but this time of people who work in packaging at red hat. Valuable – lots of notes taken. Nice to know some targets
- Mmcgrath’s clouds will rule the world talk. Pretty much what I expected but the rhcloud stuff he demoed was pretty bad-ass.
- Cloudfs talk by Jeff Darcy. May be coming to a fedorapeople.org near you soon. It’s like AFS but w/o all the hair ripping and angst.
- Gave the much-longer-than-it-should-have-been-b/c-network-flaked-out version of how to start up ec2 and rackspace cloud host instances to ricky and toshio and ricky and anyone else who was near by – I think there is much we can do from there.
- Dinner, eventually, then back to the hotel for a lot more fedora infrastructure discussion and we kinda set some goals for our hackfest.
- Collapse in a pile in the room.
fudcon day -1
Leaving for tempe in a few hours. I’ve got a laundry list of things (and people) to talk to and I have this sinking feeling I’ll not be able to get to it all (or remember everything if I do). However, I’ve got a bunch of items prepared and I’m quite looking forward to the yum/rpm/packaging discussions and the Fedora Infrastructure meeting/hackfest/talk.
I’ll be getting in tonight around 9:50ish localtime – otherwise I’ll be in airports all day.
Those who may need my attention on something know how to get in touch.
func-command: handy in a pinch
during our outage and other fun on sunday/monday in fedora I had call to use func-command to give me some info. Thought I’d post that here.
needed to find all the hosts that were mounting a certain machine via nfs:
func-command ‘grep phx2-3.storage /proc/mounts’
returned that quickly.
ditto with finding out which systems were mounting it via iscsi.
Now – the objection here is ‘can’t you do this with ssh?’. That’s true except with ssh for many of the commands I would need to:
1. enable ssh as root or sudo first and muck with all that
2. and then parse that all back.
func-command handles that for me.
Also func records all the commands it runs in its audit log:
2011-01-23 22:45:54,966 – 10.1.16.10 puppet1 3550176746 command.run called with (‘grep phx2-3.storage /proc/mounts’,)
Thought it was cool and that I would mention it.
New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.
Keep up with the latest Advogato features by reading the Advogato status blog.
If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!
