gyptazy
@gyptazy@gyptazy.ch
Wrote one for BoxyBSD that uses the proxmox api of a cluster to rebalance VMs across all nodes within a cluster for a better resource and vm workload distribution.
Nodes and VMs can be filtered (e.g. you have special nodes (like NVMe-oF, PCI passthrough, cpu pinning,…) or VMs that make use of this of have some special configurations that should not be moved.
Rebalancing can be done by memory, storage (makes only sense on local storage) or CPU usage.
The rebalancing tries to balance all nodes in a most equal way.
@gyptazy I'd love to take a look at it. Got a link?
Let me write the docs and create pkgs, I’ll push it next week: https://github.com/gyptazy/proxmox-rebalancing-service
From Cloud Chaos to FreeBSD Efficiency
https://it-notes.dragas.net/2024/07/04/from-cloud-chaos-to-freebsd-efficiency/
#FreeBSD #jails #bhyve #GCP #AWS #kubernetes #OwnYourData #Linux #NoteHUB
@samurro I'd use lxc/lxd containers - or just plain docker/podman.
@stefano cool thank you for your answer, I am aware about docker/podman but just briefly googled LXC. In my current job we barely use any containerization. We just use KVM capabilites. The interesting thing on #freebsd is the ability to use #ZFS instead of relying on older filesystems in the #linux ecosystem. This has big implications for #backup strategy et cetera.
@samurro @stefano I could be wrong, but ZFS support is no longer a unique feature of BSD. OpenZFS is now a shared codebase and many Linux distros package it. Paired with #zfsbootmenu, it's a fantastic experience. I run zbm (root on zfs) with both a laptop and desktop. I also run a raidz pool on the desktop that replicates to an offsite pool attached to a raspberry pi. Aside from the USB enclosures I've connected to the pi, it's been very stable for years.
@ahoyboyhoy @samurro yes, ZFS is working on Linux, too.
On FreeBSD it's a first class citizen and has been for many years. This means that all the system tools are aware of it and that it's (more) tightly integrated into the OS
@stefano An uncanny parallelism with WordPress 🙈
@stefano i suspect that their next level in evolution will be switch from php to something else.
Being serious now, i'm really grateful for showing a way.
Wish to be good enough and work with stuff like this.
This for sure improves life and work of others.
Have a nice day!
@sourcerer thank you! Have a nice day, too!
@sourcerer @stefano If I had a PHP stack, I'd be really psyched about trying out https://frankenphp.dev/ -- the ability to smash your entire PHP app and server into a single binary is a great fit for these types of environments.
@stefano very interesting post! We've been using FreeBSD for a couple of years now mostly for crucial services (aka too important to ever fail), and I couldn't be happier.
On the other hand, looking at what k8s has given us for day to day production workloads, it is mostly the DevOps/GitOps part. Can you apply GitOps principles to the jails based approach you outlined in your blog post?
@riaschissl thank you! It can surely be done - but some more manual work, depending on what you need.
Also, have a look here: https://www.reddit.com/r/freebsd/comments/1bjggpi/has_anyone_noticed_the_great_podman_oci/
@stefano for some reason, every time I've tried getting Alpine vms to work on FreeBSD they fail to boot because they can't find the kernel, both with your blog post settings and with the default profile (I was using vm-bhyve at the time). Eventually I gave up and tried Debian, and it worked right away. Never managed to figure out why though.
@linus I'm now using UEFI and that is quite reliable. I've had the same problem with some old Alpine Linux releases
@linus this is one of my Alpine Linux's vm-bhyve config:
loader="uefi"
cpu="4"
memory="4G"
network0_type="virtio-net"
network0_switch="public"
disk0_type="nvme"
disk0_name="disk0.img"
@stefano managed to get uefi boot working at least after installing 'bhyve-firmware', but for some reason after running 'setup-alpine' it looks like it keeps booting from the ISO. do I need to manually disconnect it somehow? on boot, the log lists both "bhyve devices" (img file) and "bhyve iso device" (the alpine iso)
@linus yes, you need to remove the iso (just edit the vm .conf file "vm config vmname" and remove the iso )
@stefano There are lots of topics and classics in this story!
A WP plugin allowed remote code execution, it uses unmonitored autoscaling in the cloud, and dev team decides "we need moar power!". Classics on top of classics!
I believe this have to happen like every day around the world.
Now, I'd like to point that is not that Alpine/Linux is less secure than FreeBSD. It's just more popular, so vulnerable WP plugins execute Linux binaries if they can.
I am a firm defender of both cloud and self-host. In one hand, It's very hard to have the physical security and electrical/network redundancy of an actual datacenter. On the other hand, no cloud provider will offer truly data ownership, sovereignty and vendor independence . So, both approaches have their places.
I've seen AWS billings that would buy very nice hardware, every month. And salaries of "cloud experts" are no cheaper that old-school sysadmins. Specially if you have to pay their cloud certifications too.
I hope that manager learned the lesson.
@release_candidate Alpine Linux is perfectly fine (it's one of my favourite distributions).
I think that the "half way" solution is: get a server (maybe leased) inside a reliable datacenter but keep the control of the entire workflow. So you also know how to keep full and reliable backups of your data.
@gyptazy A shared image would be great. There is not a single source that provides this
@gyptazy Solaris gives me 'nam like flashbacks 🙈
@DesRoin lmao wut
@hyperreal trust me the old SunOS/Solaris systems were a pain 🤪
@DesRoin @hyperreal Slowlaris
@gyptazy Never studied history of Solaris etc. but i'm 100% sure once in my journey i heard about Open Indiana
@gyptazy Are you based on https://github.com/richlowe/arm64-gate/ ?
@gyptazy
Nice, the project needs.more devs. So Vagrant boxes are good!
@gyptazy very cool! One question : is that running in little endian mode? Or big endian?
Do you have any Fediverse/Mastodon client suggestions for iPhone? Reminders from last time okay. Gripes:
Mastodon: No auto update (Load missing posts)
Mammoth: Overall great but refuses to post photos
tooot: Jumps all over the place
Mona: Boost is a paid feature?!?
@dexter I've used Ivory and currently use Mona. I'm okay with paying keep development going on them.
I played around with Ice Cubes for a little while, but ended up back at Mona.
@dexter Using Ice Cubes. Not great, but most tolerated so far.
I use the web clients, saves as webapp for easy access
@dexter I quite like "Toot!" to be honest.
@dexter Ice Cubes works pretty well here.
Excited to announce the release of NotiMail 1.0!
📧 NotiMail now includes:
* Log rotation based on size or time for efficient log management.
* Thread-safe email processing for seamless multi-account handling.
* Enhanced configuration validation to ensure smooth operations.
* Continued support for multiple push providers including Apprise, NTFY, Gotify, and Pushover.
Upgrade now and experience extended battery life, swift notifications, and reduced data consumption.
Check out the full changelog and installation guide on our GitHub!
https://github.com/draga79/NotiMail
#NotiMail #Email #Notifications #Tech #OpenSource #Python #Apprise #NTFY #Gotify #Pushover #SQLite #OwnYourData
This looks interesting, but I have an unrelated question:
Where did the picture for the link preview come from? I don't see it anywhere in the repo.
I ask because I've been seeing a lot more of these in the link previews for text-based things and am curious if this is a new (ab)use of AI and the servers are just automatically generating them for social media now.
@suetanvil I have set the picture in the GitHub repository settings.
Latest 𝗩𝗮𝗹𝘂𝗮𝗯𝗹𝗲 𝗡𝗲𝘄𝘀 - 𝟮𝟬𝟮𝟰/𝟬𝟳/𝟬𝟭 (Valuable News - 2024/07/01) available.
https://vermaden.wordpress.com/2024/07/01/valuable-news-2024-07-01/
Past releases: https://vermaden.wordpress.com/news/
#verblog #vernews #news #bsd #freebsd #openbsd #netbsd #linux #unix #zfs #opnsense #ghostbsd #solaris #vermadenday
Latest 𝗩𝗮𝗹𝘂𝗮𝗯𝗹𝗲 𝗡𝗲𝘄𝘀 - 𝟮𝟬𝟮𝟰/𝟬𝟳/𝟬𝟭 (Valuable News - 2024/07/01) available.
https://vermaden.wordpress.com/2024/07/01/valuable-news-2024-07-01/
Past releases: https://vermaden.wordpress.com/news/
#verblog #vernews #news #bsd #freebsd #openbsd #netbsd #linux #unix #zfs #opnsense #ghostbsd #solaris #vermadenday
@vermaden maybe you can add https://www.qualys.com/2024/07/01/cve-2024-6387/regresshion.txt to your list for today?
Its already there from the beginning, should I add it 2nd time there? :)
@vermaden aah I found it with grep, a normal reading didn’t pick it up - sorry!
Enhancing FreeBSD Stability With ZFS Pool Checkpoints
Discover how to enhance the stability of your FreeBSD system using ZFS pool checkpoints, providing robust backup and recovery solutions.
https://it-notes.dragas.net/2024/07/01/enhancing-freebsd-stability-with-zfs-pool-checkpoints/
#FreeBSD #RunBSD #ZFS #DataRecovery #BackupStrategies #IT #SysAdmin #NoteHUB
Very interesting. I was knowing this feature, but without a "guide".
Thanks for sharing.
#FreeBSD #RunBSD #ZFS #DataRecovery #BackupStrategies #IT #SysAdmin #NoteHUB
@stefano It's not well documented, but zfs
also supports taking atomic snapshots of collections of datasets. The underling snapshot ioctl that the zfs
command uses takes an nvlist of snapshots to create.
This means that you don't need pool snapshots to create a consistent atomic snapshot of a set of filesystems and can create a snapshot of a subset of all of your ZFS filesystems (for example, all of the ones that contain system data, but not home directories, allowing you to roll back base-system + ports upgrades, without losing user data).
@stefano When doing zpool import --read-only=on --rewind-to-checkpoint <pool>, where does it mount to? If I can zpool import a checkpoint and have it have its root start at /backupdata or something like that, it would make my backup tool so much easier
@stefano I think I am misunderstanding this...I wonder if I could do zpool import -R /backupdata ... and have the read-only copy be at that location
Let’s continue to create great projects & code and contribute back to the #opensource community! Happy #coding! Have fun!
Hey #jq users,
Anyone knows how to turn a #JSON field containing HTML content into a clean text-only excerpt?
So far I did
jq -r '.[] | [ @uri "\(.status.content)" ]'
which gives me the URL-encoded content. But I don't get how to strip the HTML tags and only keep let's say 20 characters from that .status.content field.
Any clue?
is only content encoded or everything?
sed 's/<[^>]*>//g' could help to get rid of the html tags.
@gyptazy the whole content of one of the fields I gather. That's why I' d rather do it inside jq.
@joel Pipe with htmlq : https://github.com/mgdm/htmlq 😅
@ledeuns yeah, I' ve seen it. But I'd rather do it in jq to prevent another pipe/fork 🤔
@joel with jq i guess regexp is closest and we all know how that usually ends. but fq has a proper html parser so you can do something like this:
$ fq '.a |= ([html({array:true}) | .. | select(has("#text"))?."#text" | trim] | join(" "))' <<< '{"a": "<html>hello <b a=b>world</b></html>"}'
{
"a": "hello world"
}
{array:true} is an option to the html decoder to use nested arrays of ["element", obj with attr and text, children] instead of fancy but quite "lossy" object tree.
@wader so you mean I should replace the whole process with fq, instead of using jq? Or just using fq for the HTML sanityzing part?
@joel yes if want to parse html inside a filter then fq can do it. using jq and fq for just html i guess would be same as how you would use htmlq etc.
but i don't know what env this is running so maybe installing additional tools is a nono?
@wader seems like fq exist in pkgsrc so this can be an option :)
@joel 🥳
There is documentation about html at https://github.com/wader/fq/blob/master/doc/formats.md#html or via $ fq -h html
@gyptazy following this 🧵 👩💻
@gyptazy interesting 🤔
The Live/Warm Migration features are in the review and/or in the works - https://docs.google.com/document/d/1_0LEJqHT6v089YUI3PTRnNCLH8L14UZclvs2x_YWjtY/edit - details here.
@gyptazy I was looking at the vm-bhyve migration code just last night. It is not “live” but close to it, using ZFS snapshots to migrate the storage as efficiently as possible.
Every time this comes up on the calls:
1. Of course we all want it or think we want it.
2. If an HA environment, aren’t your apps already HA ready and can tolerate brief outages?
3. You generally need identical hardware at both ends. Did your cloud provider surprise you with a newer system, knowing you’ll be happy?
4. Using PCI Pass-Through? That’s generally a no-go.
5. I am probably forgetting some.
1. Yep and it’s every customer's first question when we talk about virtualization solutions and turns directly in a no-go. I’d really like to place FreeBSD based systems more often at customer's environments.
2. That often depends on the customer and his setups. But many of them did the way and also moved directly into the fully containerized world.
3. No, you do not need. You just need to ensure that you use the function set of the oldest system.
4. Sure, but that a special corner cases. We have the same with any other similar things like CPU pinning, SR-IOV…
@Toasterson @gyptazy @dexter @oxyhyxo That makes sense!
@Toasterson @gyptazy @bdha @dexter @oxyhyxo
That's definitely learned experience on their part. Triton has no shared-block-storage ("all storage is local") and no live migration. Now Oxide comes along and it has shared storage... why? To support live migration!
@danmcd @Toasterson @gyptazy @bdha @dexter @oxyhyxo I can confirm that from the beginning of the company a simplified chain of reasoning was:
- want to support automatic OS updates which need host reboots
- want updates not to disrupt guests, so that people don't need outages to update
- outage free reboots require us to live migrate workloads
- low-pause or ideally pause-free live migration requires storage to be replicated and accessed over the network
And thus: Propolis and Crucible!
@dexter @gyptazy @oxyhyxo I will say, having live migration support is really useful operationally because there’s always users who haven’t architected to support for even a couple seconds downtime.
But we’ve only got between 2/3 & 3/4 live migration support, mostly because of SR-IOV, ARM, Local NVMe etc.
And it’s not really feasibly in FIPS regulated regions because the overhead of memory sync with FIPS encryption is high.
@josephholsten @gyptazy @oxyhyxo Sadly, exactly that. My fear is that they’ll get a best a false sense of HA.
Don’t love, I cut them out and put them to my stamp collection :)
Thanks :) you should also fetch them :)
I love getting such mails! Thanks a lot for shipping them @mischa@exquisite.social & @OpenBSDAms@bsd.network!
PS: Just have a look at the stamp ❤️
#OpenBSD #FreeBSD #NetBSD #BSDCafe #BSDNetwork #BSDPub #OpenBSDAms #OpenBSDAmsterdam #BSDCommunity #Community
This was especially funny on mirrors and cdn archieves.
A modern, minimal, flexible, and easy-to-expand FreeBSD Jail manager built with love by experienced users for both neophytes and experts.
I tested it in the past, and liked it. Testing it again now and I like it. No databases, just the jail's dataset and a single .conf file to migrate or backup a jail.
Thank you, @antranigv
#Kubernetes #K8s #Container #Linux #NetApp #Trident #VM #SLES #Rancher
https://gyptazy.ch/blog/harvester-a-more-modern-alternative-to-proxmox/
You can find more about here:
Call Recording: https://m.youtube.com/watch?v=YhKQsPVM8sQ&t
Build + Bin: https://gyptazy.ch/howtos/howto-build-freedive-a-freebsd-gui-management-tool/
Written by @harshad@sharma.io - thank you very much!
#BoxyBSD is now sponsoring: Freedive
#Freedive is a beginner/hobbyist/tinkerer friendly, mobile-first web interface to run FreeBSD as a personal/public server.
Hey Thomas,
thanks a lot and happy to hear that it helps you and the open-source projects :)
People spending their spare-time and efforts into building and sharing valuable software should not be limited by missing resources. Depending on the needs, even small projects can easily cost too much money for the required hardware and everything behind it - just to keep it running.
Have fun and keep up the great work that you guys are doing! Thank YOU ALL
@BoxyBSD @stsp @gyptazy My thanks to @gyptazy for going out of his way to help. I can’t stress enough just how friendly and informative he’s been as we’ve been provisioning these machines.
This will really help #gameoftrees out, as I plan on having a manual CI system in place to compile changes across the three core BSDs ({Free,Net,Open}BSD).
Currently we’re relying on the infrastructure CirrusCI provides. This is still excellent, but it has been in contention of late with how much we’re able to use those servers. #boxybsd is going to help massively here.
I plan on writing up the process of how I’m going to use these machines.
Thanks again, @gyptazy
#BoxyBSD is now sponsoring the open-source project: Game of Trees
#GameofTrees (Got) is a version control system which prioritizes ease of use and simplicity over flexibility and mainly targeting #OpenBSD users.
Project: https://gameoftrees.org
CC: @thomasadam @stsp @gyptazy
Tags: #opensource #freebsd #openbsd #netbsd #GoT #gameOfTrees #VCS
#BoxyBSD is now sponsoring the open-source project: Telescope
#Telescope is a w3m-like browser for #Gemini.
Project: https://telescope.omarpolo.com
CC: @thomasadam @op @gyptazy
Tags: #opensource #freebsd #openbsd #netbsd #gemini #gem #capsule #gemcapusle #geminicapsule #geminiprotocol #gopher
Well done! I love your overviews! They’re technical and an eye catcher at the same time and always attracting people! Well done!
Longer text is also something what I encountered in the German translation, curios how it will look like rendered :)
Thanks for your work and efforts :)
However, currently I'm playing around with some completely different things like #Harvester - running #SLE with #Kubernetes, #Longhorn and #KubeVirt. I spawned a three node cluster just in minutes and could use live migrations.
Multi tenant solutions are a bit more tricky and the #Rancher integration is not that easy as expected and took additional 30 minutes. Currently, I like the approach of #HCI and also the more modern one which feels more like infrastructure as a code. I'm aware that this might not fit all scenarios and use cases and it's just having a look aside of the typical solutions. But it looks pretty promising and interesting.
#cloud #virt #virtualization #proxmox #vmware #esx #esxi #clonos #bhyve #jails #container #linux
https://gyptazy.ch/blog/clonos-an-alternative-to-proxmox-based-on-freebsd/
#bsd #virtualization #zfs #ipv6 #livemigration #bhyve #jails
Which slightly tends more to the second option. Do users prefer the possibility of easy adjustments (because it's easy to understand, test and re-use) or prefer performance aspects over it.
For business it might be different, so I might have a look into it again.
That's more a topic regarding the software lifetime cycle and feature set.
Working with static files to simply copy, you can of course also ship a whole dedicated Py environment. However, I would never do this because this wouldn't be upgraded in any way unless YOU also upgrade and ship it (and the users also updates it). Therefore, I make Py as a dependency on the underlying OS.
Currently, I'm more in Rust, but Go is also pretty nice.
Example: https://github.com/gyptazy/QualvoSec/tree/main/packaging/client
I think many prefers Python because it's easy to understand and to adjust/fix/add new features to many ones. Rust maybe for more performance and still easy to write. Go is probably a a perfect mix between Python and Rust. I'm not sure which lang should be my primary one when crafting new tools and apps.
OMG I just finished the FreeBSD Day live stream and I am still shaking. I tripped on words 🙈
(sorry folks, haven't been speaking in public much lately)
Wusstet ihr schon, dass Kindern bei uns niemals Geld im Weg stehen soll?
Unsere Freizeiten haben realistische Teilnahmepreise, aber niemand muss sie bezahlen. Eine formlose Mail mit Wunschbetrag ab 0 € reicht.
Falls ihr Kinder oder Jugendliche kennt, die sich über eine Tech-Ferienfreizeit freuen würden, aber die Familie denkt, es passe nicht ins Ferienbudget, sagt es weiter!
#Armutsbetroffene #IchBinArmutsbetroffen #FediLZ #EduNRW #EduRLP #Coding #Bildung #Kinder
Latest 𝗩𝗮𝗹𝘂𝗮𝗯𝗹𝗲 𝗡𝗲𝘄𝘀 - 𝟮𝟬𝟮𝟰/𝟬𝟲/𝟭𝟳 (Valuable News - 2024/06/17) available.
https://vermaden.wordpress.com/2024/06/17/valuable-news-2024-06-17/
Past releases: https://vermaden.wordpress.com/news/
#verblog #vernews #news #bsd #freebsd #openbsd #netbsd #linux #unix #zfs #opnsense #ghostbsd #solaris #vermadenday
Join us in celebrating #FreeBSDDay! Since June 19, 1993, FreeBSD has been a cornerstone of open-source innovation, powering systems and applications worldwide.
We have a week of activities planned to celebrate the developers, supporters, users, and contributors. We invite you to join us in the celebration by sharing your experiences.
#FreeBSDDay #OpenSource #Innovation #TechCommunity
https://freebsdfoundation.org/freebsd-day/
That shouldn’t be that much, see my German translation:
https://github.com/clonos/control-pane/pull/67
@lovisix@social.zdx.fr French?
@stefano@bsd.cafe / @tarastella@mastodon.bsd.cafe Italien?
@vermaden@bsd.cafe Polish?
@michel@social.recondo.com.br PT?
https://github.com/clonos/control-pane/issues/65#event-13191075473
Yes and no. I feel the same like @immibis@social.immibis.com tgat I only see interesting things that got boosted by my bubble (which shares the same mindset and interests). Without this, I would probably still follow 2 people. This works as a snowball system and increases my following count, which now makes it hard to see the really interesting things of people that I’m really interested to.
What I really hate is boosting useless stuff like „yes, thanks“,… it just annoys and blows up the timeline. I only boost something if it is (imho) really valuable
That’s my POV
Yes, you read that right. While the overall numbers might suggest growth, a deeper look reveals a worrying trend: the monthly active users on the Fediverse have plummeted to half of what they once were. Even the big profiles are feeling the pinch with dwindling interactions. Could it be that the Fediverse honeymoon is over?
Meanwhile, Twitter/X is seeing a resurgence. More and more users are flocking back, reigniting the platform with a surge of interactions. Is this the beginning of the end for the Fediverse, or just a bump in the road?
Let’s hear your thoughts!
#Fediverse #Mastodon #Federated #socialmedia #Twitter #X #Posting #Interactions #MAU #monthlyactiveusers #social #socialising #tech #techbubble #alternatives
More and more single user instances are joining the Fediverse while more and more users are leaving - sure, there’re coming more and more users, but the overall user count is nothing compared to the falling MAU count.
I can also see that more people are active in Twitter/X again. Also most interactions occur there which currently makes the Fediverse less interesting.
Friends of the #BSDCafe, #FreeBSD, #OpenBSD and #NetBSD enthusiasts and the entire #Fediverse,
Curious about #BoxyBSD?
Join Florian - @gyptazy - as he dives into its creation, architecture, and community support.
Perfect for BSD beginners and pros alike!
📅 10/06/2024
🔗 Watch now: https://www.youtube.com/watch?v=XEHL4skVq3U
🔗 More about BoxyBSD: https://boxybsd.com
🔗 More about the BSD Cafe: https://bsd.cafe
#BoxyBSD #BSD #OpenSource #VirtualMachines #BSDCafe #BSDCafeVideo #BSDPub #IT #VM #SysAdmin #RunBSD
Did I already mention how I loved back the days where we just had XMPP and all transports worked quite well.
Next week is FreeBSD week!
June 19th has been declared FreeBSD Day, and we'll be celebrating all week long. Each day, we’ll highlight FreeBSD content, user stories, and more! On FreeBSD Day itself, we will be streaming on YouTube as Kim McMahon interviews members of the FreeBSD community.
Stay tuned and join us in celebrating 31 years of FreeBSD!
I’m not sure if this reaches you, because also the postings on social medias stopped. However, maybe someone has some more information - and if you are feeding this I just want to say thank you for the time and efforts you put all the time into this project and I also hope you’re doing well so far (because we haven’t seen you for a while anymore).
Sometimes things get too much or our focus simply changes - it’s ok! Just hope you’re good so far!
I created some ready-to-use container images for #RV64:
* #Debian Trixie
* #Ubuntu Jammy (22.04)
* Ubuntu Mantic (23.10)
* Ubuntu Noble (24.04)
* #Fedora 39
https://gyptazy.ch/misc/riscv-container-images-for-podman-docker/
#riscv64 #rv #risc #docker #podman #images #containerreg #registry #userland
We have passionated people and developers that were brought together by the BSD Cafe, sharing the same mindset, having the same ideas - boosting up the community. Especially the #BSD one and also trying to attract #FreeBSD, #OpenBSD, #NetBSD and all the other awesome flavours to beginners. We see people working and developing new things to make beginners life easier, using #BoxyBSD service for this (like @harshad@sharma.io does with #Freeway), syncing in the #BSDCafe Matrix chat and discussing those things in the #BSDPub calls with the community. Providing deep detailed information, like you did yesterday with your blogpost where #Proxmox on #Linux got compared to #bhyve on #FreeBSD (both running on #ZFS, but also #LVM/#EXT4) are the important things to boost the visibility of BSD! Also keeping up and together the community is very important. Without this place, many things would have never happen!
Everyone puts his time and efforts in pushing the community, to make it a little bit better by each day.
So, thank you and everyone here!