news aggregator

A large change found its way into the Opensolaris Code base: Headlining with with integration of "PSARC 2009/534 SMB/CIFS Standalone DFS" a whole load of features and fixes were put back to the kernel.

Surely the integration of the Distributed File System feature is the most interesting one, enabling the user to host the root of a DFS filesystem on a Solaris Server with the in-kernel CIFS implementation, enabling the admin to separate parts of a CIFS directory tree to separate servers.

Vem aí a 1ª Semana de Educação e Artes Digitais, uma realização da Casa da Arvore – Projetos Sociais que reunirá do dia 5 a 9 de abril deste ano em Palmas, pesquisadores, educadores, alunos e artistas representantes de projetos inovadores desenvolvidos em várias partes do Brasil. São iniciativas que têm provocado profundas reflexões sobre o universo de possibilidade de produção de conhecimento na era digital, através da arte e da educação. A programação conta com três oficinas voltadas para artistas, estudantes, profissionais liberais e demais interessados em experimentar novas possibilidades de criação a partir de novas tecnologias de informação, comunicação e mobilidade. Para fazer sua inscrição envie um email com o nome, idade, telefone, área de atuação para os e- mails abaixo e aguardar instruções:

Jornalismo na web 2.0 – Nacho Duram (SP) jornalismoweb2.0@gmail.com

Animação Básica em Blender 3D – Erick Henrique (TO) anima.blender.3d@gmail.com

Processamento de som e imagem em tempo real – Jarbas Jácome (PE) processamento.tempo.real@gmail.com

Os painéis serão gratuitos e aberto a toda sociedade.

As oficinas serão cobradas R$ 10,00 e o limite de vagas é 15 pessoas por oficina.

1ª Semana de Educação e Artes Digitais – Palmas – TO
de 05 a 09 de Abril na Av. Teotônio Segurado, 402 Sul, Área Verde, Palmas-TO

An old school SA practice...

This is probably the geekiest blog title I've used - but today's blog is a short look at two variations on the old sysadmin practice of separating /var from /, inspired by recent "how do I do this?" calls. Why do it? How was it done before?

This was traditionally done to ensure that growing space consumption in /var, perhaps caused by core, log or package files, didn't exhaust critical parts of your file system. This could happen because some program kept dumping core or generating log entries. Exhausting space would add further injury by causing other failures.

Several techniques can prevent such problems. One method is to use coreadm to put core files somewhere else, and to use logadm and /etc/logadm.conf to rotate log files on a schedule that is consistent with your disk space and retention policy. But, the biggest hammer and most complete solution was to keep /var in a separate file system by giving it a dedicated UFS file system on its own disk slice. That way, even if something went amok and filled /var, it had no effect on other file systems.

The disadvantage, of course, is the hassle of creating and sizing separate disk slices. You had to plan how many slices you needed and how big they were, and if you got them wrong it was really inconvenient to change them. You might have one slice and file system too big, wasting space you really needed in another slice, but reallocating space was a drag. Having storage allocated into little islands was a real time-waster, especially on the itty-bitty disk drive capacities we used to live with.

ZFS, and in this case ZFS boot, pretty much eliminated this inconvenience - as I'll discuss in a moment. But now, an old joke...

Before I go into the two examples that came up, a classic joke from mathematics or science class.

The professor is in the front of the classroom and writes an equation on the blackboard (I'm picturing the professor I had when studying Fourier transforms in EE class, but I won't try to do his accent.) Pointing to it, he tells the class, "As you can see, this theorem is clearly trivial."

Turning back to the blackboard he pauses for a moment, puts his hand on his chin and says "Hmmm.... just a moment." Now he starts working on the equation's derivation, covering blackboard after blackboard with equations - everything from &#945 to &#969. He fills all the blackboards in the classroom, mumbles "excuse me, I'll be right back," and then goes into an adjacent empty classroom to use its blackboards.

Twenty minutes pass. Finally, the professor returns to the classroom. He beams at the students with a big smile and says "I was right. It is trivial!"

I think this may be relevant to the rest of the post! :-) The trivial case with ZFS root file system

I'll start with the straightforward case first. I was contacted by a long time friend (who has exceptional knowledge of Solaris and other operating systems, but is new to ZFS) who had wanted to restrict /var for a fresh installation of Solaris 10 that he had just done. He used ZFS boot and selected the option that allocated a separate ZFS dataset for /var, and wanted to know if there was an easy way to control its size.

I thought that this should be easy with a ZFS quota, but to be sure, I brought up a new instance of Solaris 10 under VirtualBox to run through the steps and get the right ZFS dataset name. I allocated the separate /var (it's an option you specify during install), and after installation completed I logged in and issued the following commands: # zpool list NAME SIZE USED AVAIL CAP HEALTH ALTROOT rpool 15.9G 4.16G 11.7G 26% ONLINE - # zfs list NAME USED AVAIL REFER MOUNTPOINT rpool 4.62G 11.0G 34K /rpool rpool/ROOT 3.12G 11.0G 21K legacy rpool/ROOT/s10x_u8wos_08a 3.12G 11.0G 3.06G / rpool/ROOT/s10x_u8wos_08a/var 65.6M 11.0G 65.6M /var rpool/dump 1.00G 11.0G 1.00G - rpool/export 265K 11.0G 23K /export rpool/export/home 242K 11.0G 242K /export/home rpool/swap 512M 11.5G 42.0M -

Right - all I should need to do is set a quota on rpool/ROOT/s10x_u8wos_08a/var, so let's do that. I picked a quota slightly larger than the amount of space already consumed so I could easily test filling it up by creating dummy files with random data. I did that once to make sure I didn't mess up the syntax, and once more in earnest to exceed the quota: # zfs set quota=80m rpool/ROOT/s10x_u8wos_08a/var # zfs get quota rpool/ROOT/s10x_u8wos_08a/var NAME PROPERTY VALUE SOURCE rpool/ROOT/s10x_u8wos_08a/var quota 80M local # dd if=/dev/urandom of=/var/XX1 bs=1024 count=10000 10000+0 records in 10000+0 records out # zfs list rpool/ROOT/s10x_u8wos_08a/var NAME USED AVAIL REFER MOUNTPOINT rpool/ROOT/s10x_u8wos_08a/var 75.3M 4.67M 75.3M /var # dd if=/dev/urandom of=/var/XX2 bs=1024 count=10000 write: Disc quota exceeded 4737+0 records in 4737+0 records out # # ls -l XX* -rw-r--r-- 1 root root 10240000 Mar 17 14:15 XX1 -rw-r--r-- 1 root root 4849664 Mar 17 14:16 XX2 # zfs list rpool/ROOT/s10x_u8wos_08a/var NAME USED AVAIL REFER MOUNTPOINT rpool/ROOT/s10x_u8wos_08a/var 80.1M 0 80.1M /var

Mission accomplished: the second file reached the quota allocated to this ZFS dataset as required. The only odd thing (in my opinion) is the odd spelling "Disc" instead of "Disk" in the message write: Disc quota exceeded. So, if I'm building a Solaris system and want to keep /var from exhausting disk space, all I need is one command to set the quota. Sweet. A less trivial case, with zones

Shortly after the preceding example, I was contacted by a customer who wanted to do something similar to control /var within Solaris Containers. He tried to create the zone with /var defined as a delegated ZFS file system using legacy mounts. There seems to be a chicken-and-egg situation about what parts of the zone's filesystem must already be mounted before the zone can boot, but then you can't delegate it to the zone. Instead, I created a ZFS dataset and assigned it to the zone's /var: # zfs create rpool/zones/vartest # zfs list rpool/zones/vartest # cat varzone.cfg create set zonepath=/zones/varzone set autoboot=false add net set physical=e1000g0 set address=192.168.56.164 end add fs set dir=/var set special=/zones/vartest set type=lofs end add inherit-pkg-dir set dir=/opt end verify commit # zonecfg -z varzone -f varzone.cfg # zoneadm -z varzone install A ZFS file system has been created for this zone. Preparing to install zone <varzone>. Creating list of files to copy from the global zone. Copying <2899> files to the zone. Initializing zone product registry. Determining zone package initialization order. Preparing to initialize <1062> packages on the zone. Initialized <1062> packages on zone. Zone is initialized. Installation of <2> packages was skipped. The file </zones/varzone/root/var/sadm/system/logs/install_log> contains a log of the zone installation.

So far so good. After booting the zone without incident, I set a quota and fill it up (note: this is a much bigger /var because I'm building a zone in a Solaris instance with a bunch of additional software in /var/sadm/pkg ) # zfs list rpool/zones/vartest NAME USED AVAIL REFER MOUNTPOINT rpool/zones/vartest 274M 9.31G 274M /zones/vartest # zfs set quota=300m rpool/zones/vartest

Within the zone, I exhaust allocated space using the same method as before: # dd if=/dev/urandom of=/var/xx1 bs=1024 count=100000 write: Disc quota exceeded 26369+0 records in 26369+0 records out

So, I was able to create a separate /var for the zone, and manage its space independently from the zone's root. WARNING: I do not know if this is a supported or recommended procedure, even though it seems to work. My recommendation is that it's more important to impose a quota on the zone's ZFS-based zone root, in order to control its total accumulation of disk space. That protects other zones and other applications that may be using the same ZFS pool.

Conclusions

Separating /var was especially important with the small boot disk capacities we had to work with in Ye Olde Days, and perhaps became less important with the large disks we have now. However, this becomes important again due to the availability of relatively low capacity Solid State Disk (SSD) boot drives being used for fast local boot disks with low power consumption, and because of virtual environments in which a single Oracle Solaris instance might host many containers, each with its own /var and pattern of space consumption.

So, maybe this is a useful Old School idea that has new, and slightly different relevance today.

RPM of the ffmpeg-php extension is available in remi repository for Fedora ≥ 11.

This extension works with ffmpeg version 0.5 available in RPM Fusion (free) repository and PHP 5.3.2. This RPM is build from the SVN snapshot (version 0.6.3, revision 676). Installation : yum --enablerepo=remi install php-ffmpegFor now, the old version 0.4.9 available in old fedora version (and EL) is not compatible. I'm waiting for your... Lire php-ffmpeg-0.6.3

Olá amigos humanos
Venho com muita alegria divulgar os primeiros trabalhos finalizados por nossos alunos do núcleo TelinhaAnimada. Com isto finalizamos a primeira fase da oficina de ilustração 2D digital com Inkscape. Espero que gostem dos resultados, e dêem boas vindas aos novos humanos criativos que entram no nível LARVA do GEDJA. Parabéns pessoal, estou orgulhoso de todos!

Galeria TelinhaAnimada

El día de ayer salio la nueva edición de Tuxinfo con la segunda nota de Fedora Proyect hecha por nuestra comunidad.
En este caso el tema es el Usuario objetivo de Fedora.
Pronto también estaremos participando del podcast semanal de la revista
Bájate la revista digital desde aqui

To avoid any unnecessary kitten killing, use Tech Talk PSE instead.

(from BoingBoing)

Hadoop AvatarNode

HDFS clients are configured to access the AvatarNode via a Virtual IP Address (VIP)
When PrimaryAvatarNode is down,  the Standby AvatarNode takes the relay
The Standby AvatarNode ingests all committed transactions because it reopens the edits log and consumes all transactions until the end of the file
The Standby AvatarNode finishes ingestion of all transactions from the shared NFS filer and then leaves SafeMode
The VIP switches from Primary AvatarNode to Standby AvatarNode

In "blue" the AvatarNode before the fail
In "red" the AvatarNode after the fail
The servers are alternatively Primary or Standby AvatarNode.

See the publication from Dhruba Borthakur

          

Mahaswami Software (based in Bengaluru, India) uses its homegrown Rapid Application Development framework to deliver quality applications in quick time. The framework leverages JRuby, Rails, and the J2EE platform along with Test Driven Development and Continuous integration tools. Mahaswami offers product development services and specific consulting on JRuby/Rails based application development. The Mahaswami team actively contributes back to the Ruby and Rails community.

And they picked GlassFish for a web-based supply chain management product for a large enterprise application service provider in India. They picked JBoss instead of GlassFish because they loved the web-based admin console and high performance.

Watch more details in the following video:

Here is what the customer (Sify Technologies) has to say about their experience:

We were pleasantly surprised by this team's fantastic ability to deliver complex solutions with great agility, and have gained an edge to our product development efforts.

Do you have any JRuby/Rails/GlassFish consulting requirements in Bengaluru ? Mahaswami Software is your one stop shop for providing all the services.

See other similar success stories here.

Technorati: mahaswami bangalore bengaluru india jruby rubyonrails glassfish stories

Mahaswami Software (based in Bengaluru, India) uses its homegrown Rapid Application Development framework to deliver quality applications in quick time. The framework leverages JRuby, Rails, and the J2EE platform along with Test Driven Development and Continuous integration tools. Mahaswami offers product development services and specific consulting on JRuby/Rails based application development. The Mahaswami team actively contributes back to the Ruby and Rails community.

And they picked GlassFish for a web-based supply chain management product for a large enterprise application service provider in India. They picked JBoss instead of GlassFish because they loved the web-based admin console and high performance.

Watch more details in the following video:

Here is what the customer (Sify Technologies) has to say about their experience:

We were pleasantly surprised by this team's fantastic ability to deliver complex solutions with great agility, and have gained an edge to our product development efforts.

Do you have any JRuby/Rails/GlassFish consulting requirements in Bengaluru ? Mahaswami Software is your one stop shop for providing all the services.

See other similar success stories here.

Technorati: mahaswami bangalore bengaluru india jruby rubyonrails glassfish stories

Sir Ken Robinson on education and creativity /link to talk

So the FAD was epically awesome, and Paul sent us all home with gadgets This was the toy that I wanted the router bridge set up for; the bridge worked wonderfully after reviving the not dead(1) linksys WRT54Gv5 and putting dd-wrt micro on it. Anyways, yes, SIP Phones for all! And real cool ones at that. I got a Polycom IP 501:

It’s a nice, _extremely_ hackable phone, and works wonderfully with Fedora Talk. I guess I have to add some stickers to it too

(1): I told konqueror to remember the wrong password and username combo and for whatever reason the router never spit out a 403 forbidden, leading me to believe it was dead. oops!

Two Configuration Methods to Rule Them All

The phone is primarily designed to be used in corporate environments, and judging by the main configuration methods, this is (sometimes painfully) clear. You can choose between either setting it up via web gui or by hacking the configuration files and setting up a boot server

WebGUI is fun, i guess

I started off trying to configure this bad boy using the Web GUI since it came with the SIP application preinstalled into the OS (more on that below).

The Web UI is pretty nice, if a little unwieldy. I noticed that it takes about two minutes after the phone boots (that is strange to say ) for the Web UI to start up and become accessible. At this point, I put in all of the details taken from the fedora talk setup pages, and put them into the forms:

This mostlyworked. I was able to place outgoing calls, but the phone refused to register with Fedora Talk, and I appeared as Unavailable to everyone.

ZOMGWTFTFTP

At this point, after trying for a while to make the WebUI bend to my will, I gave up and decided to wipe to Factory Defaults. Unfortunately, I didn’t realize how this would effect my phone: without a boot server, the phone was at that point a fancy paper weight. FAIL. So, it was time to learn how to set up a bootserver using tftp, which wasn’t that hard as it turns out…. You have to, of course, find the boot rom and applications on the Polycom website, which was a fun task. You’ll want to download "SoundPoint IP and SoundStation IP SIP 3.1.6 Legacy [Combined]" then run

yum install tftp-server

chkconfig tftp on

mkdir tmp && cd tmp

unzip ../spip_ssip_3_1_6_legacy_release_sig_combined.zip

cp -rf * /var/lib/tftpboot

cd ../ && rm -rf tmp

At this point, you have a working TFTP boot server. congratulations! When you boot your phone, before it has the chance to autostart, jump to the setup dialog and enter the default password (456). At this point, you can set up its network settings; i left DHCP stuff and the IP Gateway alone, but on the Server menu, you’ll want to set the Type to Trivial FTP, and the Server Address to the address of, you guessed it, your server! Leave username and password blank. Now allow it to boot, it should load and install the application, and spit you out at an…. unconfigured phone! yay! Wait, what?

The config files from the deep

In the root of your tftp are a number of config files… The design of the entire system was to allow the config files to cascade to make it easier for enterprise deployment. This is cool and all but….

1. FlashROM config
2. <MAC-address>.cfg
3. 000000000000.cfg
4. phone1_316.cfg
5. sip_316.cfg

That’s a lot of config files. So, I basically did all the configuration in sip_316.cfg and 000000000000.cfg which is recommended against in case you ever update your bootrom. Oh well, I don’t. So I just tinkered with sip_316.cfg and phone1_316.cfg…. Basically, these are GARGANTUAN xml files describing everything from the SIP login details, to the frequency of the dial tone, to the shading of the buttons. omg. I’ll just point you in the right direction in these files, but you can pretty much do anything you want with them just by looking at the various entries’ names.

phone1_316.cfg contains the server details in phone1->reg->reg.1.*:

<reg reg.1.displayName="Fedora Talk" reg.1.address="sip:rrix@fedoraproject.org" reg.1.label="ftalk" reg.1.type="private" reg.1.lcs="" reg.1. csta="" reg.1.thirdPartyName="Ryan Rix" reg.1.auth.userId="rrix" reg.1.auth.password="$WITHHELD$" reg.1.auth.optimizedInFailover="" reg.1. musicOnHold.uri="" reg.1.server.1.address="talk.fedoraproject.org" reg.1.server.1.port="5060" reg.1.server.1.transport="DNSnaptr" reg.1.server.2. transport="DNSnaptr" reg.1.server.1.expires="" reg.1.server.1.expires.overlap="" reg.1.server.1.register="1" reg.1.server.1.retryTimeOut="5" reg.1. server.1.retryMaxCount="5" reg.1.server.1.expires.lineSeize="" reg.1.server.1.lcs="" reg.1.outboundProxy.address="talk.fedoraproject.org" reg.1. outboundProxy.port="5060" reg.1.outboundProxy.transport="DNSnaptr" reg.1.acd-login-logout="0" reg.1.acd-agent-available="0" reg.1.proxyRequire="" reg.1.ringType="2" reg.1.lineKeys="3" reg.1.callsPerLineKey="1" reg.1.bargeInEnabled="" reg.1.serverFeatureControl.dnd="" reg.1.serverFeatureControl.cf="" reg.1.strictLineSeize="" reg.1.tcpFastFailover=""/>

Which leaves us with:

1. displayName can be whatever you want
2. address is sip:fas_username@fedoraproject.org
3. label can be whatever you want, it’s what is displayed on the phone
4. thirdPartyName is what appears to other users when you call them
5. auth.userId is fas_username
6. auth.password is the VoIP Password, NOT your FAS password.
7. server.1.address is talk.fedoraproject.org
8. server.1.port is 5060
9. outboundProxy is talk.fedoraproject.org 5060

I deleted the config details for server.2. and server.3. since they didn’t matter to me.

Then in sip_316.cfg I enabled SNTP support so that the time was correct…

<SNTP tcpIpApp.sntp.resyncPeriod="86400" tcpIpApp.sntp.address="tick.ucla.edu" tcpIpApp.sntp.address.overrideDHCP="0" tcpIpApp.sntp. gmtOffset="-7.0" tcpIpApp.sntp.gmtOffset.overrideDHCP="0" tcpIpApp.sntp.daylightSavings.enable="1" tcpIpApp.sntp.daylightSavings.fixedDayEnable="0" tcpIpApp.sntp.daylightSavings.start.month="3" tcpIpApp.sntp.daylightSavings.start.date="8" tcpIpApp.sntp.daylightSavings.start.time="2" tcpIpApp. sntp.daylightSavings.start.dayOfWeek="1" tcpIpApp.sntp.daylightSavings.start.dayOfWeek.lastInMonth="0" tcpIpApp.sntp.daylightSavings.stop.month="11" tcpIpApp.sntp.daylightSavings.stop.date="1" tcpIpApp.sntp.daylightSavings.stop.time="2" tcpIpApp.sntp.daylightSavings.stop.dayOfWeek="1" tcpIpApp. sntp.daylightSavings.stop.dayOfWeek.lastInMonth="0"/>

address is the address of the NTP server you want to use. I left the rest of it alone.

A few final words

I suppose that the way I did this (with a TFTP boot server to configure the phone) was overly complicated in the long run, it gave me a _huge_ amount of control over what I could do with the device. Digging into the phone will be fun; it looks like the main sip application is a Java applet with the source code included. Sweet, if only I enjoyed hacking Java

I’ll enjoy hacking some of the other parts of this phone, and will detail them when I do… I plan on adding a Fedora background image to the phone’s display, for one, and possibly a page for the microbrowser which can query FAS for extensions. More on that later, though

In the mean time, I’ve cooked up a little script to general the XML dictionary files used as the Address Book within these phones. After you edit the file to add your FAS account information and the list of folks you want to be in the book, it will spit out a 000000000000-directory.xml, which you should copy to the root of your tftp server and rename it to match your phone’s mac address (available from menu->status->ethernet).

More later!

=-=-=-=-=
Powered by Blogilo

OpenOffice.org是Linux平台上广泛使用的办公套件，各大Linux发行版都的软件库都有OpenOffice.org，而且不少发行版已经把OpenOffice.org列为标准组件。虽说都是同一套源代码，但是各个发行版对OpenOffice.org都略有区别。例如Debian、Ubuntu、openSUSE等发行版都集成了Go-Oo的补丁，而Fedora则是完全利用OpenOffice.org的初始源代码重编译，仅仅加入一些修复bug的补丁，基本不会对性能进行优化和改进功能。

以下操作都是在终端里的root用户完成。

安装Go-Oo首先要卸载原来的OpenOffice.org

yum remove openoffice*

添加Go-Oo的yum源，到这里下载一个rpm包，然后安装。如果是x86_64的用户，还需要修改一下repo文件，把/etc/yum.repos.d/GoOo.repo里baseurl的链接修改为http://go-oo.mirrorbrain.org/stable/linux-x86_64/ 。

紧接着安装openoffice.org-ure 1.6.0

yum install openoffice.org-ure-1.60

随后需要添加一个exlude到yum.conf，因为Fedora自带的openoffice.org-ure的版本号跟随OpenOffice.org的版本号3.2.0，比1.6.0高很多，但是安装前前者无法让Go-Oo运行，所以需要排除这个软件包。

echo “exclude=openoffice.org-ure” >> /etc/yum.conf

然后就是正式安装Go-Oo：

yum install openoffice.org3-zh-CN.x86_64 openoffice.org3.2-redhat-menus ooobasis3.2-{write,calc,impress,math,draw,base}

{}里面的都是OpenOffice.org的组件，用户可以根据需要来安装。稍等片刻，安装就会完成了。

使用Go-Oo替换，速度上明显会感觉到提升。特别是使用了微软模板的演示文档，速度和兼容性有了质的飞跃。所以我还是渴望Fedora能够整合Go-Oo的补丁。Fedora原版和Go-Oo似乎都没有解决到与Gnome的字体渲染一致的问题，虽然我已经安装了OpenOffice.org的Gnome界面整合包，还是请牛人帮忙解决。

My last day of employment at Red Hat was March 17th.  Deciding to leave Red Hat was the hardest decision of my life.  Red Hat and Fedora has been my mission and obsession for my entire working career (plus a few years in college.)  But I decided that it is time for me to depart for full-time graduate school later this year in order to further my personal growth.  I am excited about studying toward a MBA degree.  I already took the GMAT exam, and currently I am in the process of applying to schools in the Boston area.  Meanwhile I will be helping my parents with the family business, studying classical music, and getting some much needed rest.

This big personal decision forced me pause and think not only about my own career, but also how far Fedora has come since I was originally hired over 6 years ago.  Red Hat hired me early 2004 to grow the community that was started as an undergraduate project back during 2002.  The early years of Fedora were a bit painful and confusing to both people within Red Hat and in the community.  Red Hat as a company had much internal confusion about the importance of engagement with the community.  This is in stark contrast to today where the company now clearly understands the importance of community engagement and prioritization of resources toward these community ecosystems, both upstream and home-grown at Fedora.

Reading my comments from over 3 years ago it seems that the goals of Fedora have not changed much.  The goals may not have changed, but the pace of succeeding at these goals has increased.  For the past several years Fedora has been the exemplification of the goal: Rapid Progress of Free and Open Source Software.  ext4 default, IBUS, Kernel Mode Setting, Nouveau 3D acceleration, Slick new mobile broadband interface in NetworkManager, KVM virtualization, GNOME Shell are only a tiny sample of features that happen in Fedora before any other distribution.

Fedora has long ago surpassed my initial goal in 2002: Allow developers to work together collaboratively to improve a centralized repository of packages.  Back then I couldn't have imagined the size and scope of the project today.  I feel so proud of what the team and the community has together accomplished.  Great job everyone!

While I am leaving Red Hat in order to further my personal growth, I will continue to be part of the Fedora Project.  I will continue to be a co-maintainer of pidgin and spamassassin, and occasionally help out with Release Engineering and Infrastructure.  I also have a prototype project in mind to eventually revamp LTSP in Fedora with a far more efficient approach to thin clients and remote desktops.

What is next for me?  Things are extremely busy for the next 2 months.  Then I will be taking some classes in June learning how to repair musical instruments.  School begins in September.  I can't wait! =)

If you’ve ever talked to me about the topic of weight loss, you’ll know that for a span of fifteen months, I walked around Madison SD for various things including employment and consumed a lot of water.

I had a very good routine going on during that time. I managed to walk 8 miles a week just to get to and from one job. I also moved around a lot at a second job where I was a housekeeper for a four seasons resort. In addition to moving around so much, I practically had a water bottle taped to my hand, including many more at home for easy switching. Although I don’t remember my eating diet that well, I imagine it was decent, though not the best. I can’t help but imagine what all these did for my budget. GIven the fact that this was during the height of our high gas prices, I likely saved myself a lot of money.

What does this all mean you ask? During those fifteen months, I managed to shed 40-50lbs without a lot of extra effort. Surely if I did some sort of running or gymrat work, the process would have gone a lot faster, but I wanted to be lazy and most of the methods I used I “had” to do. I knew that I had to go to work so that made a lot of motivation to do the walking.

This all changed in August of 2007. It’s at that point where I decided to move away from Madison and back in with my parents for a “break”. This completely broke my beloved routine. I no longer had my motivators to push me along and I would need to make myself do it. Ask anyone who has tried and they’ll say that’s not always an easy task.

Move forward to January of 2008. It was at this point that I made the move over to Sioux Falls SD. While not my ideal situation, I did have an instant job at one of the local KFC’s and it helped me to stay afloat while I looked for other work along the lines of my studies from college. Naturally, being around that kind of food and being subject to it on a full time work schedule basis didn’t really help myself diet wise. However, it wasn’t devastating because I was still up on my feet consistently and moving around. I can confirm that some pounds slowly crept back on me, and I was there for a little bit over a year.

After my time ended at KFC, I moved onto my current position as a front desk security officer at Citibank SD. At the time, I thought the new environment would be a good thing for me and I’d enjoy it a lot better. The ways I was wrong with this is that I am proverbially tied to the front desk and can not roam freely around beyond a small radius, including restroom breaks whenever I want. That caused me to need to limit my water consumption rate. I am fully free to roam and walk around the lobby of the building I am in, but there isn’t much space for trying to get a heart rate going. The last negative thing with this job is that it has an early start time before the sun is risen, and I’m naturally a night owl. That said, I tend to consume things that will help prevent myself from nodding off, thus food/beverages with sugar or caffeine.

Although I’ve recently already taken some strides to help curb this issue, I weighed myself tonight and the scale returned the weight of 230lbs. I hate to admit that I’m back to the weight I was at when I first started to lose it back when still in Madison.

So, at this point I have to ask what exactly can I do to help reverse this gain, in my current environment. I am starting to go for quick walks when I get breaks at work. The big one would be my lunch break. I’m starting to do quick walking tours of the buildings at work and opting out of stopping by the cafeteria for something to eat. This gets me up on my feet and moving, as well as a fringe benefit of not spending some money on food. During this walk, I use staircases as much as I can and never rely on elevators. I am also looking to take shorter path walks during my afternoon break shortly after shift change or some of the other officers. On top of “at work” walks and with spring/summer coming up, I want to go for walks around my neighborhood during my off time. I already took one 2 mile walk this afternoon and will try my best to stay motivated to do that regularly.

Adding to my efforts to walk more, I hope to increase my water consumption away from work, as well as reduce extraneous consumption of sugars while at work. To counteract the risk of nodding off, I have recently started bringing in work-safe podcasts on my audio player. The benefits of this idea is that they require active listening, instead of passive listening that music can produce. The music can fade into the background of my mind and just be there. If I’m having to pay attention to people talking i order to comprehend, this keeps my mind more focused and awake. Finding a regular bedtime that allows me to get proper sleep could also go a long ways in helping at work. Hopefully I can start also altering my food diet away from work as well. Who couldn’t eat a bit better when at home?

Through these actions that I’m already enacting or planning to enact, as well as any suggestions that you readers can produce, I hope I can once again curb this weight issue and perhaps even start bringing it back down to what it used to be. I please urge any of you to submit your ideas to me either through a comment below or use a way to contact me from my contact page. I’ll definitely consider any and all suggestions.

As I wrote some minutes ago over some Fedora mailinglists I had the idea of having some kind of Fedoraproject Band thingy. I know there are often geeks like me around who also have some kind of creative blood. So maybe its a good idea to put the Fedorians all over the world together to one band (reminds me of “one band, one sound…” anyway).
Paul encouraged me to init that idea and Mel also gets that shiny eys as I mailed her some time ago. So, lets ROCK IT!
So if there are others around who want to produce some cool tunes together take a look at the Fedora Band wikipage for more info.

Oh, and it haz nottin to do wid thaz band… as far as I know

After talking with a few people recently and doing some self-analysis, I feel like it’s time to make a major shift in what I do within the Fedora Project. My Fedora résumé so far has consisted mostly of wiki czaring,1 package maintenance and other odds-and-ends jobs others kindly ask me to do.

I’m presently concerned with the second in that list — a combination of increased stress and decreased time available due to school and the speed of discussion on package maintenance and release engineering is a losing game. In the next few weeks, I’ll be checking all of my packages and determining which ones have dead or slow upstreams or bugs that I can’t resolve on my own. Those packages will likely be orphaned, and if nobody wants to care for them, so be it.

The two others? Wiki czaring is fine, but I need to improve on it a bit (see the footnote), and I always enjoy the random problems that I can help quickly solve for people. This being said, development on mw, supybot-fedora and other convenient software is (hopefully) Not Going Away™ any time soon.

With the pushing away of my first Fedora love, package maintenance, I’ve found something new to focus on. Through my internship with Red Hat last year, I discovered that there is a large deficit of good statistics about our community. There’s a large deficit of good statistics about most free software communities, according to some random Google keywords I just tried, apart from “this is how many times our product has been downloaded.” I really loved the opportunity to combine my self-proclaimed mad Python skillz with answering other people’s questions, such as:

• How many contributors does Fedora really have? And according to these standards/filters?
• How often is the wiki edited and when?
• How many “things” has this random dude over here done? Do we consider that “active”?
• How many vague statistically-related questions can we come up with on devel@l.fp.o or during a marketing meeting?

Some of these, obviously, have no answer. Yet.

When I finally graduate from high school, I’ll be pushing full swing into answering these sorts of things. Until then, you can help me make Fedora a better place by simply telling us what you want to see tallied up. I asked this about 9 months ago and I got a lot of responses — thank you. But with recent discussions about the future of Fedora and a lot of claims about our user and contributor bases not being backed up (not pointing fingers), I think there are even more questions that can be answered. Please add your statistically-inclined questions to [[Statistics 2.0]] and I’ll do my best in the near future to get them answered with statistics on our community.

I also love help. (Shout out to joshkayse who is taking the lead on making it simple to find a single contributor’s actions within Fedora, taking inspiration from Mel’s FAS scraper.)

Quick summary: Maintaining packages is a drag (for me) right now. I like taking questions and answering with numbers. I graduate soon. Ask questions.

1 While writing this I decided to Google for “fedora wiki czar“. What I found was a mysterious character who was appointed as such in a community touting full transparency. Mel brought this to my attention the other day — I really suck at providing transparency into the process of administering the wiki. It’s pretty much on a whim. It shouldn’t be this way.

Formatted notes available at http://blogs.sun.com/glassfishpodcast

• Supporting slides (Virtual Conference)
• Java Persistence 2.0 (JSR 317)
• Java EE 6 Tutorial
• GlassFish v3

Music: Rodolphe Raffalli - Je me voyais deja.

From about 2001 until about 2009 I used evolution as my mail client. I got addicted to how it handled vfolders and got comfy with it. Then vfolders self-destructed and started telling me I had MAXINT new messages while not actually having ANY mail and well, it stopped being useful for me. So I switched back to my old stand by of alpine. It works fine more or less there is only one feature I miss from evolution – the ability to have multiple windows open when I’m composing a mail. So I can get info out of one mail and paste it in to use in the one I’m composing.

I setup claws-mail this weekend to see how that would work for me. It looked very hopeful until I went to setup multiple imap servers and it refused to see the other imap server at all. It’s like it didn’t even exist. So I shelved it back away and went looking for something else. Nothing has sprung up so far. We’ll see.

Suggestions are welcome, but here’s what I’ve tried:

thunderbird -

kmail -

alpine

mutt

claws

evolution