This asus has a Broadcom BCM4313 wifi card. The linux kernel that ships with Slackware 13.37 comes with an open-source driver for this wifi card. Unfortunately, it does not come with the firmware for the card, so the driver is useless. If you try to download and install the driver from Broadcom it crashes the machine (even if you blacklist every other module).
The solution is to download and install the firmware according to the driver's README (which you can find at /usr/src/linux-2.6.37.6/drivers/staging/brcm80211/README). Once the drivers are copied to /lib/firmware/brcm/ you need to make symlinks "bcm43xx-0.fw" and "bcm43xx_hdr-0.fw" to the files which look closest like those.
Of course the git repository that has the files is not working, so you have to pull them from somewhere else. You can get the archive from debian here: http://packages.debian.org/sid/firmware-brcm80211 (download the source package's .tar.gz file and extract it)
Once you set up the firmware, just reboot and the machine should attempt to load the brcm80211 module and the right firmware automatically. Don't use the "wl" driver from broadcom as it will crash the machine. Add "wl", "b43" and "ssb" to the /etc/modprobe.d/blacklist.conf file just in case it tries to load those.
UPDATE
Apparently, that stuff doesn't work the way it should. You have to upgrade to the latest 2.6 kernel (2.6.39.4 as of this writing) and load the 'brcmsmac' driver, as this is the new driver used by the BCM4313 on the latest 2.6 kernels. Blacklist all the other drivers first ('wl', 'brcm80211', 'b43', 'ssb', 'b43-legacy', 'bcma'). I'm not sure if this is because the latest firmware is incompatible with older driver versions, but it kept crashing my machine to do anything but use the latest kernel and the brcmsmac driver with the latest firmware. What a pain in the ass.
Tuesday, October 25, 2011
Monday, October 24, 2011
pNFS is in Linux 3.x!
I totally missed it, but pNFS is officially in Linux 3.0 and beyond. If you need simple, stable, parallel network filesystem that is included with vanilla Linux kernels, now you have it. Any NFS 4.1 compatible client should be able to use servers set up with pNFS.
Here's the docs I found so far on it:
http://wiki.linux-nfs.org/wiki/index.php/PNFS_Setup_Instructions
http://wiki.linux-nfs.org/wiki/index.php/Configuring_pNFS/spnfsd
http://wiki.linux-nfs.org/wiki/index.php/PNFS_Block_Server_Setup_Instructions
http://wiki.linux-nfs.org/wiki/index.php/Fedora_pNFS_Client_Setup
Here's the docs I found so far on it:
http://wiki.linux-nfs.org/wiki/index.php/PNFS_Setup_Instructions
http://wiki.linux-nfs.org/wiki/index.php/Configuring_pNFS/spnfsd
http://wiki.linux-nfs.org/wiki/index.php/PNFS_Block_Server_Setup_Instructions
http://wiki.linux-nfs.org/wiki/index.php/Fedora_pNFS_Client_Setup
Thursday, October 20, 2011
most new startup companies are stupid
Let me go down the list of some kinds of startup companies from Start-Up 100 and why they're stupid.
Advertising and Marketing
Right off the bat, a bunch of useless bullshit. I don't ever WANT to see an advertisement or marketing. I want to be able to find that shit if i'm looking for it, but I don't want any of it to just show up somewhere.
Audio and Media
More stupid "web 2.0" websites based around music and other crap. Internet radio has existed for well over a decade. I don't need another place to not find the music I want to hear. (Pandora sucks, Spotify sucks, Grooveshark sucks... it all sucks. I'll turn on Shoutcast or Last.FM Radio if I want random music that I kind of like)
Education, Recruitment and Jobs
First of all, if you didn't get a normal education, some web 2.0 shell of a company probably isn't going to educate you any better. We have Dice and Monster, and people who are competent at what they do will network and find jobs in person like normal.
Enterprise: Security, Storage, Collaboration, Databases
Finally, startups intended for technology. Too bad all of them suck. Most people i've seen who try to develop startups have not worked very long in tech so they design or implement poorly and if they survive it's from sheer luck. Most of these solutions are crap or unnecessary.
Finance, Payments and Ecommerce
Again, i'm pretty sure all the big contenders have already been created. It'd be interesting if they actually had a new way to deal with finance or ecommerce, but most of it's been done and there's not a lot of room for innovation.
Gaming, Virtual Worlds
Ok, here's something that actually has promise. Make a stupid game which is addicting and make a billion dollars like Rovio.
Social Networking and Collaboration
JUST. LET. IT. DIE.
Social networking is a fad. You know what the original social network was? AOL. Just let the shit die. God I hate social networks.
Travel and Transport
I guess there's still a few niche/boutique businesses you could start in this space. But if it's another "how to look up cheap flights" website, just kill yourself.
What I would like to see more of are startups that are intended on bettering mankind, or fixing a common problem, or pioneering a new technology (a *real* new technology, not just a new shitty website or NoSQL garbage tool nobody wants). Medical device startups are really cool. Startups that develop technology for the 3rd world are cool. I'm still waiting on somebody to build a company that just services new companies, giving them turn-key solutions to build new networks and support them. I'll go work for them.
Advertising and Marketing
Right off the bat, a bunch of useless bullshit. I don't ever WANT to see an advertisement or marketing. I want to be able to find that shit if i'm looking for it, but I don't want any of it to just show up somewhere.
Audio and Media
More stupid "web 2.0" websites based around music and other crap. Internet radio has existed for well over a decade. I don't need another place to not find the music I want to hear. (Pandora sucks, Spotify sucks, Grooveshark sucks... it all sucks. I'll turn on Shoutcast or Last.FM Radio if I want random music that I kind of like)
Education, Recruitment and Jobs
First of all, if you didn't get a normal education, some web 2.0 shell of a company probably isn't going to educate you any better. We have Dice and Monster, and people who are competent at what they do will network and find jobs in person like normal.
Enterprise: Security, Storage, Collaboration, Databases
Finally, startups intended for technology. Too bad all of them suck. Most people i've seen who try to develop startups have not worked very long in tech so they design or implement poorly and if they survive it's from sheer luck. Most of these solutions are crap or unnecessary.
Finance, Payments and Ecommerce
Again, i'm pretty sure all the big contenders have already been created. It'd be interesting if they actually had a new way to deal with finance or ecommerce, but most of it's been done and there's not a lot of room for innovation.
Gaming, Virtual Worlds
Ok, here's something that actually has promise. Make a stupid game which is addicting and make a billion dollars like Rovio.
Social Networking and Collaboration
JUST. LET. IT. DIE.
Social networking is a fad. You know what the original social network was? AOL. Just let the shit die. God I hate social networks.
Travel and Transport
I guess there's still a few niche/boutique businesses you could start in this space. But if it's another "how to look up cheap flights" website, just kill yourself.
What I would like to see more of are startups that are intended on bettering mankind, or fixing a common problem, or pioneering a new technology (a *real* new technology, not just a new shitty website or NoSQL garbage tool nobody wants). Medical device startups are really cool. Startups that develop technology for the 3rd world are cool. I'm still waiting on somebody to build a company that just services new companies, giving them turn-key solutions to build new networks and support them. I'll go work for them.
Saturday, October 8, 2011
note to self for change management system
if a hack like extra privs is applied to a system to allow a dev to fix some issue in production or something, should be a system in place to automatically revoke privs after a given time. or specify a date/time range that the privs should be added, so you can specify "during maintenance window sunday 5am-9am developer Steve gets weblogic sudo access". as a matter of principle, all changes should be allowed to have date ranges applied to control when the changes happen. if the date/time starts but has no end, assume end time is indefinite. if start time and end time are the same, change is only applied once.
all account access should have defined end dates (for example, contractor steve has a 6 month contract, so all his access should have an expiry time set). BEFORE access is revoked email alerts will be generated for 2 weeks out, 1 week out, 3 days out, 1 day out before access expires to alert somebody before his access goes out the window. most configuration should not have expiry times because it's assumed if it is in config management it's meant to be there indefinitely, but for quick hacks where we know we don't want it to be there long we can set expiry times and will get alerts before it expires.
all account access should have defined end dates (for example, contractor steve has a 6 month contract, so all his access should have an expiry time set). BEFORE access is revoked email alerts will be generated for 2 weeks out, 1 week out, 3 days out, 1 day out before access expires to alert somebody before his access goes out the window. most configuration should not have expiry times because it's assumed if it is in config management it's meant to be there indefinitely, but for quick hacks where we know we don't want it to be there long we can set expiry times and will get alerts before it expires.
Subscribe to:
Posts (Atom)