betabug... Sascha Welter

home english | home deutsch | Site Map | Sascha | Kontakt | Pro | Weblog | Wiki

Entries : Category [ digital ]
Things having to do with digital world items go in here.
[digital]  [language]  [life]  [security]  [media]  [zope]  [tourism]  [limnos]  [mac]  [athens]  [travel]  [montage]  [food]  [fire]  [zwiki]  [schnipsel]  [music]  [culture]  [shellfun]  [photography]  [hiking]  [pyramid]  [politics]  [bicycle]  [naxos]  [swim] 

07 March 2012

Service Announcements: Currently No Commenting

Gotta fix things

A short service announcement: Due to a persistent SPAMmer who managed to get by my first line of defenses, commenting is for the moment turned off. Sorry for the inconvenience. This spammer doesn't manage to actually post comments, but I get a lot of notifications (and moderated comments are in the system). I'll fix things soon.


Posted by betabug at 11:58 | Comments (0) | Trackbacks (0)
20 March 2012

An Evening at the Hackerspace

Learning lots of stuff

I spent this evening at the hackerspace.gr, in the company of my friends graffic and tralala and a few other good people there. The time spent there was not only fun, but very productive for me:

In difference to the last 2 times I was there, it was a quiet evening, but totally worth it, for fun and profit!


Posted by betabug at 23:41 | Comments (0) | Trackbacks (0)
25 May 2012

Back to Bookeen

Got my ebook readers back

The screen on my first ebook reader (Bookeen Cybook Opus) broke a long time ago. Some months ago I also broke the screen on the ebook reader I bought as a replacement (Bookeen Cybook Odyssey). So I calculated about sending them in for a repair. Now this kind of thing isn't covered by the warranty, even though some people complain that these screens break way too easily (not just Bookeen's, but all current ebook readers). The repair isn't cheap, getting them both repaired costed almost as much as buying one new one. In the end I went with the repair because a.) this way I have two readers for the money and b.) I don't like the "throwaway culture" that is all too abundant these days.

I got my readers back this Wednesday. There were some delays in sending out the devices (on my side). Also the confirmation mail that told me where to pay for the repair got stuck in my spam folder (something that hasn't happened for more than a year). I paid for the repair in the online shop on Monday and on Tuesday FedEx was knocking on the door for the first time... impressive.

My first impression was that these readers weren't repaired but replaced. So much for avoiding the "throwaway culture", but fine with me. Especially on the Odyssey, there are some changes to the hardware: There is a different on/off switch and USB connector.

So I'm back to reading on the ebook reader. Of all things, I picked up Moby Dick again. Then I switched to some short stories I had found online, then to some technical manuals. Business as usual on the ebook reader. I'm a bit over cautious with the screen though, but I guess I'll relax soon enough.


Posted by betabug at 10:00 | Comments (0) | Trackbacks (0)

Convert CHM to epub with Open Source Software

Command line information

Speaking of ebooks and technical manuals in my last post: I had some stuff that was in "CHM" format. Now in theory calibre can convert from CHM to epub, but it never worked for me. All I ever got was the contents and maybe the first page. Given a bit of command line use, there was an easy solution for me.

I used chmlib (in my case the OpenBSD chmlib-0.40p0 package) and htmldoc (the htmldoc-1.8.27p6 package). First with extract_chmLib the chm file was converted to a directory full of html files. In the next step I ran htmldoc with the --continuous option to make one big html file out of all the chapters and parts. In this case the parts were named so that they went in proper order, I don't know if that is usually so. I only had to do some minimal cleanup in the html.

This big html file would already have worked in the ebook reader, but I went one step further and converted it with calibre to epub format.


Posted by betabug at 10:20 | Comments (0) | Trackbacks (0)
27 May 2012

One Year of hackerspace.gr Party

Good times in these times

Saturday evening I was at the "one year of hackerspace.gr" party... at the hackerspace. Besides the good people I know through the hackerspace itself, I met there also Panos and Javier. I haven't been at the hackerspace from day one (as I was busy with other stuff in those times, see previous post), but I still felt entitled to show up for the party!

The evening started with the last two talks of the hackfest 11 that was right before the party. Then it just went over to party mode, which means that small groups of people were having interesting discussions on all kind of technical or non-technical stuff.

One thing that comes to mind right now was the story of the guy who had rigged a chainsaw motor to his bicycle. No, I didn't saw the guy, neither the bike, but someone there described the encounter. The "cutting parts" were removed from the chainsaw.

Then there was the (very tasty) birthday cake, with one electronic candle on top... which was blown out by Pierros.


Posted by betabug at 23:03 | Comments (0) | Trackbacks (0)
31 May 2012

SSH Mastery: OpenSSH, PuTTY, Tunnels and Keys

Book review: An introduction to using SSH

Some weeks ago, undeadly.org published an announcement for the book "SSH Mastery" by Michael W. Lucas. The book isn't expensive and I was in the mood, so I bought it at smashwords. (At smashwords I was able to get it both in PDF for reading on the 'puter and in epub for the ebook reader. No DRM either, very reasonable.)

Now, this book came probably a few years too late for me. If you aren't using the ins and outs of ssh yet, if you feel confused by all this stuff ssh does, this is the book for you. It covers OpenSSH and Putty. There's a lot of practical info in there. But then the author clearly says what the book is not:

This book is not intended to be a comprehensive SSH tome.

Too bad, since that would be really nice to have! So for myself I haven't found much new stuff in there. Well, there were some details that I tended to forget and got a good reminder. Also I had never really played much with X Forwarding. I'd like to play around more with X Forwarding just for the cool factor, but right now I don't have any setup where I could use it.

The description of how to do an ssh based VPN gave me much more confidence to try it myself one day... if opening an ssh login for root wouldn't freak me out (even with all the restrictions that you are dutifully guided to set up).

I found one big mistake in the book, it says:

Also note that all IP address bindings must be chosen before opening
your SSH session. You cannot add port forwarding to a live SSH
session, or change the IP addresses bound during a session.

Indeed we can: We use ssh escapes to do exactly this, at least for sessions with a pty. Type newline + ~C and you'll get a prompt where you can add and remove port forwardings.

In general, escape sentences are missing from the book. This is a pity, since even a beginner might find ~. useful to kick out a stuck ssh session. Apart from these few points I can really recommend the book!


Posted by betabug at 23:25 | Comments (0) | Trackbacks (0)
09 June 2012

A Simulated Network with Qemu and Sockets

Set it up fast and simple

I was trying to test some networking stuff with various operating systems. I have neither the patience nor the hardware to set up a bunch of real machines with those systems to try out things. So qemu came to my mind to simulate these things - which would be good enough for my problem case.

According to my study of the various parts of documentation on qemu, the fastest way to get this going was to set up a "socket" based network. That way, a couple of virtual machines can share a virtual network. They don't have access to the host's network connection - and therefore to the Internet though - unless you set up one of the guest OS systems as a router and set up some routing. I wasn't going to do that, as in my problem case a bit of isolation is a good thing.


My first machine is running some oldish variant of Linux. There isn't much special about it in that sense. This machine is the one that is set up as a "socket listener" in the command line:

#!/bin/sh

qemu-system-x86_64 -nographic -no-acpi \
-m 1024 -clock unix \
-net nic,model=rtl8139,macaddr=52:55:01:4e:79:28 \
-net socket,listen=:1234 \
-hda lunix_hda.img

Basically it's a normal qemu command, the only thing different from usual is that we're not specifying the default "user" type of network, we're using "socket". This means that the vm will listen on port 1234 on the host operating system (127.0.0.1 is implied here).

The second machine is running good ol' Windows 2000 - guess why a bit of isolation is a good thing. Here too, the setup is pretty much standard. Note that I'm specifying MAC addresses - apparently if I don't do that, all virtual systems will get the same MAC address, which is not a good thing in a network. The socket network setup is there too:

#!/bin/sh

qemu-system-i386 -m 512 \
-net nic,vlan=0,model=ne2k_pci,macaddr=52:54:00:12:34:58 \
-net socket,connect=127.0.0.1:1234 \
-monitor stdio w2kdisk.qcow2

I'm not sure it's necessary to specify 127.0.0.1 here, but it was like that in the docs. We do not set it to "listen", we set it to "connect" on this machine.

Now these two virtual machines will live in the same network, it's as if they were connected with ethernet cables. They do not have any IP addresses yet though - when using the "socket" networking in qemu, the built-in DHCP server from the "user" network is not activated. No problem though, on Windows just go to the control panel and set up an IP address and netmask (no router, name server or anything else required). Set up a different IP in the same network range and netmask on the other machine (on Linux with ifconfig). Now they can ping each other and as a test the windows machine can telnet to port 22 of the Linux system.

More machines could be set up to this system and you could even set up more virtual ethernet cards and hook them to more virtual networks, or route them through the host system to the Internet.

Posted by betabug at 22:10 | Comments (0) | Trackbacks (0)
19 June 2012

Airplanes now with "integrated literature consoles"

ebook on plane
Bookeen ebook reader on plane

On my flight to Switzerland I was reading with the ebook reader. After a while I got bored holding it in my right hand, then my left hand, then back to the right hand... The table was too low to place it comfortably, so I had to find another solution.

Good for me that airlines these days have "integrated literature consoles" in the backside of your front seat. These integrated consoles are of the "bring your own ebook reader" variety, but apart from that they work fine. How do they work?

The result is a very relaxed and healthy reading posture, due to the ergonomics of having the reader nearly at eye level and having your hands free.


Posted by betabug at 16:27 | Comments (2) | Trackbacks (0)
02 July 2012

Make Your JavaScript Behave

Installing jsLint into vim

Like most programmers, I'm using JavaScript, because that's the way it is these days. Given the minefield of that language, I really want a tool like jsLint to watch my code behave. So what I did was to integrate jsLint into vim. Which wasn't so difficult at all.

First step Is to get a JavaScript engine. For JavaScript unit tests I already use rhino, but it's a bit slow, especially to start up. The jslint.vim plugin recommends SpiderMonkey, so I gave it a try. There is an OpenBSD port for it, so that was no sweat. It took me a bit of looking up to discover that this will install a JavaScript prompt in /usr/local/bin/js -- neat. The jslint.vim plugin knew that already, so no configuration needed there.

Second step was to install the jslint.vim plugin itself. Here the main necessary hint was to add the command filetype plugin on to my ~/.vimrc file. In ~/.vimrc I also added these lines to be able to switch the highlighting of "complaints" off and on:

" jslint display errors off & on again
map %0 :let g:JSLintHighlightErrorLine = 0<CR>
map %1 :let g:JSLintHighlightErrorLine = 1<CR>

Instead of %1 and %0 you might want to choose some command that suits you better, I'm always at pains finding good command keys in vim that are not used for something else yet.

Now jslint was complaining about too much. The next step is to create a configuration file in ~/.jslintrc to give it some standard settings. Currently for me that's:

/*jslint browser: true, nomen: true, sloppy: true, vars: true, maxerr: 250 */
/*global jQuery, $ */

Any of these settings and any others can be configured on a per-file basis, by adding a comment like that to a JavaScript file.


Posted by betabug at 08:28 | Comments (0) | Trackbacks (0)
11 July 2012

darcs https/SSL repos on OpenBSD

Complaining, complaining, always complaining

Trying to access a repo on a https URL (using SSL) using darcs on OpenBSD, I got the error:

"Peer certificate cannot be authenticated with known CA certificates"

Even though there is a "real", "verified" certificate on the SSL web host. The problem appears to be that darcs goes through curl for these http accesses and curl does not find the proper CA root certificates.

Solution: Download the root certificates from the curl caextract page and install them in /etc/ssl/cert.pem. Seems like there was some root CA missing in the cert.pem bundle that I had there.


Posted by betabug at 10:28 | Comments (0) | Trackbacks (0)
13 July 2012

tmux book at the beach

Useful Information and a horse

We went for a late afternoon swim today after work (ok, before work too, I'll soon continue a bit). While there I was reading the tmux book I had bought a weeks or so ago.

I hadn't expected too much in that book, since I'm already using tmux, have mastered the first bit of confusion, and after all, all the info is in the man page. But then, on the first 15 or so pages, I already found a bunch of interesting information and things to try out too.

For example it made me think about organizing different projects in separate sessions and more consistently name some windows.

While I was reading these things in the last half an hour before sunset, the beach was pretty much emptied. Then there arrives a guy on a horse, riding without saddle. They go straight for the water, right in the sea. At this beach, the beach falls down pretty fast, a few steps and you are in swimming depth. The horse goes in probably till it lost step, swam a bit and they came back and leave. Impressive.


Posted by betabug at 20:16 | Comments (0) | Trackbacks (0)
06 October 2012

Queimada-Sprint

Hacking till the drinks are on fire

So, here I am, on the other corner of Europe. Together with graffic we flew via Madrid to La Coruna and then drove with Wu to his place in Lugo. Yesterday we started "sprinting" on our customer's pyramid app. Sprinting, what's that? It means that for some days we work really concentrated on programming this thing. It also means that for once we're working in the same room instead of cooperating over the net.

It's really great to work with these guys. Yesterday r0sk was working with us too. We got tons of things done already, despite "wasting" some hours at the start having a meeting about what to work on first. Now the sprint will continue till Tuesday and then we'll have our queimada and we'll be relaxing in the countryside of Galicia for a while.


Posted by betabug at 11:40 | Comments (1) | Trackbacks (0)
04 December 2012

vi or emacs?

It might actually make sense to ask this sometimes

A few days ago we were talking with someone who we thought he would join our team (in fact he did a few days later). When all seemed to be asked and answered, half out of wittiness, I asked the "big" question: "vi or emacs?"

Turns out, that question is actually quite valid in a "hiring" situation for a programmer job. It's not that I actually care what people are using (besides, vi is so clearly better, there's no contest). But the resulting discussion gave me some insight into the "geek level" and the approach to many work process related questions of the person sitting in front of me. All in a non-threatening, easygoing topic, since come on, nobody really discusses this on a serious level.

It boils down to: The choice of tool might not define good workmanship, but the approach of how someone chose their tools sure gives some insight. You can feel how sure they are of their work, what level of expertise they have, how they approach learning new tools, and so forth. What if someone would really get on a rant when posed such a question? Well, that gives you an answer too, doesn't it?


Posted by betabug at 18:39 | Comments (0) | Trackbacks (0)
22 December 2012

Public - ποιον κοροϊδεύετε με αυτήν την τιμή;

Ερευνα αγοράς

Χτες βράδυ λόγω εποχής και κίνησης είχα κολλήσει στο Mall στον Αγ. Δημήτριο. Πάω μια βόλτα από το Public εκεί να χαζεύω gadgets. Και τι βλέπω; Είχαν το Bookeen Opus (ebook reader) για "μόνο" 228 Ευρώ. Το "μόνο" είναι σε εισαγωγικά, γιατί online το αγοράζεις στο bookeen.com με 99 Ευρώ. Ποιον κοροϊδεύετε με αυτή την τιμή ρε Public; Περιμένετε να περάσει κάποιος στο χριστουγεννιάτικο στρες που δεν έχει την παραμικρή ιδέα και δεν ξέρει να ψάξει 5 λεπτά online;


Posted by betabug at 14:19 | Comments (0) | Trackbacks (0)
23 December 2012

Playing around with multipath routing on OpenBSD

Two options, two paths

Yesterday the router that was fried in a lightning strike has been replaced. After making the necessary changes to the repeater, I thought it would be a good moment to play around with network stuff, since I have now two network options. There is the reinstated ADSL connection and the mifi that I used as my Plan B.

So I went to the Networking chapter of the OpenBSD FAQ, where I had seen the sample setup for equal-cost multipath routing. I think the instructions are quite clear, but it sure is something that I wanted to try to see how it "feels" and works. Good for me that I have a splendid Laptop with OpenBSD installed, which has an Ethernet port as well as a working wifi card (ok, more or less working, but as long as the router is in the same room, it does work). So my setup is:

  1. ADSL-Connection, passing through an Access Point with OpenWRT that is configured as a repeater. The OpenWRT router has one Ethernet port and here I used that to connect to the em0 interface on the laptop.
  2. Mobile connection (3G / HSDPA, whatever the reception) in the shape of a little "mifi" access point. This little beast is actually faster than the ADSL connection, but then, it's a tiny ADSL connection. I use the iwn0 interface on the laptop to connect to this one.

On the em0 Ethernet interface I set up a static IP in the range 192.168.2.0/24, on the iwn0 wifi interface I have dhcp configured, which gives me an address in 192.168.5.0/24. The last time I tried the mifi with a static address, it didn't like it, something was blocking there, maybe that thing is configured to give access only to DHCP clients in order to be able to limit access to 5 clients? Further investigation will be needed there.

Next thing was to set the sysctl parameter for multipath routing and then to set up the routes, like in the FAQ. Easy enough, I could ping both gateways... but then, I couldn't actually get further out from one of the gateways. Looking at the routes, they had different priorities: em0 had 8, iwn0 12. Maybe this was an effect of having one of the routes created by dhclient. In any case I flushed the routes and created them new, setting -priority 8 on both of them. That did the trick.

Now looking at netstat -r they both had the same priority and both started to have increasing numbers in the "Use" column. The "P" flag for multipath was also present in the routes. Then I opened to terminals with ntop and had fun for a while watching various connections pop up on each interface. The effect could also be felt: I uploaded something big, which likely saturated one of the (tiny) uplinks, but still I was typing in ssh without any delay. Sure I have a nice pf setup with queues and ACK priorization, but with such an upload still there is a little delay noticeable. With the dual uplinks, it seems that the connections balanced out better. Definitely though, the max speed of a connection is defined by the speed of each one of the interfaces, not by the sum of both of them.

I haven't yet understood every detail of how this works, e.g. what algorithm is used to balance connections or which interface is chosen. I guess that once opened, a connection stays on the same interface, as some protocols won't take it well if your source IP jumps around (also see some explanations here). Definitely it also does not do "fail over" out of the box, but there are various solutions for that.

Conclusion: Using this on my laptop is not going to be something that I will do every day. But using an OpenBSD router for example in an office setup, where multiple people access the internet, it could be a nice option. Combine two or more cheaper Internet connections and have people not hinder each other, no matter if someone downloads or uploads some bigger files for a while. Then add some failover capabilities and connectivity through different ISPs and you will gain a little bit of uptime too.


Posted by betabug at 18:16 | Comments (0) | Trackbacks (0)
29 December 2012

An Experiment with PC-BSD on the Thinkpad X220

Not going to buy it anyway, you know

Background: There are some things that you know you're not going to switch to, but still you want to try. For myself, I know that I like the command line. I switched to OpenBSD from the Mac (ok, I still do some things on the Mac, e.g. full scale image editing), after deciding that I spend all my time on the command line anyway. I now use cwm and tmux. Still I wanted to give PC-BSD a try. For one thing, my laptop's hardware is supported very well in OpenBSD, but it's not supported perfectly. Maybe FreeBSD could improve on that, and the easiest way to find out (without having to digg into FreeBSD's config world) would be PC-BSD. And then, I just like to tinker with this stuff...


So I went and downloaded a PC-BSD image (9.1). Or rather: I tried, because the torrent would just sit there and look at me with lonely eyes. After a bit of searching it turns out that the project's torrent servers are out of order. OK, this kind of stuff happens. So I grabbed it off an ftp mirror, dd'ed it to a USB stick and there we go. I had prepared an external disk, with one big OpenBSD partition (need some more backup space) and a smaller, but still comfy FreeBSD partition.

Then I came to where the funky graphical installer lets you select the disk. That was scary. I have two internal disks in this laptop, both of which hold OS data. Sure I have backups, but if an install would wipe one of those disks, it still would mean a hassle and a lot of lost time. There are a few indications that help choosing the disks: Luckily the two internal disks were labelled with the manufacturer info... or at least with some parts of the manufacturer info. Instead of "Hitachi" and "Kingston", there was "itachi" and "ingston". I wonder how they missed such an "in your face" bug, but at least it let me exclude those disks.

Next options were to choose between the USB stick with the installer and the USB disk, which was the desired target for the install. Here I was a bit confused: The disk sizes used (in Megabytes) didn't look really familiar. The FreeBSD partition I had prepared was nowhere to be seen. At one point I thought it's the USB stick, but that is smaller. The installer showed the full disk size (in the "use whole disk" option) and it also listed the OpenBSD partition. In the end, since I had no data on the OpenBSD partition, I chose to use the whole USB disk for now (which means PC-BSD will get deleted faster, since I want that disk space). Did I mention it was scary? In fact I stopped the installer and rebooted into OpenBSD to check the disk sizes.

Once the install started, the blinkenlights and the buzz of the drive reassured me that indeed the right drive was being used. I was thinking a lot if the curses based disk format/select procedure in the OpenBSD installer is any better. I guess it can be pretty scary too, but a.) it forces you to read up on what you do and b.) it does not claim to be newbie-friendly. When I installed OpenBSD on the mSata SSD while having data on the spinning disk, I was a bit worried to chose the wrong drive, but being able to see the manufacturer info helped there. Having had some years of experience with the OpenBSD installer makes me a biased reporter in this respect though.

The actual install went through reasonable fast (USB 2 isn't so fast). Then came up two more problems, which I sometimes couldn't tell apart: For one thing, PC-BSD does not seem to know how to power down this laptop. After telling it to shut down, it just sits there with a black screen, I suppose it's halted. For a second thing: On OpenBSD there is a problem to switch off X and then you can't get back to a console, you just hang there with a black screen. Looks like FreeBSD has the same problem here. So it took several tries to get past the stage where the installer showed me which video driver it had detected (Intel) and if I want to go with it. After it got stuck on that for the first time, I was almost going to abandon the try.

Then I played around with it a bit more, asking for a different video driver (vesa gave a funny color screen and got stuck, wtf) and another with the Intel driver selected manually and finallyy got it one more stage further, where I was asked to select the time zone and set up a user. At the end of that, again a black screen. Another round of force rebooting and I got to a working system. I guess it had taken me 6-8 reboots or so, sometimes it looked like the video went bust, sometimes it looked like the installer tried to restart. The good thing is that the system didn't run any fsck, so restarting wasn't any slower than normal. I don't know, but maybe that is a result of ZFS.

So now I had a running system. Clicking around, it looked kind of ok. The LXDE desktop I had chosen looked and worked a bit like Windows, "Start-Menu" and all. Sound didn't work until I selected another sound device (from a graphical menu), but that wasn't hard to find. Installing Firefox with the graphical app installer was childs play, these are the kind of things that less technical users of the system will adore. Using Firefox I went to Youtube to try out some videos. That worked really well, both in the embedded video and in full screen the video didn't stutter and played flawlessly. I would say the new Intel video driver in FreeBSD 9.1 is really doing a good job there. This really is an improvement to what I get with the same machine on OpenBSD.

Also I guess that Firefox was using the Flash player that was installed with the system- but I haven't really checked if Youtube used that. For a non-technical user what counts is that Youtube just works. After a while I wanted to turn down the sound a bit. The hardware sound buttons didn't work. The taskbar sound slider didn't work. What worked was the sound slider in Youtube. Ugh. (In OpenBSD the X220's hardware sound buttons just work, but without an on-screen-display feedback.)

The OpenBSD iwn wifi driver for the "Intel Centrino Advanced-N 6205" wifi card in this laptop is a bit bad: Basically it works, but I need a really strong signal to connect. We're talking about "access point has to be in the same room" kind of strong signal. A little bit further off and there will be some hassles with disconnects, a bit more and no connection at all. I did not see the FreeBSD version of iwn to improve on that a lot (it didn't connect to the upstairs wifi, where the white macbook with Mac OS X connects perfectly and OpenBSD's iwn on the X220 fails totally), but I didn't experiment more to determine if there is a small improvement.

Frozen screen on trying to suspend the X220 with PC-BSD 9.1

Next hardware check: suspending the laptop. Closing the lid din't have any effect at all. Maybe that has to be set up somewhere (in OpenBSD you have to set it up). So I went to the menu and told the machine to suspend from there. I got a funny screen (see pic) and the machine froze instead of suspending. Same thing for "hibernate" from that menu. With OpenBSD the X220 suspends/resumes works just fine (with one small caveat: after resuming you can't switch to a console from X any more).

Time to round up this short experiment. I powered down the system... oh, in fact, I told it to shut down, then waited a bit to assume that it was done and held the power button till the laptop force shut down.

I have a little theory of mine: Every OS is just a different set of compromises. If a system works for you is determined by which compromises you are willing (or even liking) to take or not. That might be a question of how much Open Source you like your system to be, how much tinker friendly, how much it has to "just work" or "just run the stuff I need". Especially with a laptop, the grade of hardware support is another big source of compromising. Buy a Mac with OSuX and the hardware should interact perfectly with the Software from the same brand (but pay for it and take things like nonreplaceable batteries and ever increasing OS bloat, see what kind of compromises I'm talking about?)

First of all, as expected, PC-BSD is a nice system for someone wanting to work in a real GUI environment, moving around with a mouse and clicking on graphical menues. Not my choice right now, but I knew that much from the start.

As for the hardware: On the plus side, there is a much improved video driver. I'd like to have that, but I can live with the non-accelerated driver in OpenBSD. On the downside, there is no suspend (big outch), the question of why it doesn't power down (only a nuisance), and the non-working sound buttons. I didn't notice a big change with the iwn wifi driver. Looks like in that respect I'm better off with OpenBSD, since my big nuisance there (iwn) didn't improve, while I'd compromise a not-so-important improvement (the video driver) against a big downside (not suspending).

Note to (especially future) readers: If you like PC-BSD, give it a try yourself. By the time you found this blog post (which might be a lot later), there is a good chance that they improved on the hardware support, also you might not have the same exact laptop as me.

Posted by betabug at 21:26 | Comments (1) | Trackbacks (0)
Prev  ...  4   5   6   7   8   9   10   11   [12]   13   14   Next