This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.
OK, this is a bit of a diversion. My last post was about my perceived shortcomings of Windows 8, or specifically, the Metro interface.
Since then, work on my new system came to an abrupt halt. It seems that around the time I was typing my previous post, I was also having a heart-attack. This entailed a trip to ED, where the problem was identified, and entailed my having to have an emergency procedure to fit a stent to fix a blockage in one of the coronary arteries that was stopping 90% of the blood getting through to my heart. As Kurt Vonnegut would say, so it goes. In hindsight, this had been building up for some time, I can see, and while leaving me feeling better than I had felt for some time, it left me weak. I was ordered to rest up for a fortnight, and now I am now gradually getting back on my feet.
Two things have happened since my last post. One is the receipt of my Raspberry Pi.
Now, I know it’s for kids, but how could I resist a a linux-compatible ARM-based computer the size of a credit card for less than $50/£30? It runs Debian better than my Nokia n900 did, my only other gadget capable of running Debian on ARM.
Yah, it’s just another linux desktop, XFCE this time. But…
I am rarely so impressed with anything as I am the RPi. It has to be the easiest computer I have ever set up.
I would recommend any parent who wants to encourage their kid to do more with a computer than do social media stiff to get them one of these, over and above any other computer.
The steps were:
1. Get the RPi ($49)
2. Get a box to put it in ($11)
3. Get a power supply that fits a micro-USB socket
(I re-used my Nokia n900 charger, but I also have a TomTom one that will support with a USB->microUSB cable)
4. Download Raspbian (Debian-ARM for the RPi) & uncompress it.
5. Flash a 4GB SD card with the Boot & OS image.
6. Next fit it all together:
I had a spare HDMI-DVI cable to plug into a Samsung monitor
I had a Ethernet cable to plug into a hub
I had a Dell USB keyboard with built-in USB hub, to run keyboard & mouse through one of the two USB ports.
Plugged the SD card in, connected the power, and up came Debian ARM, to the RPi configuration screen.
NB. OK, I admit, one problem: booting using the USB keyboard. If I booted with it plugged in, I’d end up at a blank screen as soon as the rasp-config menu came up.
I had to boot without the keyboard, and then plug it in once at the rasp-config menu. Then, I could go through the menu, and input the localisation & keyboard mapping. On reboot, the system came up fine with the keyboard plugged in.
7. Work through all the relevant items on the raspberry configuration, including updates, and you will eventually reboot into LXDE.
8. I’m not keen on LXDE. Didn’t like it on the n900, don’t like it on the RPi. So, I ran a debian update & upgrade, installed XFCE4, synaptic update manager, X11vnc, chromium, etc., and eventually ended up rebooting into XFCE, and able to log in using a Remmina from the laptop.
So, this is what the final XFCE desktop looks like via VNC:
Eventually, I’ll set the RPi to run in headless mode, in a way that I can run the desktop remotely. What for? Well, already I’m seeing this could act as a file server, print server, although limited to external USB drives, possibly a NAS, or a music player. But, the graphics capabilities are very good (1680*1050 – full HD), so running headless seems a wee waste of the GPU. Thin client, maybe? Or as a tiny TV-server. The possibilities are limitless, and people are already working on setting these up as XBMC boxes. It will run Apache easily, and people have set them up as small LAMP servers. Very impressive little computer. Given this is a single-core 700MHz ARM cpu, it highlights the potential of these processors in something like the ARM 1GHz dual-core in the Xoom, or the quad-core 1.5GHz Samsung S3. Unfortunately, unlike the n900, to be able to run linux on these, you have to root the phone/tablet, which risks screwing up the 3/4g connection (along with warranty & RTM potential for fixes). It is a shame, because despite the potential power of these new generation of ARM devices, Android prevents using their full potential in much the same way Apple do with iOS, and same with Microsoft Windows 8.
Which neatly brings me back to my last piece about Windows 8, where I have had a slight change of heart. In fact, having read further into other people’s experience, I have begun to re-think how I feel about Unity & Gnome 3.x.
I do still think it is the ugliest interface I have yet come across, and I suspect that at some level Microsoft realise this too – hence the 60′s style psychedelic embroidery that is available with Metro. But, I now appreciate that the Windows 7 desktop is hidden under there, accessible by pressing the Windows button and “I”. The Metro interface, while acting as the input screen for tablets, is also what you used to find under the traditional Windows ‘Start’ button, and using Windows+I/C/etc you can get to change settings, access control panel functions, or go to a desktop with its own panel, where you can pin the apps you want there. It’s a bit like the way linux/unix has different screens. So, yes, quite a clever solution to running the same OS on a tablet/phone and desktop; well thought out.
This helped me see what Gnome is all about. And Gnome is ten times nicer than Windows 8, aesthetically. If only they had come up with that kind of solution, it could have saved them a lot of grief. Keep the old desktop ‘behind’ the new environment, rather than forcing people to have to use the fallback, thereby locking them out of new features they might actually appreciate. And, here’s the rub, because it takes me back to Android, and the inability to run a proper Linux desktop from Android without having to root the phone/tablet. It should be quite easily implemented so that those who want to could run a traditional desktop like XFCE from an Android device, if required, just as it should be possible to run a Gnome 2.x like desktop from within then Gnome 3.x shell (rather than using that as a fallback). What is a bit frustrating is that the potential to do this has been there in linux all along, but the opportunity was missed – and Windows have done precisely this. Shame.
And, ironically, the most ingenious use I have seen for the Raspberry is to do much the same thing. One of the drawbacks of the Raspberry is the lack of small cheap monitors. people have made use of small 3.5″ car-reversing monitors, but even a 7″ monitor costs over $100, and beyond that at 19″ you are looking at something in the region of $150, that isn’t portable. So, somebody has come up with a way of using the USB ports to connect to a tablet, and using the USB to network and VNC client-server, run Raspbian on the Pi, using the tablet as both monitor and keyboard, and touch-screen interface. Way to go, but with one snag… to achieve this, you need sudo/su access to the tablet, which means rooting the tablet, which not everybody wants to do.
So, I finish as I began, with a raspberry.
I had hoped to be writing about my progress on the new 16GB i7 3.5GHz quad-core PC I built, except the company I bought it from insisted on assembling it for me, so I let them. Only to find they had stuck a sticker on the back that said the warranty was void if I opened it.
Anyway, the short story is I eventually got everything up and running, using Debian Wheezy with kernel 3.6.5 I had to download, build and install, because the mobo was so new the stock 3.2 kernel had a few problems with it. That and having to build and install custom networking firmware & driver, because Gigabyte have changed the on-board network card from one supported in the kernel (I checked beforehand…) to one not supported yet (grrrr). Virtualbox installed, along with extensions, guest additions worked on an old copy of XP I had lying around, and freeBDS as well as Debian-GNU/HURD, both of which require the hardware virtualisation I built the PC to meet, both work, as well as Debian-kfreeBSD (flaky), but pcBSD still has problems installing to a virtual machine, which I’ve not figured why yet. Lots more to go, but as a diversion, I decided to buy a copy of Windows 8, stick it on a virtual machine, and see how that goes.
My first impressions of Windows 8:
The problem with it is that although I assume with all their millions, Microsoft did do some good human interface design work, but my guess is it was all based on how people would use a tablet; so, when it comes to working on a desktop, most people will be used to a different way of working. In order to use Windows 8 on the desktop, people will have to change the way they have worked for the past 25 years. This happened with the shift from applications written for terminals, early PC’s and CPM & MS.DOS (such as IBM Displaywrite, Supercalc, Harvard Graphics, WordStar, WordPerfect, etc.) to the MS & Apple windowing systems we all know and love-hate. Back then, there was a bit of resistance, but, these new metaphors were fairly intuitive and easy to use, and had the advantage that where before you really had to use key-stroke commands, you still could use key-stroke commands, but didn’t have to. There is a reason both Apple and Microsoft did so well using the windows metaphor for so long: it worked, people found it easy use, and it reflected how they worked in the world. These are key requirements of human interface design. There is a saying “if it ain’t broke, don’t fix it”. So, what they have done is taken something that has worked well for many years, and replaced it with something that doesn’t (from what I can see). They have done this, not as an option, but as how you WILL work.
I am sure that this will work fine for people doing fairly ‘trivial’ tasks, like browsing and e-mail, or if they only tend to use one or two ‘serious’ applications. But anything more than that, I don’t see how more complex tasks would be achieved. I am sure it will be possible, but not sure how. For example not sure how you would have a browser, e-mail, word, and powerpoint open, and read a pdf while you edit a jpeg file and a document, draggin things between them. That was a way of working that reflected how people do things in the world. People do not tend to work by flipping between screens, carrying what they have copied with them. I may be wrong, maybe I am missing something, but without breaking the interface design in order to force it work the way I am used to, if I used Windows, I’d feel that Microsoft were forcing me to learn a new way of working.
To be honest, as much as I dislike Gnome 3.5 on linux, it does what Windows 8 is trying to do much better, and looks far more polished and professional, with much better compositing engines available. Gnome 3.5, like Unity, broke the traditional computing metaphor for some reason over a year ago (wanting to get onto tablets I suppose), but the only people who like it are the people who designed and developed it, who all tell the rest of us we are wrong for not liking it. I suspect the same is true for Windows 8. The designers and developers must like it, as do the management, and no doubt the fanboys will like it, and the sales people will have to, along with anybody would would be made to drink Kool-Aid if they dissent. I’m sure that guys for whom computing is all about downloading porn, people who venture little further than iTunes, Explorer and Mail, it will work. Maybe this heralds a shift in computing where any ‘serious’ computer users still using Windows will migrate to poper operating systems like OSX & Linux, and leave the ‘trivial’ stuff to Windows, Android & iOS. When I say trivial, I mean in terms of the types of applications you can use – I know of at least one person who runs a lot of their business via a web-browser on a tablet.
My thought about Gnome/Unity was that I couldn’t see an interface like that appealing to sysadmins, or even programmers; so, I am wondering where they will take this new metaphor when it comes to Windows Server. I cannot see this will be that acceptable for sysadmin tasks. OK, a lot will be done on the command-line, but one of the things Windows NT did introduce was the idea that you could do sysadmin using a windows manager. While you could do that with UNIX, it was never a pretty sight. So, I’ll be interested to see whether they release the next Windows Server version with Tiles, traditional Windows, or a choice.
Maybe I’m just getting old, I hear a whisper in my ear. I saw the shift from mainframes+terminals to UNIX-Servers+PC’s running MS.DOS to standalone PC’s running MS.DOS to standalone PC’s running Windows to UNIX/Windows-Servers+PC’s running Windows to Windows-Servers+Windows Thin Clients. And now this, and the smoke-and-mirrors magic of “The Clod” (no typo – lump of protocols and services that have been around in UNIX for donkey’s years, bundled together and re-branded as the “next-big-thing” ).
It seems that Apple are quite capable of supporting two forms of their OS, one on desktops/laptops, the other on tablets/phones; Linux is av aailable on the desktop, and with Gnome/Unity is aiming at tablets. Google uses linux to develop separate environments for laptops (ChromeOS) and for tablets (Android). I don’t really understand why Microsoft didn’t do Windows 8 as a fork of Windows for mobile devices, and retain the traditional desktop for those who want it; perhaps the memories of the disastrous Windows CE/PC2000/Mobile were too hard to bear – so rather than fail spectacularly at a mobile operating system again, they have teamed up with Nokia (who spectacularly failed to bring out a successful touch-screen device), and bet on the future of Windows on a successful mobile operating system, and not worrying about desktops until later. I guess we’ll have to see what the take-up is, and whether they have to backtrack and rapidly bring out Windows 9, like they did Windows 7, in order to finally get people to stop using Windows XP in 2014 (when extended support for it finally ceases).
In a year or two, will we all have taken our PC’s & laptops to charity shops and municipal tips, having traded them in for touch-screen TV’s, PC’s, notebooks & tablets? The typewriter metaphor in computing has been around for over 60 years, I learned to type on a manual, then electric, typewriter; the typewriter goes back 150 years. I remember doing data input by punching holes in cards – using a computer keyboard for data input was a huge improvement. Even on tablets, the keyboard metaphor is there for interacting in complex ways; and as an after-market accessory, a bluetooth keyboard is necessary for any substantial text-based activity. It would be naive to think that the system of working that has developed to one where we type on a keyboard, looking at what we type, copying things in from other places, all visible on our screen, will be superseded by tablets that quickly. Whereas as a mobile interface for warehouses and factories, on the road, the tablet will be revolutionary.
Now I have installed it, will I use it for my few remaining Windows applications? I don’t think so, after Windows XP, I prefer Windows 7.
This is not definitive, is incomplete, focuses on certain developments, omits Windows completely, and any suggestions would be welcome.
|1975||Version 6 UNIX||Bell Labs|
|1978||1BSD||Version 6 UNIX||PDP||Berkely UC|
|1979||Version 7 UNIX||Version 6 UNIX||Bell Labs (AT&T)|
|UNIX/32V||Vesrion 7 UNIX||Bell Labs (AT&T)|
|1982||System III||PWB/UNIX + UNIX/32 etc.||VAX + PDP||AT&T|
|1983||System V R1||System III + 4.1BSD||VAX + PDP||AT&T|
|GNU||OS development + Free sotware||GNU-O/S|
|1984||System V R2||SVR1||Xenix/386||AT&T|
|1986||System V R3||SVR2||AT&T|
|Mach||4.3BSD||VAX + SUN-68030 + IBM-PC/RT||CMU|
|Mach 2.5||Mach-2||CMU + OSF|
|XNU||Mach-2.5 + BSD 4.3||NeXTstep|
|L3||rewrite of Mach||x86||GMD+SET|
|SVR4||SVR4 + 4.3BSD + Xenix + SunOS||x86 + SPARC||AT&T|
|GridPad||MS-DOS (1st touch-screen tablet)||8086||(Jeff Hawkins for) Samsung|
|GNU-Hurd||GNU + Mach-3|
|Psion S.3||SIBO (EPOC+OPL)||NEC v30||Psion|
|1992||GNU/Linux||GNU + Linux|
|Zoomer||GEOS (Grid-pad based tablet)||Palm+Tandy+Casio (Hawkins)|
|Newton||NewtonOS (pen-input tablet)||ARM||Apple (with Palm apps)|
|1994||OSF/1||AIX + Mach-2.5 + 4.3BSD||Alpha||OSF|
|GNU-Hurd||GNU + Mach-4||GNU|
|OpenStep||Mach-3 + 4.3BSD||NeXTstep|
|Simon||Personal Communicator||1st Smartphone w/touchscreen||IBM/BellSouth|
|n9000 Communicator||GEOS v.3||i386 clamshell PDA/smartphone||Nokia+HP|
|Pilot||PalmOS (Zoomer development)||Motorola 683xx||(Jeff Hawkins for) Palm|
|MkLinux||Mach-3 + Linux||OSF + Apple|
|1997||Psion S.5||EPOC32||32-bit ARM||Psion|
|LighSurf||First MMS camera-phone||Philippe Khan – Lightsurf Tech|
|Mac Rhapsody||MkLinux + OpenStep|
|Mac OSX Server 1||XNU+Darwin+MkLinux+4.3BSD+Mach.3||PowerPC|
|2000||Darwin||OSX Server 1 + Rhapsody||PowerPC||Apple|
|Ericsson R380||Symbian||EPOC-based smartphone|
|Nokia 9210||Symbian||EPOC-based smartphone|
|2001||OSX.10.0||Darwin+OSX Server 1||PowerPC||Apple|
|Kyocera 6035||PalmOS PDA + cellphone||1st smartphone marketed in USA||Kyocera+Palm|
|2003||Android Inc formed||Linux based mobile OS development||Android|
|Nokia 9500||Symbian||smartphone with camera and WiFi||Nokia|
|2005||Google buys Android|
|OS2005/Maemo.1||Linux+GTK+Nokia proprietary||n770 tablet||Nokia|
|2006||OSX.10.4||PowerPC & Intel||Apple|
|2007||OS2007/Maemo.3||Linux+GTK+Nokia proprietary||n800 tablet||Nokia|
|iPhone OSX||OSX+services+media+Cocoa Touch||iPhone smartphone||Apple|
|Android source||Linux+Dalvik+Apache Harmony (Java)||Arm & Intel|
|pureDarwin||Apple-free OpenSource Darwin|
|2008||Android.1||Linux+Dalvik+Apache Harmony (Java)||HTC Dream smartphone||HTC+Google|
|Moblin||Linux based OS||Intel||Intel|
|Diablo/Maemo.4||Linux+GTK+Nokia proprietary||n810 tablet||Nokia|
|2009||Maemo.5||Linux+QT+Nokia proprietary||Nokia n900 smartphone||Nokia|
|ChromeOS||Linux based OS||Intel & Arm|
|ChromiumOS||OpenSource version of ChromeOS||Intel & Arm|
|Moblin.2||Linux based OS||Intel||Intel|
|WebOS||Linux based OS||HP/Palm|
|2010||MeeGo||Moblin+Maemo Linux merger||Intel+Arm||Intel+Nokia+LSF|
|Harmattan/MeeGo||Maemo+MeeGo Nokia fork||Arm||Nokia|
|Android.4||Linux kernel 3.x||Arm+Intel|
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.
I first had problems with the Bcom 4311 on my c300 when I first installed linux on it a few years back. The only one I could get to work was Ubuntu, and that kind of set the scene for me with Ubuntu from 6.04 on. But, the NDIS wrapper was a PITA. I retired the c300 when I started running Ubuntu on a new-ish laptop, without issues, after W7 died on an update and I discovered one of my recovery disks was corrupted. so, I used the c300 to try out Mint in the pursuit of a way of avoiding upgrades to Gnome/Unity – the Bcom 4311 was never an issue with Mint, as I recall, being based on Ubuntu.
I found the repositories on Mint never seemed to work smoothly for me here in this part of the world. So, having realized I would never get freeBSD running in a VM on an older PC, needing at least an Sandybridge i5, which would set me back about a grand, last week I wiped Mint and the old backup Ubuntu partition, and wiped it clean to try out freeBSD. I soon realised that BSD is still where linux was a decade ago when it comes to stuff like Bcom.
I was also having problems trying to get a Hauppage dvb-t working on my Dual Core 64-bit Ubuntu 12.04 laptop. I decided to see what would happen in the 32-bit version, so replaced freeBSD on the c300 with 32-bit Ubuntu 12.04. With a bit of tweaking getting VLC working with the Hauppage dongle was pretty painless compared to the 64-bit HDMI laptop. Lesson learned, when people say not everything works on 64-bit distro’s, they aren’t lying.
But, the Wireless was not working, despite the system identifying the need for proprietary drivers and (supposedly) installing them. I gritted my teeth to work out why the Wireless was not even switching-on. I was expecting a gruelling journey, as clearly what used to work in 10.10 wasn’t happening in 12.04; I really didn’t want to have to go back to a NDIS wrapper, especially when it used to work without that much pain, and with the stuff I’ve been reading about Bcom making the technical info available drivers to be built under Linux.
So, prepared to be pulling my hair out, I stumbled across this method from the first post-Unity-exclusive release, 11.10:
Even though I’ve never bypassed apt-update for this kind of thing before, it was a real treat. It worked just as well on 12.04. Thanks.
It leaves me with a question – why?
Why when this was working before, does it no longer work, and why do we have to remove the ‘wrong’ driver, and install the ‘correct’ one?
Sorry, but that kind of functionality/usability is far more important than sexy-interfaces like Gnome-3 or Unity, and if distro-providers lose sight of that, then they are stuffed.
Take getting DVB to work – nobody is going to take up Linux if they get a hint of what a hassle getting it working properly is going to be. Sure, MS Media Center is horrible, and a bit unreliable, and the system freezes now and then, but at least it just works (more or less) out of the box. It wouldn’t be too bad if there was a straightforward step-by-step explanation, but usually, those either simply work, or lead you into a brick wall that you then have to keep hitting your head about. There are so many different solutions for so many different releases in so many different distributions, it is actually quite frightening. I would have thought if anything was a priority for linux development, it would be getting DVB working consistently and properly.
I know it is not the distributor or developer’s fault, they can only address issues that relate to hardware they have access – but when hardware is stated to work on linux, both on linux sites devoted to video linux and manufacturers – and sometimes it can… but often it may not… the whole documentation process has to be called into question. I would question who has actually got some of this stuff working – because I see so many unresolved threads asking the question, so many snitty referrals to other threads, that often even don’t solve the problem, that I don’t thnk anybody has some of these working – otherwise they would simply post something like the guy I linked to above, which is basically, do this, this, and this, and it should work. That is why it is such a relief to come across such an elegant solution, because I could have wasted hours trying things out suggested by people who don’t really know.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.
These are a sample of things I have legitimately collected from beaches on both sides of the world.
I don’t believe taking a small piece of dead coral from a beach has any detrimental effect on the environment. I say this, because I made a comment somewhere about how a tiny piece of sun-bleached coral had been confiscated at customs when I declared it. The argument was that tourists doing on islands like those in Vanuatu will effectively lead to whole beaches being depleted of sand, the cycle of sand-formation being damaged by such thoughtless people. Is there any evidence for this idea that a few tourists can disrupt the cycle of sand-creation (which takes place over vast periods of time) I wonder? What is the research in peer-reviewed journals on this – how on earth would you even begin to prove such a hypothesis? We live in an age where what people believe is often stated as being in some way factual.
Nope, customs don’t allow these sorts of things in because they are worried about the beaches of Vanuatu. There is a list of things that are not allowed past customs, and coral is one of them. The customs guy agreed that as it was dead coral from the beach, it shouldn’t be an issue, but because I declared it, and coral is a protected species, even though it was not a living organism at the time I found it, it still could not come in. That is why it was taken away, and I was not fined. Had it been live coral, I would have been. Shells are not allowed because the country is extremely protective of its own flora and fauna, and shells have the potential to contaminate shellfish in national waters, which is a big industry here.
The biggest threat would not be tourists carting the odd shell or bit of coral away with them, it would be the death of the coral reefs themselves, which has little to do with tourists, but the fishing industry – particularly Japan’s, but also other trawling fleets operating out of Asia. Compared to the damage done by that, the little tourists might take is going to be negligible. This sounds more like anti-tourist snobbery to me, which is not unusual amongst ex-pats living in countries frequented by tourists. When I had to spend a lot of time in Spain, I avoided British holiday-makers like the plague, and here I tend to avoid any visiting Poms I might come across, if I can.
The ecology is damaged in the Pacific, just like everywhere else, despite what the glossy brochures might try and sell you, and regardless of any band-aids people might try and put on it. The water might look clear, and the fish amazing while snorkeling, but the radio-activity from experiments starting in the 1940′s, and the disaster in Japan last year all has a toll – it all had to go somewhere; then there’s trawling, off-shore drilling, etc. Anything individuals might do, like pick a piece of coral on the beach, is going to be pretty insignificant compared to that stuff. What we used to sometimes do in the UK, given that a FAR BIGGER PROBLEM IS PEOPLE’S RUBBISH is take plastic bags with us, and collect up the crap other people leave behind. That is a positive way of doing something constructive about a REAL PROBLEM.
If theories about global warming and rising sea-levels are to be believed, it is academic anyway, because in a few years, the beach won’t be there anyway, it will be underwater. The projected rate of increase in ownership of computers, internet access, and related technology over the next decade suggests that countries will need to double the number of power stations they have now in order to feed the demand for server-farms, ISP’s, as well as domestic computing. That power has to come from somewhere, and although countries like Norway and NZ with small populations and huge national resources can tap green energy, we still have coal-fired power stations, as it is not enough. An over-populated country like the UK, with fewer such resources, no chance – same for those countries that are developing technical economies. The only realistic way to provide enough energy is burning stuff, and/or atomic energy.
So, the only way to help the environment realistically is to stop talking about how to do it on-line, log-out, and keep the laptop/PC/smart-phone turned off. But, I’m not holding my breath waiting for the stampede of environmentalists going off-line.
I originally installed VirtualBox when somebody suggested this could be a useful way I could still run some old XP programs I needed for astronomy on a Linux PC.
So, I have XP available to me as a VirtualBox guest on my Ubuntu 11.04 (Gnome 2.32) laptop host.
As well as Windows 7…
(one way of recycling the original licences…)
I also run Ubuntu 12.04, with Gnome failsafe, on an old Dell P4 PC in the garage that dual-boots XP. Ubuntu for Midi, Yamaha keyboard and Music Studio, XP for ASCOM telescope control. Ubuntu also serves to play my MP3′s held on an atom file server running Debian in the office. Rather than Unity, I installed gnome-fallback, which gives me a desktop very similar to Gnome 2.3.
I do use Unity on a little netbook, but it can also run XFCE, which I prefer.
I use Debian 6 (Gnome 2.3) on a 2GB dual-core atom-based file & print server. This also runs a 500MB XP VB guest (primarily so that we can print to a Canon multi-function Laser Printer that has no Linux drivers). The Debian file server runs NFS, CUPS & Samba (CIFS), to allow access to files, the HP scanner/inkjet, and the Canon A3 photo-quality inkjet. The workaround for the Canon MF printer is to have Chromium loaded up with GooglePrint, which means I can print from Linux+Chromium -> Chromium+Windows.
Back to VirtualBox VM’s.
Apart from Windows XP & 7, I run a copy of Debian 6 (Gnome 2.3) – which could be lost now I use it as a server as well, I guess.
I have a copy of Fedora 17 running Gnome 3.4, just so I can see what I am missing.
I have Mint 13 (Mate), which is a fork of Gnome 2.3, basically. Mint also allows me to look at their spin on Gnome 3.4 failsafe, looking like Gnome 2.32, called Cinnamon
I have Suse 12, running KDE 4.7, so that I can watch those irritating bouncy transitions when I select something to run, and look at KDE if I want to.
A recent addition was CentOS 6.3 (Gnome 2.28), because while I have Fedora, I wanted to see what the more stable Red Hat Enterprise Linux was like, and in itself as a potential alternative to Debian.
I did install OpenSolaris, but as work on that effectively ceased in 2009 after Oracle took over, I have been looking at OpenIndiana (Gnome 2.3). This is a ‘proper’ Unix derivative. I think of all the UNIX variants I worked on over the year, I was fondest of Solaris, which had many features from SVR4. I was quite taken aback the first time I loaded openSolaris, to find myself looking at Gnome rather than CDE/Motif.
VirtualBox has also allowed me to try out things like Moblin, MeeGo, ChromeOS, Android, JoliCloudOS etc.
The other ‘proper’ UNIX systems I wanted to try were pureDarwin and FreeBSD, unfortunately I could not get either of them to run as a guest, because my laptop does not support VTx or EPT.
So, I wiped Mint Debian Edition off my old Compaq Celeron Laptop, and managed to get FreeBSD running on that instead:
Darwin will boot from CD on the same laptop, but there is no obvious way of installing it from there, and so no way of installing any graphical libraries that would let me run Darwin Xmas, for example.
Apart from Windows, Fedora and Suse, these are all running variations of Gnome 2.3. Ubuntu I found works quite well in fallback mode, and Cinnamon is Mint’s version of a post Gnome 3 implementation of Gnome. All the older, stable, releases still use Gnome 2.3, including the UNIX variants. The good news from last month is that the Common Desktop Interface has now been released as Open Source, and Motif may well be soon to follow. So, once some work has been done tidying those up, no doubt we will see revitalised versions of CDS/Motif on the open UNIX systems they were originally the main X interface for – and may even some of the LINUX variants as well.
So, whoever it was put me onto VirtualBox, if you happen to read this, I’m really grateful. I’m also grateful to whoever it was I tried installing on my relatively new laptop, rather than consigning it 5+ year-old kit; I think that is what converted me to linux once-and-for all.
None of these PC’s & laptops is younger than three years-old, some nearly ten years-old. The newest PC is the atom file/print server – which cost the equivalent of US$120-130 from an online-auction site. In fact, that’s the only bit of this kit purchased in the last three years. While I am keen on recycling older equipment, there is a significant energy cost in doing this, which was the reasoning behind the atom server – it consumes up to 1/10 of the energy of even a modern tower PC. That represents a cost of $20-30 a year, rather than the $200-300 older equipment can cost. So, my focus is starting to shift from recycling older kit, to investing in more energy efficient kit in the future.
As a postcript, I had several problems running the Vbox guest additions updates to 4:20 on several distributions from a range of versions like 4.08 & 4.12 after updating the main system – CentOS being the trickiest. I had to drop into root at the terminal, install gcc and dependent libraries, then run the ‘autorun.sh’ from inside the /media directory that contained the updates. Debian, I had to drop into root, but the shell command didn’t work; in the end, I had to launch nautilus as root, and run thr autoprompt in the gui as root. With Mint, I had to log out and back into Cinnamon, and then it ran fine from the Devices menu. Fedora had to be run from a terminal, have gcc installed, then run with a ‘sudo sh autorun.sh’ command in the relevant directory. Suse involved much the same, installing the base C/C++ & kernel development in Yast before it would run properly.