Someone just sent me a link explaining that over in Russia there is a competition going whereby my Home Office is being compared with that of Microsoft’s Steve Ballmer. Unfortunately I can’t read a single word of Russian so I’ve got no clue what the battle of the offices is about but I thought it was rather hilarious! See for yourself at; http://www.fishki.net/comment.php?id=17527
There are, of course, other people who fell for the Russian joke and actually do think that this is Mr. Ballmer’s office. Unfortunately (for Steve!), he probably wouldn’t feel very much at home in my office… there are computers here that run Linux, you see. That might be too much to handle and cope with. 🙂
Another cool site, chipchick.com, a site about technology and gadgets from a girl’s perspective, also points out the difference in office styles. Well, girls, the main difference is that I prefer screens and Steve prefers Windows. That and I have to use my office to work for a living, he doesn’t have to. 🙂
From my translations it reads “We had to save the best for last in the article…nothing in the world can compare to the ZX81 Super Computer”
🙂
Hey there– I’m a reader from ChipChick and I must say you set up is pretty intense. Perhaps you may consider a bit of color instead all gray and black. Imagine the exponential level of productivity you will have =)
That should probably make the quotes file. That’s BRILLIANT. (I say “brilliant” in the most Guinness-esque manner possible.)
Sorry to ask the mundane, but is there any chance of those wallpapers being made available? 😛
Specifically the lake and woodland ones.
Sure. They’re by 9X. Get them here. 🙂
Hey Stefan,
Very nice setup.
I can read Russian and it basically says that Ballmers office sucks compared to yours,
but it also says that you are a regular web designer, I’m sure you do more than that.
ViTaL
Thanks Vitaliy! I was wondering what all those characters meant in Russian. 🙂
Though regular web designer… uhm… I don’t even do web design at all (never did), except for my own site, that is. 🙂
Very nice setup YEAH
it would be nice to spend all that money in having f**ing 6 screens to work at while there are people starving to death. people that DIES because they have no FOOD. or screens.
you americans stinks in any ways possible.
Luciano, I think you are the first racist here and failed to notice that I am NOT an American. It’s a good thing I use all those screen and hardware to keeps people employed and working and earning money and thereby taking part in their economies. People, like myself, who get the living daylights taxed on our earnings and income. Our tax of which parts are spent to provide aid and food for poor people. Go complain to your politicians and globalists and bankers if food or money doesn’t arrive where it should. Thanks.
Stefan.
That is one sick setup! And I mean “sick” in the good sense of the word. Very cool, indeed. I’m actually in the process of implementing something similar in my home office, but I must admit that I doing more because I can and less because I need to. But I’m not going into to debt to get it done. 😉
Anyway, very cool.
luciano, It is seven screens, not six. you suck.
Stefan, You, on the other hand,are my hero.
Stefan, you mention a compile-farm made of VMs hosted on the same system.
Don’t you think that the virtualization in this case simply slow things down and a single, multithreaded compile-system will perform better?
Note that this is a serious question, thank you
Multithreaded compiling on a single system might be faster when it comes down to the actual time spent compiling but I’m doing more than just compiling on the “compile farm” (perhaps build farm would be a better word) and it doesn’t run on just a single server with several VM’s.
The entire build farm runs on two server (with some redundancy) and consists of a total of 8 VM’s that are usually booted and ready to go. There are 7 additional VM’s that do some very specific building/compiling where, unfortunately because of legacy support, the libraries/SDK’s can’t be used in a stand-alone fashion (think COM libs that are scattered throughout the OS thanks to old installers of 3rd parties). Those 7 specific VM’s are booted up and shut down automatically by the build processes/scripts/tools on an as-needed basis.
If I could run the whole build system on a single system with x CPU’s and y cores it might be a little faster but the virtualization choice has little to do with the potential slight performance gains but rather with the overall management and maintenance of the system as a whole. It’s much easier to just clone part of the system (a single VM) and perform maintenance and upgrades to it without potentially having to take the entire build farm offline. It’s also more forgiving when mistakes are made because restoring/copying back an entire previous configuration in the form of a VM is less invasive and disruptive than having to restore an entire system that does everything in a self-contained fashion.
The other big reason for the virtualization of the build farm is portability. I can easily take the whole batch of VM’s and relocate them to any system. All it takes is a few 100GB worth of transfers over the network. Very useful when hardware maintenance on the servers is required or when replacing an entire server which I seem to do at least once every 12-24 months (and servers can and will break down or fail so the portability is even more important because backups alone only go that far). The portability is going to be a major time saver for me since I’ll be moving to the other side of the planet soon and will be without most of my setup. I’ll have to buy a lot of new systems and the time to build up an entire single system build farm would take days. Now I just take 2 of those WD Passport 1TB USB drives with me (one in carry-on and one in checked-in luggage, both identical copies) that contain the whole build system of VM’s and all I need to do is copy them over from the external little disks.
Several of the VM’s, usually split over two physical systems, with 2-4 cores allocated to each, will do the actual compilation. Another two VM’s manage the source control (including daily, weekly, and in active projects even hourly, backups and synchronizations). Two more VM’s hold the project management portals (bug tracking, documentation, etc.) and another VM runs the ‘master’ build system that is in control of the entire farm. The last of the 8 main VM’s in the farm process results, files, installers, and gathers up the binaries and does all the boring stuff like archiving, putting it in the right places for daily/weekly (alpha, beta) builds and also takes care of producing final deliverables for releases.
Now, for the really fun part, while all this usually runs on just two servers, there have been certain projects in the past where a few more VM’s were involved to do the building processes and those VM’s would be started on any other system that had the resources and availability (a little Windows Service I wrote way back allows the master builder to check which systems are capable and available and would instruct it to start/shutdown any VM it required at the time). That way the entire build system was split over 3 to 5 systems which I guess you could see as a “multiple multi-threaded compile farm”. 🙂
thanks for your explanation, in fact your system seems to gain a whole bunch of flexibility in exchange for a mere bit of efficiency (however i can’t imagine not going crazy with so many VMs 🙂 )
Wow this is an epic setup what OS you running there and are all the monitors connected to one hostcomputer