It's not so much the sound that I would worry about as the shock. A sonic boom is a shock wave. This turbine is in a tiny enclosure that would have to absorb that shock.
I just think it's an interesting engineering problem to deal with.
1 million RPM at the diameter of a dime. An American dime has a diameter of 17.91 mm. At 1 million revolutions per minute, a point on the edge of the turbine blade will travel:
10**6 * 2 * pi * (0.01791 / 2) meters per minute
or 56,265.9 meters per minute, or 937.8 meters per second.
No, this was an Office 2000 install. I know; I installed it.;-)
Excel has problems with "large" workbooks for some value of large. I've seen this through several versions, going back to at least Excel 95. Once a workbook gets above about 500Kbytes/several thousand rows with multiple worksheets, it gets unhappy. Not very confidence inspiring that they haven't been able to nail this problems down in 10 years.
I have a client who uses Excel extensively. They've built a spreadsheet that they've been steadily adding to over the past year. Yesterday, Excel just rolled over and died on them. This was a 6,000+ row spreadsheet with formulas, various flavors of highlighting, etc. that contained a year's worth of data. I don't know how they managed to save it, but if you tried to open it with Excel you'd get the friendly(?) "Microsoft Excel has encountered a problem - do you want to send a bug report to Microsoft?"
They were desparate: they (of course) had no backup except for the original source data, meaning it would take them days to re-assemble the spreadsheet. They asked me to "fix it." I had had problems like this in the past, and usually saving the file as a.csv then back again as a.xls would fix it, but this time I couldn't even open the file. I figured it was toast.
Then I tried OO.o. I opened it with "Spreadsheet" (offtopic aside - part of me wishes the OO.o guys had more clever names for their components, and part of me is glad they don't waste their mental energy on such trivialities:-). It opened just fine. I saved it as an Excel 95 format document, then tried opening it from Excel. It opened just fine.
I'll never get my client to move to OO.o (they are a 10+ year Excel user and are basically computer illiterate and petrified of ANY kind of change), but it's nice to have it as a tool that actually works for those times when Microsoft falls down on the job.
...is a good project management application. I just scanned SourceForge.net but didn't find one. IMHO this is sorely lacking in the Open Source world. So much so that I've thought about writing my own (I wrote one that was curses(3)-based back in the early '80s:-). Does anyone have any pointers to a decent[1] project management app? Or should I start coding?;-)
[1] decent == Can track resources, tasks, costs; can perform some sort of resource auto-leveling; can report resource conflicts; supports GANTT charts; has a relatively easy-to-use UI.
Every one of the 1,252 songs on my iPod is from my own CD collection or purchased from iTunes.
Fuck you, Ballmer.
(Note to moderators: you may choose to moderate this as redundant, but I'd like to see all iPod-owning Slashdotters post their own statistics just to prove the point. Oh, and once more, Fuck you, Ballmer.:-)
My ex-wife has built an ad hoc accounting firm in her home. They had a P III-450 running Windows 98 as one of the workstations, and were doing file sharing off of it. It was painful.
Finally, I convinced them to get a newer computer for that employee. They bought a refurbished Dell - a Celeron - but with "only" 128MB of RAM, and running XP Home. I added a 512MB DIMM, which vastly improved performance. One happy employee.
Then, I took the P III-450 and rebuilt it with Red Hat 8, upgraded to kernel 2.4.26. I put a recent build of SAMBA and CUPS on there, and now they have a rock-solid file and print server that performs much better than Personal File Sharing on Windows 98. One happy accounting firm.
That doesn't seem right. Even our Windows 2000-based database server, running SQL Server 2000, ran for 15 months before we finally retired it for new hardware.
Of course, the server had no access to the Internet, so we didn't have to worry about the patch-o-the-month. Otherwise we wouldn't have made it 2 months without a reboot.
I don't run into malware on my personal systems, or my systems at work. But I do computer maintenance on the side of a series of Curves for Women gyms, and their computers were getting hit constantly. I finally had to encourage the owners to adopt a zero-tolerance policy: downloading anything or surfing to any site other than a list of previously approved sites (the Curves corporate web site, Yahoo mail, and a few others) is a fire-able offense. The computers run a gym-management software package that is based on Microsoft Access and, therefore, requires Windows.
Since adopting the zero tolerance policy and installing McAfee Personal Firewall running at the "tight security" level, I haven't seen malware on any of the systems.
McAfee lets you delete the specific executable that comprises the malware application, but doesn't help you delete the associated data files. I find I have to go in afterward and clean up after McAfee does its thing. (I suppose I should get a copy of Ad-Aware or one of the other anti-adware apps, but, frankly, most of them LOOK like the ads they're trying to kill - these guys need to get better GUI designers and look more "professional".)
Still, it's better than nothing.
BTW, have I said lately that adware/spyware authors should DIE A HORRIBLE, PAINFUL, LINGERING DEATH!
And how do you document complex interactions in complex projects needing tens or even hundreds of programmers?
You can't. Any software project that requires tens or hundreds of programmers is doomed to failure.
All successful software projects have been developed by small teams, regardless of the size or complexity of the project. Large teams tend to run into the "Mythical Man Month" effect.
One example I can think of: Cisco's IOS. Three years ago, when I attended an executive briefing at Cisco in San Jose, I was told that Cisco had 3,500 engineers working on IOS. Yes, some of those are QA engineers, not writing code but instead developing test cases. Yes, IOS consists of many programs running on a real-time kernel, just as "Linux" consists of hundreds of user-space programs running on the kernel. However, most of those Linux programs are orthogonal in nature: if vi blows up, it's not going to take gcc, cupsd and Quake with it;-). In IOS, on the other hand, the programs tend to interact with each other, resulting in a MUCH higher level of complexity.
The briefing I attended was to address the issue of trying to get a "stable" IOS release out of Cisco. I was working for an ITSP at the time, and we had hundreds of Cisco 5300 H.323 gateways deployed around the world. We were desparately trying to deploy new functionality, but were unable to get a stable IOS release out of Cisco; every time they'd release a patch, it would break something else. After 18 months we still hadn't received a stable release, thus the briefing.
At the briefing, Cisco recognized that they had a problem. They showed us graphs of defect rates over time by release, talked about internal initiatives to address their perceived quality issues, and introduced us to their new head of software development, who promised to clean up the whole mess.
A year later, the situation had not improved, and shortly thereafter they had yet another new head of software development.
From my perspective (software engineer with 20+ years experience), the problem is simple: they have too many engineers working on too old a code base. The only solution for them is to start a "skunkworks" project with a small team to re-architect, re-design and re-implement the entire thing from scratch. Until Cisco takes this approach, IOS will continue to suffer from stability issues and ridiculously long lead times for new features.
I don't mean to pick on Cisco specifically; I'm sure that every large software development organization has the same issues, and for all I know, in the two years that have passed since the last time I dealt with Cisco's development organization, things may have improved (but I doubt it). Cisco just happens to be one that I'm familiar with.
Actually, if you read Steve McConnell's Code Complete, a lot of it is about the Macintosh port of Excel, and the coding techniques used on that project. It made for good reading; every programmer should read that book at least one during their career, preferably toward the beginning of their career...
I maintain computers for a set of Curves for Women gyms owned by a couple of friends of mine. I run into the spyware/malware problem all the time.
Each gym uses DSL to connect to the internet. While working on one of the computers this weekend, I noticed that McAfee Personal Firewall (I stopped using Norton a while ago) wasn't seeing any inbound events, unlike the other gyms where it sees 10,000 to 20,000 events per week. A little investigation showed that the DSL modem at this site has a built-in DHCP server/router/firewall/NAT function. Seems like the DSL providers are getting a clue and building necessary capabilities into the hardware that the customer has to have just to connect to the Internet.
Slashdot Mirror
Though it was mentioned in the article lead-in, I prefer the form
e**(pi)i = -1
to the form
e**(pi)i + 1 = 0.
It's not so much the sound that I would worry about as the shock. A sonic boom is a shock wave. This turbine is in a tiny enclosure that would have to absorb that shock.
I just think it's an interesting engineering problem to deal with.
1 million RPM at the diameter of a dime.
An American dime has a diameter of 17.91 mm.
At 1 million revolutions per minute, a point on the edge of the turbine blade will travel:
or
56,265.9 meters per minute, or 937.8 meters per second.
The speed of sound at sea level is 340.29 meters per second.
So this thing's blades will have a tangential velocity of mach 2.76.
I think the sonic boom when it starts up will be as much if not more of an issue as the whine from its operation...
Could the solution be as simple as putting a fine-meshed screen over the intake?
You mean like this?
The GP has obviously never heard of multiple births, either...
;-)
Technically, it is feasible (sextuplets every year for 17 years would do the job). Just not very likely.
Perhaps, but in my Start menu on Windows XP it shows:
Start->All Programs->OpenOffice.org 1.1.2->Spreadsheet
I guess the menu item doesn't reflect the name of the app...
No, this was an Office 2000 install. I know; I installed it. ;-)
Excel has problems with "large" workbooks for some value of large. I've seen this through several versions, going back to at least Excel 95. Once a workbook gets above about 500Kbytes/several thousand rows with multiple worksheets, it gets unhappy. Not very confidence inspiring that they haven't been able to nail this problems down in 10 years.
I have a client who uses Excel extensively. They've built a spreadsheet that they've been steadily adding to over the past year. Yesterday, Excel just rolled over and died on them. This was a 6,000+ row spreadsheet with formulas, various flavors of highlighting, etc. that contained a year's worth of data. I don't know how they managed to save it, but if you tried to open it with Excel you'd get the friendly(?) "Microsoft Excel has encountered a problem - do you want to send a bug report to Microsoft?"
.csv then back again as a .xls would fix it, but this time I couldn't even open the file. I figured it was toast.
:-). It opened just fine. I saved it as an Excel 95 format document, then tried opening it from Excel. It opened just fine.
They were desparate: they (of course) had no backup except for the original source data, meaning it would take them days to re-assemble the spreadsheet. They asked me to "fix it." I had had problems like this in the past, and usually saving the file as a
Then I tried OO.o. I opened it with "Spreadsheet" (offtopic aside - part of me wishes the OO.o guys had more clever names for their components, and part of me is glad they don't waste their mental energy on such trivialities
I'll never get my client to move to OO.o (they are a 10+ year Excel user and are basically computer illiterate and petrified of ANY kind of change), but it's nice to have it as a tool that actually works for those times when Microsoft falls down on the job.
...is a good project management application. I just scanned SourceForge.net but didn't find one. IMHO this is sorely lacking in the Open Source world. So much so that I've thought about writing my own (I wrote one that was curses(3)-based back in the early '80s :-). Does anyone have any pointers to a decent[1] project management app? Or should I start coding? ;-)
[1] decent == Can track resources, tasks, costs; can perform some sort of resource auto-leveling; can report resource conflicts; supports GANTT charts; has a relatively easy-to-use UI.
Every one of the 1,252 songs on my iPod is from my own CD collection or purchased from iTunes.
:-)
Fuck you, Ballmer.
(Note to moderators: you may choose to moderate this as redundant, but I'd like to see all iPod-owning Slashdotters post their own statistics just to prove the point. Oh, and once more, Fuck you, Ballmer.
Riverside, Iowa is the alleged birthplace of James T. Kirk, not William Shatner (though sometimes I think Shatner himself gets the two confused ;-) ).
My ex-wife has built an ad hoc accounting firm in her home. They had a P III-450 running Windows 98 as one of the workstations, and were doing file sharing off of it. It was painful.
Finally, I convinced them to get a newer computer for that employee. They bought a refurbished Dell - a Celeron - but with "only" 128MB of RAM, and running XP Home. I added a 512MB DIMM, which vastly improved performance. One happy employee.
Then, I took the P III-450 and rebuilt it with Red Hat 8, upgraded to kernel 2.4.26. I put a recent build of SAMBA and CUPS on there, and now they have a rock-solid file and print server that performs much better than Personal File Sharing on Windows 98. One happy accounting firm.
That doesn't seem right. Even our Windows 2000-based database server, running SQL Server 2000, ran for 15 months before we finally retired it for new hardware.
Of course, the server had no access to the Internet, so we didn't have to worry about the patch-o-the-month. Otherwise we wouldn't have made it 2 months without a reboot.
Shouldn't that be:
Dupe, Dupe, Dupe, Dupe of URL....
I suppose I should have qualified my situation.
I don't run into malware on my personal systems, or my systems at work. But I do computer maintenance on the side of a series of Curves for Women gyms, and their computers were getting hit constantly. I finally had to encourage the owners to adopt a zero-tolerance policy: downloading anything or surfing to any site other than a list of previously approved sites (the Curves corporate web site, Yahoo mail, and a few others) is a fire-able offense. The computers run a gym-management software package that is based on Microsoft Access and, therefore, requires Windows.
Since adopting the zero tolerance policy and installing McAfee Personal Firewall running at the "tight security" level, I haven't seen malware on any of the systems.
McAfee lets you delete the specific executable that comprises the malware application, but doesn't help you delete the associated data files. I find I have to go in afterward and clean up after McAfee does its thing. (I suppose I should get a copy of Ad-Aware or one of the other anti-adware apps, but, frankly, most of them LOOK like the ads they're trying to kill - these guys need to get better GUI designers and look more "professional".)
Still, it's better than nothing.
BTW, have I said lately that adware/spyware authors should DIE A HORRIBLE, PAINFUL, LINGERING DEATH!
It was (1983-1999). The result was The Phantom Menace. 'nuff said.
You can't. Any software project that requires tens or hundreds of programmers is doomed to failure.
All successful software projects have been developed by small teams, regardless of the size or complexity of the project. Large teams tend to run into the "Mythical Man Month" effect.
One example I can think of: Cisco's IOS. Three years ago, when I attended an executive briefing at Cisco in San Jose, I was told that Cisco had 3,500 engineers working on IOS. Yes, some of those are QA engineers, not writing code but instead developing test cases. Yes, IOS consists of many programs running on a real-time kernel, just as "Linux" consists of hundreds of user-space programs running on the kernel. However, most of those Linux programs are orthogonal in nature: if vi blows up, it's not going to take gcc, cupsd and Quake with it
The briefing I attended was to address the issue of trying to get a "stable" IOS release out of Cisco. I was working for an ITSP at the time, and we had hundreds of Cisco 5300 H.323 gateways deployed around the world. We were desparately trying to deploy new functionality, but were unable to get a stable IOS release out of Cisco; every time they'd release a patch, it would break something else. After 18 months we still hadn't received a stable release, thus the briefing.
At the briefing, Cisco recognized that they had a problem. They showed us graphs of defect rates over time by release, talked about internal initiatives to address their perceived quality issues, and introduced us to their new head of software development, who promised to clean up the whole mess.
A year later, the situation had not improved, and shortly thereafter they had yet another new head of software development.
From my perspective (software engineer with 20+ years experience), the problem is simple: they have too many engineers working on too old a code base. The only solution for them is to start a "skunkworks" project with a small team to re-architect, re-design and re-implement the entire thing from scratch. Until Cisco takes this approach, IOS will continue to suffer from stability issues and ridiculously long lead times for new features.
I don't mean to pick on Cisco specifically; I'm sure that every large software development organization has the same issues, and for all I know, in the two years that have passed since the last time I dealt with Cisco's development organization, things may have improved (but I doubt it). Cisco just happens to be one that I'm familiar with.
Actually, if you read Steve McConnell's Code Complete, a lot of it is about the Macintosh port of Excel, and the coding techniques used on that project. It made for good reading; every programmer should read that book at least one during their career, preferably toward the beginning of their career...
I thought it was more like the Federation broadcasts in Starship Troopers. "Would you like to know more?"
Those were the funniest (and only redeeming) scenes in that otherwise god-awful movie...
I maintain computers for a set of Curves for Women gyms owned by a couple of friends of mine. I run into the spyware/malware problem all the time.
Each gym uses DSL to connect to the internet. While working on one of the computers this weekend, I noticed that McAfee Personal Firewall (I stopped using Norton a while ago) wasn't seeing any inbound events, unlike the other gyms where it sees 10,000 to 20,000 events per week. A little investigation showed that the DSL modem at this site has a built-in DHCP server/router/firewall/NAT function. Seems like the DSL providers are getting a clue and building necessary capabilities into the hardware that the customer has to have just to connect to the Internet.
In my opinion, you need a firewall for anything that can connect to the Internet. Period.
I'm all for lawyers in space. Of course, lawyers in space with space suits, well, I'd have to think about that some more. ;-)