I've been involved in multiple situations in previous jobs where we contracted for companies who turned out to have extremely dubious codebases, generally provided by external companies, acquired with small acquisitions and so on.
The most notable one was Sonix (then bought by 3Com). Both Sonix and 3Com had definite policies and good practices but they had ended up with code, stripped of headers that came from Phil Karn's KA9Q.
I've also had a case where a companiy had reference code/windows drivers and wouldn't release it to help a Linux driver not because they didnt want Linux drivers but because their own engineers worked from an internal source pool that they thought was very unlikely not to contain large amounts of 'acquired' code.
On the whole it has been the smaller companies that cause the most problem - but not always. A look at Harald Welte's list of people who settled GPL violations contains some very large and reputable computer companies.
OpenDLM is the lock project. There are also people workin on an SSI project.
I don't think its fair ro say we are close to VMS, we have an awfully long way to go yet 8)
Re:... compared to InterMezzo, CODA or oMFS?
on
Red Hat announces GFS
·
· Score: 2, Informative
Intermezzo and CODA try and solve a different problem (the one AFS does), they replicate data as much as possible without violating coherency and at a file level.
GFS instead gives everyone access to the same disk at the same time rather than replication. Both methods work well for different data sets - so yes GFS and oMFS are similar
I think the other people have covered the basics pretty well - plug lots of computers into one fibrechannel or possibly firewire disk or disk array.
The second really interesting use is with virtualisation - imagine if you want all your S/390 virtual machines to share the same bsse file systems for efficiency (given the price IBM charge for mainframe disks;)) or the same with uml, Zen, etc
It shows how much the music industry owns the US government that the bill cites the 31% loss claims of the music industry in a law, and ignores the fact that all the academic studies, as well as the statistics gathering companies rebut the data rather convincingly.
Do you want the head of - The engineer who made the original mistake - The document writer who may have caused it - The manager who failed to do enough checking - The QA people for missing it
and so it goes on.
Kudos to Intel for simply saying "We screwed up" and recalling products. They seem to have learned much from the old Pentium FPU errata handling.
Presumably its outsourced to texas then. Nothing says Indiana can't outsource its government to Texas does it ? I mean we seem to have outsourced ours to Washington
Most of the HF bands are used for commercial purposes, some slots are allocated for limited private use (eg CB, remote control toys, 49Mhz walkie-talkies), and lots of it is used for emergency and longer ranger services where VHF/UHF simply won't do the job. This includes people like emergency services.
Amateur radio is probably more relevant now than since the 1940's. Its real reason for existance beyond the first uregulated days of "gee isnt this neat" was to provide a steady supply of wireless operators to draft in the event of a war. Its not the only reason but its a major reason it survived.
Shortwave radio communication over any long distance (commercial, military and hobbyist) often deals with weak signals. Each broadband power line adds to the background noise cumulatively raising the problem. One power line won't trash your TV signal (unless you are very close), but each one adds noise until all you have is snow.
Its like people talking in the background - a couple of people don't do much harm but when you try and talk across a room full of quietly talking people two things happen
1. The cumulative background noise reduces the signal 2. You turn the volume up (as the amateur radio people will have to and although entitled too don't wish too because it causes other users problems)
When you turnt he volume up, they all have to talk louder, so you get a fight between high and higher BPL power (to avoid radio wiping out internet, and higher and higher radio power for the same reason). At which point nobody can communicate usefully and lots of third parties are harmed.
HF interference isn't just an amateur radio problem either - you might well find you get 802.11 dead zones if you are near a power line using it. You may not be able to use radio controlled toys in an area with too many power lines and so on. Finally HF is essential to things like flying medical services and some rural communcation systems.
It all gets quite messy when this happens because good radio practice is the lowest possible power. The lower the power you can use the more people can use the same frequency. If everyone has to use 1KW then you'll get a lot less frequencies.
I'd also say their description of the FCC is in tune with its historical decision making - just look at the monopolisation of US commercial radio and the continued unneccessary exclusion of most small transmitters which could exist and other countries have proved are not a problem. Of course BPL background noise might well wipe out the scope for very low power radio stations too.
BTW: BPL trials in the UK (way before the US) were shelved for several reasons but intereference was a big one.
It shouldn't be insoluble - one nice property of radio is that if you can get the BPL encoding frequencies high enough then the interference problems become much less of an issue.
(PS: I defy you to find a radio astronomer who won't use expletives when asked abtut BPL..)
Let us hope the anti-spyware legislation in the US has bothered to require that the license/install permission is clear. As a word of warning btw some Linux vendors do seem to like adding autorun functionality and on older systems defaulting it to enabled.
Liner notes will also create them a problem since they then have to deal with returns from people who do not accept the license, aside from the problem that many discs are sold to minors which means that if you sell it to a 12 year old who puts it in Dad's computer you might not have the authorization to do so.
Seriously - when was the last time a footballer got promoted from the team to do the paperwork, how many CEO's secretaries outrank the CEO.
If you are the natural team leader then its unlikely the team will listen ot the manager anyway, they'll listen to you. So don't hire yourself a manager, hire yourself an assistant. Someone who goes to meetings for you, plans schedules for you and lets you get on with the real job. That doesn't have to be someone who is in charge of or controlling what you do but someone who enjoys doing the bits you don't and you can work alongside.
So many IT companies seem to screw this up. Good project managers are great people to have but they don't have to be in charge.
I found the assertion that 64bit PC's don't use the BIOS rather amusing. Evidently bits of Intel still haven't managed to bring themselves to admit the existance of Athlon64 just yet.
RAID5 is a much bigger performance drain in most setups, its also pushing boxes up to 4+ disks (realistically raid5 you need a hot spare) and that pushes it out of 1U/2U and mini cases.
IDE is so cheap you might as well just buy two big sata drives for most usage. Do make sure you buy two drives from two different vendors - its really embarrassing when you use two identical drives with near serial numbers and they fail the same day.
Also keep external backups. One place I worked we lost an entire array and the hot spare to a PSU failure. No backups.. thankfully it was the usenet spool
Radio stations provide value, so record companies pay them to play music. Now the RIAA is discovering that ClearChannel is more powerful than them and they are stuck in the middle of a problem of their own making.
I would like to see more innovation in the music sales model - things like radio stations being given free plays providing they give the URL you can buy the CD etc.
If you pass laws against offshoring then companies will simply avoid them. If they can't offshore they can outsource (as many already do). If they can't outsource they can move their executive out of the USA too.
If you want to take jobs back from other parts of the world you have to be better at it than they are. Software is currently mostly labour costs so that makes it very very hard indeed. Other jobs can be done many ways - even heavy industry has fought back by things like mechanisation. If you want the callcenter jobs back - automate it, get machine voice handling to the point it doesn't need many people.
- A thinkpad 600 with 192Mb of RAM - A VIA C3@533Mhz with 512Mb of RAM
Both are running Fedora 2 both are most definitely usable. There are only a few changes I've made to get that to happen - firstly I rebuild Gnome with gcc -Os, secondly I don't start up the 500 fascinating daemons I seem to get by default now days.
OpenOffice chugs on the TP600, but the VIA is very happy. It's not quite the same as a dual opteron with scsi where "startx" produces the entire running desktop in 2 seconds.
I've also been benching the systems. The 2.6 kernel is snappier than 2.4, and Gnome 2.6 is using less RAM than 2.4. The biggest bottleneck is disk seeking - Gnome loads a lot of scatter little files when starting up and disk heads are still constrained by little problems like momentum.
With XFce I can go down to about 48Mb and have a snappy desktop. Open Office isn't very funny at 48Mb but XFce but abiword is usable.
The first recorded programmable computer systems I am aware of that had control structures (loop count) were loom machines which while never used von-neumann style (humans punched the instructions the machine didnt weave new tapes) had the basics we consider today although very ad-hoc since they were built for real work rather than by computability theorists.
Selecting a "first" is extremely hard. If your definition is turing completeness then speech is turning complete so people probably win (although I'll leave turning completeness of animal brains to someone who knows more about the field 8)).
Personally I think that like a lot of other things in the universe there isn't a first because it evolved step by step.
NX is page table level. It is very hard to make use of segment based protecting in operating systems using 32bit flat paging modes (like all of the modern ones). Solar Designer produced some patches that try and do this at least for non exec stack but they require magic kernel side exception fixups.
Under Linux at least you can ask for executable mapped pages. This is what the fixed X loader does for x86 now. Most non x86 processors have execute bits on page table entries and POSIX/SuSv2 therefore have a MAP_EXEC bit in mmap so you can say "I want to run this"
The gnome 2.6 one also handles remote files. In fact one thing I am really really glad to see finally supported properly is WebDAV and https:// webdav too. It makes a lot of remote working much much easier.
All we need now is a decent webdav server that handles userids properly.
Real men know they can just drag and drop the file fron nautilus onto the editor. It's amazing what you learn by watching users 8)
There seem to be two things in Gnome 2.6 that annoy people - the spacial mode in nautilus (which is configurable anyway) and the file selector. I'd dearly like to see the whole file selector business go away and be replaced by a nautilus window of the right kind of files in the right location (where location is relevant)
After all why should someone have to learn *two* ways to select files ?
All the 2.6 distros so far have the problem. The 2.6 kernel changed the way the kernel thinks about partition geometry for setting up tables. Parted and friends had a few problems with the change.
It bites very few boxes because almost nobody uses C/H/S nowdays unless they force it in the BIOS
One of the other problems with testing this sort of bug is that Windows XP gets upset if you try and reinstall it 100 times.
Slashdot Mirror
I've been involved in multiple situations in previous jobs where we contracted for companies who turned out to have extremely dubious codebases, generally provided by external companies, acquired with small acquisitions and so on.
The most notable one was Sonix (then bought by 3Com). Both Sonix and 3Com had definite policies and good practices but they had ended up with code, stripped of headers that came from Phil Karn's KA9Q.
I've also had a case where a companiy had reference code/windows drivers and wouldn't release it to help a Linux driver not because they didnt want Linux drivers but because their own engineers worked from an internal source pool that they thought was very unlikely not to contain large amounts of 'acquired' code.
On the whole it has been the smaller companies that cause the most problem - but not always. A look at Harald Welte's list of people who settled GPL violations contains some very large and reputable computer companies.
OpenDLM is the lock project. There are also people workin on an SSI project.
I don't think its fair ro say we are close to VMS, we have an awfully long way to go yet 8)
Intermezzo and CODA try and solve a different problem (the one AFS does), they replicate data as much as possible without violating coherency and at a file level.
GFS instead gives everyone access to the same disk at the same time rather than replication. Both methods work well for different data sets - so yes GFS and oMFS are similar
Microsoft OLE DB Provider for SQL Server error '80040e14'
/mmrp/includes/common.inc, line 108
Line 1: Incorrect syntax near '='.
Guess its running on an HP laptop they've not fixed yet 8)
I think the other people have covered the basics pretty well - plug lots of computers into one fibrechannel or possibly firewire disk or disk array.
;)) or the same with uml, Zen, etc
The second really interesting use is with virtualisation - imagine if you want all your S/390 virtual machines to share the same bsse file systems for efficiency (given the price IBM charge for mainframe disks
It shows how much the music industry owns the US government that the bill cites the 31% loss claims of the music industry in a law, and ignores the fact that all the academic studies, as well as the statistics gathering companies rebut the data rather convincingly.
Do you want the head of
- The engineer who made the original mistake
- The document writer who may have caused it
- The manager who failed to do enough checking
- The QA people for missing it
and so it goes on.
Kudos to Intel for simply saying "We screwed up" and recalling products. They seem to have learned much from the old Pentium FPU errata handling.
Presumably its outsourced to texas then. Nothing says Indiana can't outsource its government to Texas does it ? I mean we seem to have outsourced ours to Washington
But you can't use it for email - if one spammer sends you an image in an email or a bit of midi you've breached the contract by downloading your email
Alan
Most of the HF bands are used for commercial purposes, some slots are allocated for limited private use (eg CB, remote control toys, 49Mhz walkie-talkies), and lots of it is used for emergency and longer ranger services where VHF/UHF simply won't do the job. This includes people like emergency services.
Amateur radio is probably more relevant now than since the 1940's. Its real reason for existance beyond the first uregulated days of "gee isnt this neat" was to provide a steady supply of wireless operators to draft in the event of a war. Its not the only reason but its a major reason it survived.
Alan
Shortwave radio communication over any long distance (commercial, military and hobbyist) often deals with weak signals. Each broadband power line adds to the background noise cumulatively raising the problem. One power line won't trash your TV signal (unless you are very close), but each one adds noise until all you have is snow.
Its like people talking in the background - a couple of people don't do much harm but when you try and talk across a room full of quietly talking people two things happen
1. The cumulative background noise reduces the signal
2. You turn the volume up (as the amateur radio people will have to and although entitled too don't wish too because it causes other users problems)
When you turnt he volume up, they all have to talk louder, so you get a fight between high and higher BPL power (to avoid radio wiping out internet, and higher and higher radio power for the same reason). At which point nobody can communicate usefully and lots of third parties are harmed.
HF interference isn't just an amateur radio problem either - you might well find you get 802.11 dead zones if you are near a power line using it. You may not be able to use radio controlled toys in an area with too many power lines and so on. Finally HF is essential to things like flying medical services and some rural communcation systems.
It all gets quite messy when this happens because good radio practice is the lowest possible power. The lower the power you can use the more people can use the same frequency. If everyone has to use 1KW then you'll get a lot less frequencies.
I'd also say their description of the FCC is in tune with its historical decision making - just look at the monopolisation of US commercial radio and the continued unneccessary exclusion of most small transmitters which could exist and other countries have proved are not a problem. Of course BPL background noise might well wipe out the scope for very low power radio stations too.
BTW: BPL trials in the UK (way before the US) were shelved for several reasons but intereference was a big one.
It shouldn't be insoluble - one nice property of radio is that if you can get the BPL encoding frequencies high enough then the interference problems become much less of an issue.
(PS: I defy you to find a radio astronomer who won't use expletives when asked abtut BPL..)
Let us hope the anti-spyware legislation in the US has bothered to require that the license/install permission is clear. As a word of warning btw some Linux vendors do seem to like adding autorun functionality and on older systems defaulting it to enabled.
Liner notes will also create them a problem since they then have to deal with returns from people who do not accept the license, aside from the problem that many discs are sold to minors which means that if you sell it to a 12 year old who puts it in Dad's computer you might not have the authorization to do so.
Seriously - when was the last time a footballer got promoted from the team to do the paperwork, how many CEO's secretaries outrank the CEO.
If you are the natural team leader then its unlikely the team will listen ot the manager anyway, they'll listen to you. So don't hire yourself a manager, hire yourself an assistant. Someone who goes to meetings for you, plans schedules for you and lets you get on with the real job. That doesn't have to be someone who is in charge of or controlling what you do but someone who enjoys doing the bits you don't and you can work alongside.
So many IT companies seem to screw this up. Good project managers are great people to have but they don't have to be in charge.
I found the assertion that 64bit PC's don't use the BIOS rather amusing. Evidently bits of Intel still haven't managed to bring themselves to admit the existance of Athlon64 just yet.
RAID5 is a much bigger performance drain in most setups, its also pushing boxes up to 4+ disks (realistically raid5 you need a hot spare) and that pushes it out of 1U/2U and mini cases.
IDE is so cheap you might as well just buy two big sata drives for most usage. Do make sure you buy two drives from two different vendors - its really embarrassing when you use two identical drives with near serial numbers and they fail the same day.
Also keep external backups. One place I worked we lost an entire array and the hot spare to a PSU failure. No backups.. thankfully it was the usenet spool
Radio stations provide value, so record companies pay them to play music. Now the RIAA is discovering that ClearChannel is more powerful than them and they are stuck in the middle of a problem of their own making.
I would like to see more innovation in the music sales model - things like radio stations being given free plays providing they give the URL you can buy the CD etc.
If you pass laws against offshoring then companies will simply avoid them. If they can't offshore they can outsource (as many already do). If they can't outsource they can move their executive out of the USA too.
If you want to take jobs back from other parts of the world you have to be better at it than they are. Software is currently mostly labour costs so that makes it very very hard indeed. Other jobs can be done many ways - even heavy industry has fought back by things like mechanisation. If you want the callcenter jobs back - automate it, get machine voice handling to the point it doesn't need many people.
My main desktop machines are
- A thinkpad 600 with 192Mb of RAM
- A VIA C3@533Mhz with 512Mb of RAM
Both are running Fedora 2 both are most definitely usable. There are only a few changes I've made to get that to happen - firstly I rebuild Gnome with gcc -Os, secondly I don't start up the 500 fascinating daemons I seem to get by default now days.
OpenOffice chugs on the TP600, but the VIA is very happy.
It's not quite the same as a dual opteron with scsi where "startx" produces the entire running desktop in 2 seconds.
I've also been benching the systems. The 2.6 kernel is snappier than 2.4, and Gnome 2.6 is using less RAM than 2.4. The biggest bottleneck is disk seeking - Gnome loads a lot of scatter little files when starting up and disk heads are still constrained by little problems like momentum.
With XFce I can go down to about 48Mb and have a snappy desktop. Open Office isn't very funny at 48Mb but XFce but abiword is usable.
The first recorded programmable computer systems I am aware of that had control structures (loop count) were loom machines which while never used von-neumann style (humans punched the instructions the machine didnt weave new tapes) had the basics we consider today although very ad-hoc since they were built for real work rather than by computability theorists.
Selecting a "first" is extremely hard. If your definition is turing completeness then speech is turning complete so people probably win (although I'll leave turning completeness of animal brains to someone who knows more about the field 8)).
Personally I think that like a lot of other things in the universe there isn't a first because it evolved step by step.
Alan
NX is page table level. It is very hard to make use of segment based protecting in operating systems using 32bit flat paging modes (like all of the modern ones). Solar Designer produced some patches that try and do this at least for non exec stack but they require magic kernel side exception fixups.
Under Linux at least you can ask for executable mapped pages. This is what the fixed X loader does for x86 now. Most non x86 processors have execute bits on page table entries and POSIX/SuSv2 therefore have a MAP_EXEC bit in mmap so you can say "I want to run this"
The gnome 2.6 one also handles remote files. In fact one thing I am really really glad to see finally supported properly is WebDAV and https:// webdav too. It makes a lot of remote working much much easier.
All we need now is a decent webdav server that handles userids properly.
Real men know they can just drag and drop the file fron nautilus onto the editor. It's amazing what you learn by watching users 8)
There seem to be two things in Gnome 2.6 that annoy people - the spacial mode in nautilus (which is configurable anyway) and the file selector. I'd dearly like to see the whole file selector business go away and be replaced by a nautilus window of the right kind of files in the right location (where location is relevant)
After all why should someone have to learn *two* ways to select files ?
Actually in most cases windows will simply remove the existing boot loader.
I guess you could install Linux, install windows and install linux again or rescue the Linux install 8)
All the 2.6 distros so far have the problem. The 2.6 kernel changed the way the kernel thinks about partition geometry for setting up tables. Parted and friends had a few problems with the change.
It bites very few boxes because almost nobody uses C/H/S nowdays unless they force it in the BIOS
One of the other problems with testing this sort of bug is that Windows XP gets upset if you try and reinstall it 100 times.