In my case I couldn't even tell them why I was calling because it was collections, I can't tell them who I am until they confirm they are the customer (not anyone elses business that they have debt with Cap One). What is even worse and I hate more, is the systems that call you without there being a person, it is a message and if you want more info press a button and we'll connect you with someone. Man they can't even pay someone to call you now (outbound calls typically were in the 20sec range per for collections, $10/hr/180 gives ~5 cents per call in employee cost and they won't even pay that). That is why I won't do business with them, if my business isn't worth 5 cents to you then I have a pretty good idea what the level of support for the product is going to be.
Is it just me or does Apple tend to spend so much time making things pretty that they are obsolete before they ship (as opposed to a PC that is obsolete 1 month after coming out;))? I want my month back!
Seriously though, my business laptop has a 100GB HDD and I purchased it about 15 months ago (and at the time a 100 GB drive was about mid range). A new laptop that isn't even available yet is going to ship with an 80GB, what the heck? No optical drive, I don't think that has been the case for ~10 years. Conclusion: you can make things really pretty and small if you want to use old hardware/remove major components, duh.
2. When the help desk guy was assigned to make the followup call, he didn't notice and find it odd that the original call came in 10 years ago? He didn't call his supervisor over and say, "hey I think somebody made a mistake here! Maybe we should just close this out." They probably have an autodialer, the agent didn't even see the ticket before the system called the guy. I worked for a Capital One call center for a while. I was real nice when the systems were slow: "Hi... is... Steve Johnson there" I must have sounded retarded but it was actually that I was waiting for the account to come up so I knew who "I" just called.
Tying the format closely to the tools meant to process it, to the point of embedding the processing code in the data, is one of the design blunders perpetuated by Microsoft, which gave us such wonderful 'innovations' as Word Macro Viruses. I think it is the other way around. MS uses an object model for the app and embeds the data in it. Example excel.workbooks.item(i).currentsheet.offset(1,2).value = "Total". The underlying structure of the app determined how the data is accessed in Office. Can the data and the structure be seperated? Sure with some sort of markup language I suppose (MS XML anyone?). I actually like having things organized the way they are now, I find it pretty easy to guess where things will be, eg. currentsheet.buttons, where would the drop downs be, why currentsheet.dropdowns. What would be real nice and probably is what they are trying to do, would be to encapsulate office into.Net so that it isn't a pesky COM object that you can't get intellisense for in the.Net languages.
I'm not an expert in the region and growing of asian produce, however, when I was in Japan visiting a friend I was amazed at how low tech the farming is there something ca 1920 US. They were planting rice by hand. We were biking through rural areas and I noticed a lot of older people that were bent over at a near 90% angle on the side walks. I asked my friend what was wrong with them, and he said they were farmers who's backs have become deformed from bending over all the time.
The reason for this stupidity according to my friend: Japan (as admittedly a lot of countries do) has protectionist policies in place forcing rice and other crops to have to be produced in Japan. The cost of japanese rice versus the average is 7 times, but they have to produce it. Apparently it is cheaper to ruin the backs of people in a first world country then to risk not having control of your food supply - because you buy it from the close third world countries - in the advent of a war. Admittedly if I was like Japan and had crazies for neighbours (North Korea/China, government not the people) whom I recently pissed off by invading I might be worried about war preparedness as well.
If the person isn't interested they aren't interested. Software is not a religion, you aren't right and have to beat them over the head until they agree with you. If they like commercial software let them use it. For example I already know how to use Office, I know my Office documents will work with my co-workers office documents. I make $40 an hour. "Free" is very relative, if I have to spend 5 hours figuring out how to use the software it just cost my employer $200, we get office Pro for ~140 (we have 3k+ workstations and are on the healthcare discount plans).
Anyways, some people don't like to learn how to use software they just want it to work the way they are used to (part of the problem with Office 2007 IMHO). If they are willing to try it I let them know there are free alternatives, but if they want Vista they can go out and buy it. It's up to them and what they are comfortable with.
The free arguement (free as in speach or as in beer), doesn't work. If that was the way people thought everyone would be brewing their own beer because it is cheaper and they have control over it. Saving money for money's sake is stupid. You have to have something better to spend it on (or save it), if the good feeling you get from buying something that everyone else uses or the savings of time learning new software is worth it for you then that is where you spend the money. If it's not then spend it elsewhere, but no anwser is "right" in an absolute sense.
A hospital with 3k employees, probably only around 300 are remote access types though (the doctors, managers and IT). The problem is peoples positions change. I was working an office only job prior to this one and signed on for a loose 2 year contract (I get 10% off both cable and internet but the proviso is if I drop either services I get retroactively billed for the discount amount for the period), it was fine for my use and I wasn't even aware of the packet filtering at time of sign up. But man when I started doing DB queries that pull a few MB of data interactively did I ever notice the lag. Things lately have gotten better. I know they were increasing my pipe size (from 6 to 8Mbps) and in general the whole system. I don't know if the policy was stopped or they just have enough bandwidth now to properly service the customers in my area and so the filtering high water mark hasn't been hit lately.
The real problem is that an ISP can change their policies at any time, so what you recommend this month, and people sign up log contracts with - pay for installation or whatever that gives them a lock in effect - can be the worst solution a month later. Since they advertise "up to XMbps", and the terms include "will not use it for a server" among other clauses they are covered. Heck IM can be considered a server. So you create a stink and they accuse you of breaking the terms of use.
Is it the postal service's responsibility to open every package and check what's inside, in case I'm trying to send you a photocopied novel? Exactly. Does the mall have to inspect your car to make sure you own it and there isn't any stolen goods in it? What responsiblity does the one providing the pipe have for what you put through it? It doesn't seem like any to me. If they want to sue me (or I suppose criminal if it is considered fraud or whatever) then they can hire an investigator, get a warant/suepeona etc like the rest of the world. If you don't have a case get out of my way.
wouldn't all these techniques be rendered useless by encrypted tunneling software short of making encryption over the internet illegal in itself? While not making it illegal per say many ISPs reduce the bandwidth for encrypted traffic for just this reason. Torrents announce themselves as such as part of the protocol, they were getting throttled, so they got smart and started encrypting their traffic. Guess what? My ISP started throttling all encrypted traffic, alright for say doing banking, but when I'm using VPN to get into my work and do remote desktop, db access etc, it really sucks.
What is funny is just 6 hrs ago I saw a post saying digital download of BluRay wouldn't work for years because even at 36Mbps (which isn't available) it wouldn't be a live feed but would take a day to download a show. Man tech moves fast;)
Where I really feel the pain is the hosting companies. I mean my company just purchased a new uber expensive top of the line 4Gb dual channel SAN. We aren't even close to having 4Gb/s FC cards in all our servers. At this transfer rate you'll saturate the FC cards with only 10-40 users per server, ouch. Your high end SCSI disk array? It will saturate too, unless you have a very high end RAID controller.
I think this might help promote file sharing to be honest. The idea of going to a website and hitting them with your 20MB/s load just won't appeal to anyone unless they are billing you by the MB or something (some porn). Speaking of which it will really suck for those porn companies that offer 3 day trials or whatever, in those three days you could download everything they have. Then of course everyone would throttle your rate, and the bitching about not being able to use the service your paying for would begin again.
Yes but real people don't have genitals the size of 42" screens and razor pimples the size of tennis balls... True but your arguement is flawed. If you have the 42" screen, what are you supposed to do, reduce your effective scan size when viewing porn, or have humugous pixels? If the screen gets bigger you need more resolution to "fuzz out" the jaggies caused by pixellation. As for the previous post I agree, HD is a good thing. You can sit back further and just see normal detail if you want, but you also can sit close and see the pores in the skin. It all depends on what turns you on, and if the porn industry has proven anything it is that people are turned on by a lot of strange things.
I think it is "open" in that you can install the most recent version of the software when it comes out, or a previous version
"value" as in cheaper than a normal purchase, and much better than buying the current version and having another one come out next year that you regret not waiting for
"subscription" doesn't need explanation.
This might be broading the customers that can opt in for the program, but we've had something pretty similar for years in my industry (healthcare). I think they changed the name of our program to something like Health Select or something. Anyways same "stuff" different wrapping. In our case it is going to screw us around a bit. Because our current licensing doesn't exist anymore, we can't roll over our current software on to a new 3 year lease. So we have to buy everything again at the full rate. They still are giving us around a 50% discount so I guess I can't complain too bad, I just wish marketing guys would drink some of the Kool-Aid sometimes rather than having to use the hot spin words of the year all the time, and trying to sell the same crap with a different name. It's not a new GUI, it is a "enhanced user experience".
Right on. Example bittorent clients. I searched on sourceforge, there alone I found 302 clients. Does the world really need 302 open source bit torrent clients?
I tried to start a scientific modelling program on source forge once. I got a few good people, we started designing the specs and then people started to get inactive. There in lies the problem, hurding cats indeed.
An employer has a greater chance of keeping good people than a FOSS project. First, they are the source of the employees livelyhood, the employee would at least need to spend some of there spare time looking for another job before they'd quit, thus creating a "barrier to exit". There is also the employees need for good references, which helps reduce the chances that the employee will just stop showing up one day. Chances are they will help document/train the next guy.
One could also say there is a barrier to entery as well as you have to go through an HR process where they try to screen out underperformers. It is really hard to do that in FOSS, perhaps you could do it for the really big projects, such things as must have proposed a bug fix to the project, must have been at user group meetings etc. I think I've seen similar requirements on MySQL's corporate job postings (but that is kind of a hybrid OSS project anyways). At any rate, trying this for the little 'proof of concept, we're not sure if anyone will use it projects', good luck.
Another example of open source (or at least open design) software failing to really improve would be X-windows. It was orginally a research project and never was intended for large systems. Why do all *NIX systems insist on building on top of this garbage? To this day I still have bugs, and hopefully timely patches, for X-window run away processes for the UNIX systems I administrate. This for a 30+ year old technology. Nothing quite like finding out from your users that for the last week one of the programs has been running slow, and log into the server and see X-windows was hogging a whole CPU (yes on the server where their is only GUI based apps running ~1/month).
Sorry I was incorrect. I checked with a collegue, protons can be more accurately delivered to a target, but it is pretty obtuse physics to figure it out. It has to do with the scattering cross-section of the material. Think of a pool table. Your chances of hitting something depend on the density and size of the balls on the table, the size of the ball your using, the angle etc. Similarly with particles, except because of quantum mechanics you now have a probablity density for electrons (the balls on the table), and the type and energy of the particle taking the place of the size of the que ball.
While protons at low energy get assorbed really fast, at higher energies they have enough speed to "ignore" the electron clouds they are passing through. Actually they "see" a very low cross-section for electron-proton interaction, so the probablity of them hitting something and depositing energy is low. They deposit energy at a roughly constant rate until they hit a critical velocity at which point the cross-section of the electron cloud goes way up (they are slow enough to "see" the electrons) and they interact much more frequently depositing almost all of their energy in a very narrow window. This is called the Bragg Peak
http://en.wikipedia.org/wiki/Bragg_peak . Proton therapy still relies on the electrons that get ripped off of atoms to do the damage as mentioned in my previous post.
There is still some technical challenges with proton therapy though, first: the accelerators/energies you need need larger space than current X-ray based accelerators (and most US cancer centres are small, privately run 1-2 X-ray centres), more shielding (accelerating charged particles admit radiation) over the whole length of the accelerator. Add to this, in order to paint the target with the dose you want you need to either/or/both tune the energy, or adjust the target (by progressively adding more material in front of the patient, etc. filling a bucket of water or something during the treatment to change the effective depth). These aren't impossible to overcome, just would make the treatments more complicated and time consumming than current treatments (but also potentially of better quality) and more expensive.
Another plug: more expensive, more time consumming is the norm in the field. The CT like functionality I mentioned in my earlier post, as well as the simulation methods being utilized (moving more to a tell the system what you want and it optimizes it for you, rather than guessing your self and then doing a dose calculation) are requiring more CPU horse power and more storage. Eg. at my centre we've seen the average patients record, go from mostly the initial CT + plan (~100MB) to closer to 400MB(loads of data generated during the optimization process + a couple more "CTs" after the treatment starts to verify tumor response/position. We treat about 2000 patients a year so just patient data is growing at about 1TB a year this isn't counting research, etc. Everything gets bigger, faster, higher resolution etc on a nearly yearly basis, requiring more network, SAN, server horsepower etc. All equals: job security.
I'm a physicist myself by training, currently doing IT work for a radiation treatment program. Yes there is side effects with any type of treatment. Chemo can cause organ problems among other things. Surgical can "miss something", can actually poke the tumor and cause it to spread etc. Radiation can cause new tumors to grow, can damage bones (causes them to become brittle), and other organs. Example: prostate treatments you have a choice of where to put the beam. left and right are the femeral heads (tops of the leg bone), a little above is the rectum and bladder. Common side effects (not gauranteed mind you just a chance of) permanent rectal damage which causes you to have diarhea and other gasto intestinal problems, harding of the bladder which causes it to function improperly. What you gain is: less pain due to surgery, less to no risk of infection, similar to better survival rate, no downtime (ie you can still go to work assuming your of the age that you do).
For breast cancer the story is even better. The longterm survival rates for most breast tumors is identical between a masectomy and radiation. Bonus with radiation is: you give some dose to the surrounding tissues potentially killing secondary malignancies, and of course the woman still has a breast. The best treatment as far as survival goes is a lumectomy but this usually can only be done early stage (hence all the focus on breast screening).
Proton therapy has some potential, however, the main articles claim that protons are accurant where as X-rays are inaccurant is miss leading/wrong. Protons don't penetrate as far into tissue. Thus the radiation is more targetted for superficial treatments. However, the opposite is the case for treatment at depth. Since the protons deposit their energy relatively close to the surface, you'd need a much higher total dose to treat an internal organ then with X-rays. It really would depend on the malignancy and how much you want to spread the radiation over. Different types of radiation (and different energy levels, ie. 6MeV X-rays, 15 MeV X-rays), have different dose build up and fall off profiles. It really depends on where the tumor is which one will be the best.
In typical clinical process there is the point of interest (POI) and the target volume. These differ and sometimes by a wide margin. The target (ie where the radiation actually goes) is larger than the actual area that the oncologist thinks localizes the tumor. This is to allow for alignment errors (patient moves, machine tolerances etc), plus a safety margin (typically around 7mm) to try to get the stray cancer cells around the tumor. Anyways, the physician will prescribe a certain dose at a point in the target, and a percentage at a isosurface, so say 1000 Cy, 80% at POI + 7mm. At any rate you may not use the accuracy because you want some dose "off target" for localized tumor control (as apposed to general tumor control that you attempt with chemo).
X-ray treatment machines are much more accurate than claimed. Modern treatment units have among other things MLC (multi leaf collimaters) to shape the beam. You can thing of the beam like a light bulb, they place a bunch of retangular peices of metal between the patient and the bulb to get the desired shape. These are dynamically tuned so at one angle you might get 20% of the dose at a particular shape, and the rest at a different shape say, or they can be in motion during the "beam on" painting a more uniform distribution. At anyrate, a variety of angles (now in the works to use continuous arcs as well), and leaf positions enable you to paint the target with the 3 dimensional dose distribution you want with about 3mm spatial accuracy (add ins/some systems out of the box have about 5mm accuracy even after accounting for breathing in realtime on say a lung tumor), while spreading out the dose along the healthy tissue which reduces risk to the other areas. Also, newer machines can get an add on for about 600k that will enable the unit to act as a CT scanner (typical price 2-4M),
Yeah because user workstations tend to be windows. Perhaps SSH would be better? How about you can use (insert appropriate remote protocol here) to access your systems? Funny thing, I work in healthcare, some vendors still use telnet for access and actually have ssh access disabled. That is crazy, I can see having no remote access, I can see only having SSH access, but why in the world would you choose to only have telnet access? This software is about 50k per workstation and comes from Philips, so is neither a small company thing, or a cheap little peice of junk thing. Weird.
No not at all;) Actually the last three posts I got in reply mentioned unions, and fair enough. I read a article recently (can't remember where, though I subscribe to TechRepublic and the Code Project so I wouldn't be supprised if it was one of them) about this kind of thing. The arguement was that IT people want to be treated as "professionals" not lowly hourly workers, so are more likely to take a salary position, and put up with working 50+ hr weeks even though in most places even salary employees are entitled to get paid for their time if it is over 44hrs. Anyways, it was well reasoned, and does point out some of the silly things people in the field do.
I might be a rarity, I'm an hourly IT worker. I bill for my time if I get called back in, at least 4hrs plus travel expenses. My reasoning is: getting to work once a day I pay for, if you want me back you pay for it. So far my boss has lived with that. I'm a one of though, so have some leverage. Funny, the finance department said they didn't have enough work for a second IT guy, so instead pay me OT and travel expenses out the wazoo:) As for avoiding salary my reasoning is: if my first 40 hours was worth 75k to you, then my next 40 hours is worth at least another 75k to you. A factory gets a rush job they bill the customer more, and pay their employees extra. A factory machine goes down, they pay the millwrights OT to get it up and running asap. Same should go with IT. I don't mind putting the hours in, but I'll be damned before I'll burn my gas and spend my evening working for free.
Hmm, I agree with most of what others say in that you're going to need to seperate your office and your network stuff and allow sufficient space for the network stuff to grow. A/C might not be enough, I'm looking at building a mini datacenter for my department (2 racks, 4 Sun workstations 5 PC's), even with that our plant ops said we'd need proper cooling installed. You'll also need space behind and in front of the racks so count on about 1/3rd of the datacenter space being racks, the other 2/3rds will be a server length front and back, add extra space for cooling if needed (they might be able to drop a cooler in the ceiling so you wouldn't need the floor space for it then).
Also, how important is your equipment? Assume the room were to disapear what would happen? If the answer is the department/company will be inoperable until the new equipment arrives and you are able to restore stuff from dumps then you both need some redundancy, and are going to want a proper fire suppression system. If you use a neutral gas suppression your going to need an airtight space for the servers (which will probably help with the noise too).
If your job is anything like 90% of IT professionals, you'll spend the minority of your time in your datacenter, that is what terminal services is for. Also, think about all the time spent sourcing vendors, doing project management work, any coding you have to do etc. I'd really fight to have the equipment put somewhere else if I was you. Figure out what percentage of time you need physical contact with the servers/switches and let your boss know. I mean it doesn't make sense to live with the noise and run at reduced efficiency for 80% of your work day for the sake of 300 sqft or so of floor space, if there is any way to avoid it. Also, if the sound proofing ends up sucking, you'll have to take phone calls and stuff with that crap in the background.
It is funny, companies worry about their 20k per year factory workers when it comes to noise, and environmental conditions, but the 60-150k IT guy gets the shaft. Apparently server noise has a different effect on the ears than machine noise.
We are a start up and as such we can't pay you, but we promise once we are profitable we will. Unfortunately VC and bank capital aquisitions aren't going so well as we spend all our executives' time coming up with ingenious ways to recruit people.
signed, dumb ass CEO
anonymous Inc
I mean the average person won't know were to look in the system, or won't care about the number of MRE's being sent to Zimbabwai. For the most part isn't the real threat that information will get to the people that will know how to interpret it and care? In that case won't the foreign governments/terrorists just learn how to exploit Mac's rather than Windows? This seems a lot like security by obscurity, except, well Macs aren't that obscure.
But we haven't seen anything come out of MS Research that's made a significant difference in anyone's lives. Maybe not in heavy use yet, but how about Linq, parallel support for.Net? I think these will be huge in the industry, at least as big as ADO, and pthreads has been. I seem to recall reading a few interesting articles about software engineering methodologies too (the concept of "zero bounced bugs", some theory into how to estimate the number of bugs given the historical tread of bugs over the project etc). Maybe not earth shattering but interesting.
Slashdot Mirror
In my case I couldn't even tell them why I was calling because it was collections, I can't tell them who I am until they confirm they are the customer (not anyone elses business that they have debt with Cap One). What is even worse and I hate more, is the systems that call you without there being a person, it is a message and if you want more info press a button and we'll connect you with someone. Man they can't even pay someone to call you now (outbound calls typically were in the 20sec range per for collections, $10/hr/180 gives ~5 cents per call in employee cost and they won't even pay that). That is why I won't do business with them, if my business isn't worth 5 cents to you then I have a pretty good idea what the level of support for the product is going to be.
Seriously though, my business laptop has a 100GB HDD and I purchased it about 15 months ago (and at the time a 100 GB drive was about mid range). A new laptop that isn't even available yet is going to ship with an 80GB, what the heck? No optical drive, I don't think that has been the case for ~10 years. Conclusion: you can make things really pretty and small if you want to use old hardware/remove major components, duh.
The reason for this stupidity according to my friend: Japan (as admittedly a lot of countries do) has protectionist policies in place forcing rice and other crops to have to be produced in Japan. The cost of japanese rice versus the average is 7 times, but they have to produce it. Apparently it is cheaper to ruin the backs of people in a first world country then to risk not having control of your food supply - because you buy it from the close third world countries - in the advent of a war. Admittedly if I was like Japan and had crazies for neighbours (North Korea/China, government not the people) whom I recently pissed off by invading I might be worried about war preparedness as well.
Anyways, some people don't like to learn how to use software they just want it to work the way they are used to (part of the problem with Office 2007 IMHO). If they are willing to try it I let them know there are free alternatives, but if they want Vista they can go out and buy it. It's up to them and what they are comfortable with.
The free arguement (free as in speach or as in beer), doesn't work. If that was the way people thought everyone would be brewing their own beer because it is cheaper and they have control over it. Saving money for money's sake is stupid. You have to have something better to spend it on (or save it), if the good feeling you get from buying something that everyone else uses or the savings of time learning new software is worth it for you then that is where you spend the money. If it's not then spend it elsewhere, but no anwser is "right" in an absolute sense.
The real problem is that an ISP can change their policies at any time, so what you recommend this month, and people sign up log contracts with - pay for installation or whatever that gives them a lock in effect - can be the worst solution a month later. Since they advertise "up to XMbps", and the terms include "will not use it for a server" among other clauses they are covered. Heck IM can be considered a server. So you create a stink and they accuse you of breaking the terms of use.
Where I really feel the pain is the hosting companies. I mean my company just purchased a new uber expensive top of the line 4Gb dual channel SAN. We aren't even close to having 4Gb/s FC cards in all our servers. At this transfer rate you'll saturate the FC cards with only 10-40 users per server, ouch. Your high end SCSI disk array? It will saturate too, unless you have a very high end RAID controller.
I think this might help promote file sharing to be honest. The idea of going to a website and hitting them with your 20MB/s load just won't appeal to anyone unless they are billing you by the MB or something (some porn). Speaking of which it will really suck for those porn companies that offer 3 day trials or whatever, in those three days you could download everything they have. Then of course everyone would throttle your rate, and the bitching about not being able to use the service your paying for would begin again.
"value" as in cheaper than a normal purchase, and much better than buying the current version and having another one come out next year that you regret not waiting for
"subscription" doesn't need explanation.
This might be broading the customers that can opt in for the program, but we've had something pretty similar for years in my industry (healthcare). I think they changed the name of our program to something like Health Select or something. Anyways same "stuff" different wrapping. In our case it is going to screw us around a bit. Because our current licensing doesn't exist anymore, we can't roll over our current software on to a new 3 year lease. So we have to buy everything again at the full rate. They still are giving us around a 50% discount so I guess I can't complain too bad, I just wish marketing guys would drink some of the Kool-Aid sometimes rather than having to use the hot spin words of the year all the time, and trying to sell the same crap with a different name. It's not a new GUI, it is a "enhanced user experience".
I tried to start a scientific modelling program on source forge once. I got a few good people, we started designing the specs and then people started to get inactive. There in lies the problem, hurding cats indeed.
An employer has a greater chance of keeping good people than a FOSS project. First, they are the source of the employees livelyhood, the employee would at least need to spend some of there spare time looking for another job before they'd quit, thus creating a "barrier to exit". There is also the employees need for good references, which helps reduce the chances that the employee will just stop showing up one day. Chances are they will help document/train the next guy.
One could also say there is a barrier to entery as well as you have to go through an HR process where they try to screen out underperformers. It is really hard to do that in FOSS, perhaps you could do it for the really big projects, such things as must have proposed a bug fix to the project, must have been at user group meetings etc. I think I've seen similar requirements on MySQL's corporate job postings (but that is kind of a hybrid OSS project anyways). At any rate, trying this for the little 'proof of concept, we're not sure if anyone will use it projects', good luck.
Another example of open source (or at least open design) software failing to really improve would be X-windows. It was orginally a research project and never was intended for large systems. Why do all *NIX systems insist on building on top of this garbage? To this day I still have bugs, and hopefully timely patches, for X-window run away processes for the UNIX systems I administrate. This for a 30+ year old technology. Nothing quite like finding out from your users that for the last week one of the programs has been running slow, and log into the server and see X-windows was hogging a whole CPU (yes on the server where their is only GUI based apps running ~1/month).
While protons at low energy get assorbed really fast, at higher energies they have enough speed to "ignore" the electron clouds they are passing through. Actually they "see" a very low cross-section for electron-proton interaction, so the probablity of them hitting something and depositing energy is low. They deposit energy at a roughly constant rate until they hit a critical velocity at which point the cross-section of the electron cloud goes way up (they are slow enough to "see" the electrons) and they interact much more frequently depositing almost all of their energy in a very narrow window. This is called the Bragg Peak http://en.wikipedia.org/wiki/Bragg_peak . Proton therapy still relies on the electrons that get ripped off of atoms to do the damage as mentioned in my previous post.
There is still some technical challenges with proton therapy though, first: the accelerators/energies you need need larger space than current X-ray based accelerators (and most US cancer centres are small, privately run 1-2 X-ray centres), more shielding (accelerating charged particles admit radiation) over the whole length of the accelerator. Add to this, in order to paint the target with the dose you want you need to either/or/both tune the energy, or adjust the target (by progressively adding more material in front of the patient, etc. filling a bucket of water or something during the treatment to change the effective depth). These aren't impossible to overcome, just would make the treatments more complicated and time consumming than current treatments (but also potentially of better quality) and more expensive.
Another plug: more expensive, more time consumming is the norm in the field. The CT like functionality I mentioned in my earlier post, as well as the simulation methods being utilized (moving more to a tell the system what you want and it optimizes it for you, rather than guessing your self and then doing a dose calculation) are requiring more CPU horse power and more storage. Eg. at my centre we've seen the average patients record, go from mostly the initial CT + plan (~100MB) to closer to 400MB(loads of data generated during the optimization process + a couple more "CTs" after the treatment starts to verify tumor response/position. We treat about 2000 patients a year so just patient data is growing at about 1TB a year this isn't counting research, etc. Everything gets bigger, faster, higher resolution etc on a nearly yearly basis, requiring more network, SAN, server horsepower etc. All equals: job security.
For breast cancer the story is even better. The longterm survival rates for most breast tumors is identical between a masectomy and radiation. Bonus with radiation is: you give some dose to the surrounding tissues potentially killing secondary malignancies, and of course the woman still has a breast. The best treatment as far as survival goes is a lumectomy but this usually can only be done early stage (hence all the focus on breast screening).
Proton therapy has some potential, however, the main articles claim that protons are accurant where as X-rays are inaccurant is miss leading/wrong. Protons don't penetrate as far into tissue. Thus the radiation is more targetted for superficial treatments. However, the opposite is the case for treatment at depth. Since the protons deposit their energy relatively close to the surface, you'd need a much higher total dose to treat an internal organ then with X-rays. It really would depend on the malignancy and how much you want to spread the radiation over. Different types of radiation (and different energy levels, ie. 6MeV X-rays, 15 MeV X-rays), have different dose build up and fall off profiles. It really depends on where the tumor is which one will be the best.
In typical clinical process there is the point of interest (POI) and the target volume. These differ and sometimes by a wide margin. The target (ie where the radiation actually goes) is larger than the actual area that the oncologist thinks localizes the tumor. This is to allow for alignment errors (patient moves, machine tolerances etc), plus a safety margin (typically around 7mm) to try to get the stray cancer cells around the tumor. Anyways, the physician will prescribe a certain dose at a point in the target, and a percentage at a isosurface, so say 1000 Cy, 80% at POI + 7mm. At any rate you may not use the accuracy because you want some dose "off target" for localized tumor control (as apposed to general tumor control that you attempt with chemo).
X-ray treatment machines are much more accurate than claimed. Modern treatment units have among other things MLC (multi leaf collimaters) to shape the beam. You can thing of the beam like a light bulb, they place a bunch of retangular peices of metal between the patient and the bulb to get the desired shape. These are dynamically tuned so at one angle you might get 20% of the dose at a particular shape, and the rest at a different shape say, or they can be in motion during the "beam on" painting a more uniform distribution. At anyrate, a variety of angles (now in the works to use continuous arcs as well), and leaf positions enable you to paint the target with the 3 dimensional dose distribution you want with about 3mm spatial accuracy (add ins/some systems out of the box have about 5mm accuracy even after accounting for breathing in realtime on say a lung tumor), while spreading out the dose along the healthy tissue which reduces risk to the other areas. Also, newer machines can get an add on for about 600k that will enable the unit to act as a CT scanner (typical price 2-4M),
Yeah because user workstations tend to be windows. Perhaps SSH would be better? How about you can use (insert appropriate remote protocol here) to access your systems? Funny thing, I work in healthcare, some vendors still use telnet for access and actually have ssh access disabled. That is crazy, I can see having no remote access, I can see only having SSH access, but why in the world would you choose to only have telnet access? This software is about 50k per workstation and comes from Philips, so is neither a small company thing, or a cheap little peice of junk thing. Weird.
I might be a rarity, I'm an hourly IT worker. I bill for my time if I get called back in, at least 4hrs plus travel expenses. My reasoning is: getting to work once a day I pay for, if you want me back you pay for it. So far my boss has lived with that. I'm a one of though, so have some leverage. Funny, the finance department said they didn't have enough work for a second IT guy, so instead pay me OT and travel expenses out the wazoo :) As for avoiding salary my reasoning is: if my first 40 hours was worth 75k to you, then my next 40 hours is worth at least another 75k to you. A factory gets a rush job they bill the customer more, and pay their employees extra. A factory machine goes down, they pay the millwrights OT to get it up and running asap. Same should go with IT. I don't mind putting the hours in, but I'll be damned before I'll burn my gas and spend my evening working for free.
Well Nixon also signed off on HMO's so you give some, you take some was a policy back then as well.
lol, I loved that movie as a kid. We still had a black and white TV at the time, with one of those push button channel switchers on it, nice.
Also, how important is your equipment? Assume the room were to disapear what would happen? If the answer is the department/company will be inoperable until the new equipment arrives and you are able to restore stuff from dumps then you both need some redundancy, and are going to want a proper fire suppression system. If you use a neutral gas suppression your going to need an airtight space for the servers (which will probably help with the noise too).
If your job is anything like 90% of IT professionals, you'll spend the minority of your time in your datacenter, that is what terminal services is for. Also, think about all the time spent sourcing vendors, doing project management work, any coding you have to do etc. I'd really fight to have the equipment put somewhere else if I was you. Figure out what percentage of time you need physical contact with the servers/switches and let your boss know. I mean it doesn't make sense to live with the noise and run at reduced efficiency for 80% of your work day for the sake of 300 sqft or so of floor space, if there is any way to avoid it. Also, if the sound proofing ends up sucking, you'll have to take phone calls and stuff with that crap in the background.
It is funny, companies worry about their 20k per year factory workers when it comes to noise, and environmental conditions, but the 60-150k IT guy gets the shaft. Apparently server noise has a different effect on the ears than machine noise.
50 years from now you won't care that you can't play your Crossroads DVD.
We are a start up and as such we can't pay you, but we promise once we are profitable we will. Unfortunately VC and bank capital aquisitions aren't going so well as we spend all our executives' time coming up with ingenious ways to recruit people. signed, dumb ass CEO anonymous Inc
I mean the average person won't know were to look in the system, or won't care about the number of MRE's being sent to Zimbabwai. For the most part isn't the real threat that information will get to the people that will know how to interpret it and care? In that case won't the foreign governments/terrorists just learn how to exploit Mac's rather than Windows? This seems a lot like security by obscurity, except, well Macs aren't that obscure.
This will become a requirement for grad students now?