Ok, I would be careful in jumping ship to move to nursing. Just like the late 90s did for IT, every tom, dick, and harry school is starting a nursing program. I predict a nursing glut and a reduction in salary for nurses in the next 5 or so years.
I don't know about every Virtualization push, but here we are doing it because we are getting blades that at minimum spec are so much overkill for the application we run on them. Since we like to partition out at least one server per app since app vendors like to point fingers at each other and it is just cleaner this way, virtualization is an optimal method for us.
A MBA is a ticket punch into management. You don't need one, but it does set you apart, showing that you can set down a long term goal and complete it. It should give you better writing, negotiating, and analytical skills. With COBIT becomming bigger with our industry, these skill sets will help in the long run. Of course, you could be the 60 year old programmer working a government job talking about the good old days of fortran.
Conversely, we strip the machine of useful information and then we rebuild it. My users do a bad job at not storing all of their important data on a drive.
The issue with the wikipedia becomes trying to control who can review which category. You would have to set review committees for particular subject. I do agree that a web-based solution that allows for posting of journal article. I hazard to think that you would find that much of the time comes from the peer review process. I wouldn't want articles posted until reviewed. It could work, but I think you would be surprised on how long the review process can be. However, I think it would reduce the publish time by a year. There are a lot of papers submitted that don't make press.
On revisions, I think that the posted paper, once reviewed and edited, should not allow revisioning. It is my fundamental (spelling nazi and it is a logic error to think I was not using Firefox just because I didn't check for spelling errors) belief that some researchers would try to abuse that feature, especially since I can think of a few of my compatriots who would. When papers are 'done', then they should be immutable.
Hard part is seperating the wheat from the chaft. That is what scientific journals are for, since readers can't just determine the validity on their own when the amount of writings are considered. Type 'how much time nurses have to review evidence based research per week and you get a ton of results that do not apply' and you get a whole bunch of results that don't answer the question. (We are statisticly certian that it is less then an hour per week. You heard it on slashdot first)
The fundemental question is are you going to believe JAMA with a peer reviewed article or what Joe Blow posted on the net. Sometimes the process is a shame. One of the most useful papers for my master's thesis never made it to press due to an interesting confluence of events. The problem is that the process will be horribly slow just to do the peer-review and expensive due to the cost of reviews, editors, and support staff. Then trying to get the article in a paper takes a while because the editors are trying to get the issues to flow somewhat. How would you improve a process while maintaining the quality?
I would agree with you. I am surprised that HIMSS hasn't raised unholy hell over this. I think that I am safe to recommend that users do not trust this service and not use it. I think that an electronic way of moving data between places is a good thing, but with a requirement that every data move transaction requires patent consent. Without patent consent, which is the best standard of how we should do things in health care, then data should not be shared, with a check for consent per instance of use. In addition, data access should be logged for 20 years, with unauthorized access with a hefty fine for a Google type corporation per instance with a 100 times multiplier for not disclosing immediately upon discover.
In other words, these centralized storage concepts should have a significant risk for any venture.
It is reasonable to cut access of a high permissive security user until they leave. That way, they have to work through anyone to do any work. It improves training and helps limit anything that can happen in the transition period. It is a good risk mitigation technique.
I always had a problem with a company with the value statement of 'Do no evil' who doesn't spell out what that means in detail. I was listening to Stafford's Entrepreneurial Thought Leaders series this weekend and Google.org was discussing using their engineering talent to recognize epidemics before anyone else. My guess is this is how Google plans to do it. It is clear Google intends to use this data, but I think has done a poor job defining exactly how. Add in the fact that Google has bowed to governments for information on their citizens and I end up with a cold chill. Working in the health care industry, I see the value of patient records that are easy to transfer for the patient, but I am not sure this is the way. The little security analyst in me is screaming bloody murder.
What is the percentage of the US population under 5 years old? 6.5% according to census.gov 2007 estimates
What is the percentage of the US population over 85 years old? 2% according to census.gov 2007 estimates
It is reasonable that 7.5% could not have sent emails due to age situations. (Not saying that 7.5% is exact but they would likely trend statistically to that fact) I have a 7 year old neice who hasn't sent an email. Conversely, I also have a 8 year old nephew that has sent a bunch. Different parenting styles and social-economic factors. My grandfather, who died at 82, didn't send an email in his life that anyone knows about. Considering that 26.5% of our population fall outside of the 10 to 65 year range, it is easy to believe that there there are 20% of the population that doesn't use email. I think it is a mixture of ludditism, social-cultural issues, and lack of need and interest (which is different from the other two factors). Consider the homeless, considered 1% , which probably have a low incident of email usage in their life.
As an academian, which rhymes with comedian (think about it), that has played that scene, I am appalled by..... MADD. While I will never drive when drunk due to the death of a couple of friends right out of high school, I have played many games with a glass of glenmorangie sherry cask, including driving games. I have always worried about some idiot who thinks they can drive PGR while drunk translates to being able to operate a real car. The trying to drive drunk in GTA is very hard and made me think, 'there is NO way I will get behind the wheel drunk.' I think people need to get over themselves and understand the game designer's vision when making that scene.
Or option c. Pay a ledger tech $12 an hour for 3 hours worth of work every month. The distribution groups change often and there are several different types of report packages.
On the ERP part, IBM blames Infor. Infor blames IBM. Since we can't repeat it at will, we are stuck in the middle. I do know that the attachment is good before it is sent to the SMTP gateway. Had a great comment from Infor's support thought, "You use Notes?" (Imagine the sound of incredulaity)
On the PDFs, we don't have other applications sending PDFs through the SMTP gateway. Our solution is definitely not ellegant. We produce the PDFs and manually email them out. It drives me nuts, but it is either that or purchase an expense amount of licensing for my DBMS and BI platform. (I did the cost analysis, it is still cheaper to make a ledger tech do the work)
Now, you want to hear some gripes, ask me about Sybase ASE's query engine.
Naw, he is actually really good. Notes is a Beast no matter what. We have something not work in version a, fixed in b, and broken again in c. It has been a huge issue for us. Attachments from our ERP system's notification email process corrupt at random intervals on random individuals, even though in the working directory they are fine. We can't even use the SMTP gateway for our BI platform, since our Domino server doesn't like PDFs sent in via SMTP. It generates errors if we use a dummy return address. Let alone the notes client decided today just not to run on my workstation.
To qoute my Domino Administrator: Lotus Stinking Notes
He asked about when we were getting off of it, to the CIO. No, he is not an 'exchange' guy, but he figures that anything would be better then Domino. I did mention Groupwise, which he had to conceed on that point.
Honestly, you have now become the Principle Investigator's Bitch.
That said, it depends on the lab. Back in my chemical engineering days, the PI was wonderful, but the grad students sucked. A friend from high school lamented the lack of ethics of his PI due to him doing all of the work, obtaining funding, coming up with the idea, etc... and the PI signing his name to the final product and taking all of the credit. My friend is now a PI in his own respect. My current Research Advisor, who acts as a PI, works with us on our projects, acting more as a facilitator.
(a) All documentaries are propaganda. They are designed to present the film maker's view point of a situation. Objectivity does not exist in such works, since their purpose is to influence peoples thoughts.
(b) The major premise of Expelled is that scientific debate is squashed due to view point elitism. (I actually watched the movie) It does not matter what the debate is about, but when the stance is that 'you cannot oppose us', then there is a lot of problems. We see this in a lot of fields. Climate Change is a good example, but so is Health Care Management, Health Care Economics, and a plethora of many other fields. You can try this in the classroom. Challenge a professor on a topic that they are ideologically strong in, such as I did. I asked a Sociology professor,who believed everything could be described in the Marxist Dialect, to describe the fall of the Berlin Wall in the Marxist Dialect. Her response was to scream at me and call me names. No discussion or debate, just hostile action. When I teach, I encourage debate and learning in the classroom. However, many of my fellow professors do not.
As a side point, I would recommend that you do not make assumptions on where people's viewpoints come from. It is a logical fallicy to assume that opinions that disagree with yours to be from the Bible. It is called a strawman arguement and is an invalid line of reasoning.
Hell, lets resolve this like they did back then. Give me an unit of marines, a naval squadron, and three times as many mercenaries. I will just shoot the hackers. Sing the song be damed, we'll just shoot them in the head.
What is required for a decision analysis approach, which is what a make vs buy comes down to, might be lacking from the original post (op). However, I think this just leads to a requirement for more research and work into determining a decision. I find the op to be very lacking in detail, leading to a suggestion that they tread carefully. I know the requirement for privacy and such, but consider this example:
We have a main system and BI system provided by the same vendor. To support the data load from one system to another, we created an intermediary data store, using MS SQL Server 2000 and DTS. The issue is that we have only one employee that understands developing an ETL transformation of this magnitude (80 - 100 GB range). The data store has been used also to off load the user reporting off of the main system's transactional databases. 8 other employees have been sent to training, with either no programming experience or a background in cobalt programming to MS SQL Server 05 integration services training. The only person interested in using integration services is a non-programmer looking to gap their system's issues. The cobalt programmers view integration services as not real programming and refuse to work with integration services. The initial programmer is interested in moving up in the company, pursuing further education and degrees.
There are three current solutions to choose from: (a) Use the vendor's canned transformation application, which will get the BI platform the needed data, but remove the extra data in data store beyond the needs of the BI platform. This will force all reports (120ish at last count with an expected time per report of 30 minutes and a cost of $25 per hour to convert) to be rewritten, increase system load on the main system, and reduce speed of both the main system and reporting. The cost of transformation application is approximately $15,000 (b) Let two cobalt programmers go, since we have no need for as many cobalt programmers as we have had in the past. This will allow for hiring of people more familiar with Integration Services and MS SQL Server in general, at a cost of $5,000 hiring costs, similar salary, and $2,500 in training and travel each. As always, the company could choose just to ignore the issue.
I have to take offense to this. Couple of years ago, the local paper's astrological peice listed for my birthday, 'If today is your birthday, you gonna get lucky today.' Now, yes I was dating the lady who was incharge of editing that section at the time; but by God, it was correct.
Actually, the best solution is to do a make vs. buy analysis. Make vs. buy analysis, being basic business school stuff, would be done on this problem as a comparison of how much it would cost to train up an expert, hire an expert, and various combinations of the two. Weighing time required to search vs. cost would also be useful. We did this when comparing the cost of keeping Sybase ASE or going to MS SQL Server when considering if our DBA left. Hiring a new Sybase ASE DBA costs as much as training another DBA up, but a search for a new DBA will take so much longer, due to the lack of Sybase ASE DBAs vs. DBAs expert in other systems that we could train. Conversely, it is easier to find a MS SQL Server DBA, since there is a larger pool of experienced candidates.
My organization sees these issues a lot due to the specialized clinical software we run. It is very expensive to train up staff on those systems. On the other hand, the pool of experianced candidates is very low. This causes an issue if we take an experianced candidate, not trained on the application and send them to training, because they have a habit of leaving one trained and having a minimum level of experiance. The way to mitigate this is to apply a Time in Service requirement to the training, but management feels that would be a disincentive to hire. Therefore, they look for the 'ideal' candidate, causing very long search times.
As for training a new candidate, we found the best solution is to make them responsible for some core piece. As I am in a system management/project management department, we usually make the new person the admin of a large system. This exposes them to a wide breath of how our implementation is setup. Ideally, we have a more senior system analyst act as a mentor, even though sometimes that doesn't work out due to staffing constraints.
Re:Maybe the NSA has to cut the cable to tap into
on
Third Undersea Cable Cut
·
· Score: 2, Interesting
According to a reliable source, it is plausable to hold a section of cable at a certain arc, shave the top layers of the cable off, and then apply the tap. It is also plausable for it to be done without any significant signal loss.
No, you put them on a cost plus plus pay scheme and pay them based off of project phasing. Then, since I work at a hospital, part them out after they are done.
Having self-updating financial packages is a really bad idea. Since companies tend to use software differently, with customized UDAKs and other fun policies, they need to test to make sure it will still work after the upgrade. Auto-updating is asking for trouble.
Slashdot Mirror
Ok, I would be careful in jumping ship to move to nursing. Just like the late 90s did for IT, every tom, dick, and harry school is starting a nursing program. I predict a nursing glut and a reduction in salary for nurses in the next 5 or so years.
I don't know about every Virtualization push, but here we are doing it because we are getting blades that at minimum spec are so much overkill for the application we run on them. Since we like to partition out at least one server per app since app vendors like to point fingers at each other and it is just cleaner this way, virtualization is an optimal method for us.
A MBA is a ticket punch into management. You don't need one, but it does set you apart, showing that you can set down a long term goal and complete it. It should give you better writing, negotiating, and analytical skills. With COBIT becomming bigger with our industry, these skill sets will help in the long run. Of course, you could be the 60 year old programmer working a government job talking about the good old days of fortran.
Conversely, we strip the machine of useful information and then we rebuild it. My users do a bad job at not storing all of their important data on a drive.
The issue with the wikipedia becomes trying to control who can review which category. You would have to set review committees for particular subject. I do agree that a web-based solution that allows for posting of journal article. I hazard to think that you would find that much of the time comes from the peer review process. I wouldn't want articles posted until reviewed. It could work, but I think you would be surprised on how long the review process can be. However, I think it would reduce the publish time by a year. There are a lot of papers submitted that don't make press.
On revisions, I think that the posted paper, once reviewed and edited, should not allow revisioning. It is my fundamental (spelling nazi and it is a logic error to think I was not using Firefox just because I didn't check for spelling errors) belief that some researchers would try to abuse that feature, especially since I can think of a few of my compatriots who would. When papers are 'done', then they should be immutable.
Hard part is seperating the wheat from the chaft. That is what scientific journals are for, since readers can't just determine the validity on their own when the amount of writings are considered. Type 'how much time nurses have to review evidence based research per week and you get a ton of results that do not apply' and you get a whole bunch of results that don't answer the question. (We are statisticly certian that it is less then an hour per week. You heard it on slashdot first)
The fundemental question is are you going to believe JAMA with a peer reviewed article or what Joe Blow posted on the net. Sometimes the process is a shame. One of the most useful papers for my master's thesis never made it to press due to an interesting confluence of events. The problem is that the process will be horribly slow just to do the peer-review and expensive due to the cost of reviews, editors, and support staff. Then trying to get the article in a paper takes a while because the editors are trying to get the issues to flow somewhat. How would you improve a process while maintaining the quality?
I would agree with you. I am surprised that HIMSS hasn't raised unholy hell over this. I think that I am safe to recommend that users do not trust this service and not use it. I think that an electronic way of moving data between places is a good thing, but with a requirement that every data move transaction requires patent consent. Without patent consent, which is the best standard of how we should do things in health care, then data should not be shared, with a check for consent per instance of use. In addition, data access should be logged for 20 years, with unauthorized access with a hefty fine for a Google type corporation per instance with a 100 times multiplier for not disclosing immediately upon discover.
In other words, these centralized storage concepts should have a significant risk for any venture.
It is reasonable to cut access of a high permissive security user until they leave. That way, they have to work through anyone to do any work. It improves training and helps limit anything that can happen in the transition period. It is a good risk mitigation technique.
I always had a problem with a company with the value statement of 'Do no evil' who doesn't spell out what that means in detail. I was listening to Stafford's Entrepreneurial Thought Leaders series this weekend and Google.org was discussing using their engineering talent to recognize epidemics before anyone else. My guess is this is how Google plans to do it. It is clear Google intends to use this data, but I think has done a poor job defining exactly how. Add in the fact that Google has bowed to governments for information on their citizens and I end up with a cold chill. Working in the health care industry, I see the value of patient records that are easy to transfer for the patient, but I am not sure this is the way. The little security analyst in me is screaming bloody murder.
What is the percentage of the US population under 5 years old? 6.5% according to census.gov 2007 estimates
What is the percentage of the US population over 85 years old? 2% according to census.gov 2007 estimates
It is reasonable that 7.5% could not have sent emails due to age situations. (Not saying that 7.5% is exact but they would likely trend statistically to that fact) I have a 7 year old neice who hasn't sent an email. Conversely, I also have a 8 year old nephew that has sent a bunch. Different parenting styles and social-economic factors. My grandfather, who died at 82, didn't send an email in his life that anyone knows about. Considering that 26.5% of our population fall outside of the 10 to 65 year range, it is easy to believe that there there are 20% of the population that doesn't use email. I think it is a mixture of ludditism, social-cultural issues, and lack of need and interest (which is different from the other two factors). Consider the homeless, considered 1% , which probably have a low incident of email usage in their life.
I love the sherry. The concept of a creamy scotch is just amazing.
As an academian, which rhymes with comedian (think about it), that has played that scene, I am appalled by..... MADD. While I will never drive when drunk due to the death of a couple of friends right out of high school, I have played many games with a glass of glenmorangie sherry cask, including driving games. I have always worried about some idiot who thinks they can drive PGR while drunk translates to being able to operate a real car. The trying to drive drunk in GTA is very hard and made me think, 'there is NO way I will get behind the wheel drunk.' I think people need to get over themselves and understand the game designer's vision when making that scene.
Or option c. Pay a ledger tech $12 an hour for 3 hours worth of work every month. The distribution groups change often and there are several different types of report packages.
On the ERP part, IBM blames Infor. Infor blames IBM. Since we can't repeat it at will, we are stuck in the middle. I do know that the attachment is good before it is sent to the SMTP gateway. Had a great comment from Infor's support thought, "You use Notes?" (Imagine the sound of incredulaity)
On the PDFs, we don't have other applications sending PDFs through the SMTP gateway. Our solution is definitely not ellegant. We produce the PDFs and manually email them out. It drives me nuts, but it is either that or purchase an expense amount of licensing for my DBMS and BI platform. (I did the cost analysis, it is still cheaper to make a ledger tech do the work)
Now, you want to hear some gripes, ask me about Sybase ASE's query engine.
Naw, he is actually really good. Notes is a Beast no matter what. We have something not work in version a, fixed in b, and broken again in c. It has been a huge issue for us. Attachments from our ERP system's notification email process corrupt at random intervals on random individuals, even though in the working directory they are fine. We can't even use the SMTP gateway for our BI platform, since our Domino server doesn't like PDFs sent in via SMTP. It generates errors if we use a dummy return address. Let alone the notes client decided today just not to run on my workstation.
To qoute my Domino Administrator: Lotus Stinking Notes
He asked about when we were getting off of it, to the CIO. No, he is not an 'exchange' guy, but he figures that anything would be better then Domino. I did mention Groupwise, which he had to conceed on that point.
Honestly, you have now become the Principle Investigator's Bitch.
That said, it depends on the lab. Back in my chemical engineering days, the PI was wonderful, but the grad students sucked. A friend from high school lamented the lack of ethics of his PI due to him doing all of the work, obtaining funding, coming up with the idea, etc... and the PI signing his name to the final product and taking all of the credit. My friend is now a PI in his own respect. My current Research Advisor, who acts as a PI, works with us on our projects, acting more as a facilitator.
Two points.
(a) All documentaries are propaganda. They are designed to present the film maker's view point of a situation. Objectivity does not exist in such works, since their purpose is to influence peoples thoughts.
(b) The major premise of Expelled is that scientific debate is squashed due to view point elitism. (I actually watched the movie) It does not matter what the debate is about, but when the stance is that 'you cannot oppose us', then there is a lot of problems. We see this in a lot of fields. Climate Change is a good example, but so is Health Care Management, Health Care Economics, and a plethora of many other fields. You can try this in the classroom. Challenge a professor on a topic that they are ideologically strong in, such as I did. I asked a Sociology professor ,who believed everything could be described in the Marxist Dialect, to describe the fall of the Berlin Wall in the Marxist Dialect. Her response was to scream at me and call me names. No discussion or debate, just hostile action. When I teach, I encourage debate and learning in the classroom. However, many of my fellow professors do not.
As a side point, I would recommend that you do not make assumptions on where people's viewpoints come from. It is a logical fallicy to assume that opinions that disagree with yours to be from the Bible. It is called a strawman arguement and is an invalid line of reasoning.
Hell, lets resolve this like they did back then. Give me an unit of marines, a naval squadron, and three times as many mercenaries. I will just shoot the hackers. Sing the song be damed, we'll just shoot them in the head.
What is required for a decision analysis approach, which is what a make vs buy comes down to, might be lacking from the original post (op). However, I think this just leads to a requirement for more research and work into determining a decision. I find the op to be very lacking in detail, leading to a suggestion that they tread carefully. I know the requirement for privacy and such, but consider this example:
We have a main system and BI system provided by the same vendor. To support the data load from one system to another, we created an intermediary data store, using MS SQL Server 2000 and DTS. The issue is that we have only one employee that understands developing an ETL transformation of this magnitude (80 - 100 GB range). The data store has been used also to off load the user reporting off of the main system's transactional databases. 8 other employees have been sent to training, with either no programming experience or a background in cobalt programming to MS SQL Server 05 integration services training. The only person interested in using integration services is a non-programmer looking to gap their system's issues. The cobalt programmers view integration services as not real programming and refuse to work with integration services. The initial programmer is interested in moving up in the company, pursuing further education and degrees.
There are three current solutions to choose from: (a) Use the vendor's canned transformation application, which will get the BI platform the needed data, but remove the extra data in data store beyond the needs of the BI platform. This will force all reports (120ish at last count with an expected time per report of 30 minutes and a cost of $25 per hour to convert) to be rewritten, increase system load on the main system, and reduce speed of both the main system and reporting. The cost of transformation application is approximately $15,000 (b) Let two cobalt programmers go, since we have no need for as many cobalt programmers as we have had in the past. This will allow for hiring of people more familiar with Integration Services and MS SQL Server in general, at a cost of $5,000 hiring costs, similar salary, and $2,500 in training and travel each. As always, the company could choose just to ignore the issue.
Which choice should we make?
I have to take offense to this. Couple of years ago, the local paper's astrological peice listed for my birthday, 'If today is your birthday, you gonna get lucky today.' Now, yes I was dating the lady who was incharge of editing that section at the time; but by God, it was correct.
Actually, the best solution is to do a make vs. buy analysis. Make vs. buy analysis, being basic business school stuff, would be done on this problem as a comparison of how much it would cost to train up an expert, hire an expert, and various combinations of the two. Weighing time required to search vs. cost would also be useful. We did this when comparing the cost of keeping Sybase ASE or going to MS SQL Server when considering if our DBA left. Hiring a new Sybase ASE DBA costs as much as training another DBA up, but a search for a new DBA will take so much longer, due to the lack of Sybase ASE DBAs vs. DBAs expert in other systems that we could train. Conversely, it is easier to find a MS SQL Server DBA, since there is a larger pool of experienced candidates.
My organization sees these issues a lot due to the specialized clinical software we run. It is very expensive to train up staff on those systems. On the other hand, the pool of experianced candidates is very low. This causes an issue if we take an experianced candidate, not trained on the application and send them to training, because they have a habit of leaving one trained and having a minimum level of experiance. The way to mitigate this is to apply a Time in Service requirement to the training, but management feels that would be a disincentive to hire. Therefore, they look for the 'ideal' candidate, causing very long search times.
As for training a new candidate, we found the best solution is to make them responsible for some core piece. As I am in a system management/project management department, we usually make the new person the admin of a large system. This exposes them to a wide breath of how our implementation is setup. Ideally, we have a more senior system analyst act as a mentor, even though sometimes that doesn't work out due to staffing constraints.
According to a reliable source, it is plausable to hold a section of cable at a certain arc, shave the top layers of the cable off, and then apply the tap. It is also plausable for it to be done without any significant signal loss.
No, you put them on a cost plus plus pay scheme and pay them based off of project phasing. Then, since I work at a hospital, part them out after they are done.
Having self-updating financial packages is a really bad idea. Since companies tend to use software differently, with customized UDAKs and other fun policies, they need to test to make sure it will still work after the upgrade. Auto-updating is asking for trouble.