I was going to suggest ipMonitor (which Solarwinds acquired a few years ago). They just changed the licensing so it's $1000-$2000 for an unlimited number of monitors. The system has a basic mapping engine, NOC environment, sound/email/pager/exe/etc alarms, complex SNMP, WMI, and SQL alarms for everything fmor basic up/down to user experience types of alarms, etc. I never had to look into distributed pollers and I don't think it supports that. The biggest con I ran into was that the reporting engine seemed slow and there was no way to get at the raw data to export it out to other tools.
Other pros include wizards (enter an IP to scan and it will suggest a list of monitors for that device by scanning for SNMP, WMI, etc), ability to create dependancy chains so that you would receive alerts from 5 devices when their shared switch to the backbone went down (you would just receive the ones for the switch), ability to create "Smart" groups which were basically dynamic groups that include all devices/monitors that meet a set of search criteria, etc.
Not sure how it scales, but we had this running against 100+ servers and network switches from a little virtual server w/ maybe 1GB of RAM and it didn't seem to be hurting. We also used Cacti for another perspective into traffic flows an such to give us another dimesnion to use when troubleshooting (monitoring + flows + logs).
You mean tech-high concentration like RTP? I don't know the numbers for RTP proper, but I've seen numbers for Raleigh that said overall average income of $57k and overall average house cost of $175k.
As others have already commented, Agile is one option but unless you want to be the guy in the corner pounding on screws with a hammer, I'd suggest reading a little about several methodologies and then dig deeper into one methodology when you have a situation that fits. Remember, the OP is in a single person environment, that alone is going to make daily team meetings a little challenging (or at least a little odd to watch). Other options include Critical Path, which can be executed with a waterfall model but could just as easily outline an iterative model or a TDD model. A methodology that I like is Critical Chain Project Management, which follows lean manufacturing principles rather than lean product development (like most Agile methods).
In any case, my comment to the OP would be to learn some basic project management topics from books then spend an afternoon on wikipedia learning the basics about a number of different methods. Then choose one to statr with and see how it works. Don't eb afraid to throw it away and choose a different one for your next project, they all build experience and breadth of knowledge.
Unfortunately, while I personally like the XMPP protocol and think it would normally be an excellent solution, I think you have uncovered the biggest flaw. Preventing the clients from talking to the outside world is going to be nearly impossible unless you keep them on a network that doesn't route to the outside world. For instance, GTalk uses SSL over port 443 so if you want any type of secure web transactions with the outside world then your also going to be allowing secure chatting. Even if you go through and block obvious XMPP hosts that are using non-standard ports (443, 80, etc) it will require ongoing attention as other sites start their own services.
I believe "logic" is the answer your looking for, not observable evidence.
If the world ends, then it no longer exists. If the world exists, then it could not have ended. Therefore the world can only end once because afterwards it will no longer exist to end a second time.
So to the original authors point, the world will continue not ending until the day it actually ends, at which point it won't be around to continue not ending any further (or to end a second time).
I believe the only observation that his/her logic rests on is the observation that the world has not ended yet, and therefore the day it does end has not occurred yet.
If you have enough to purchase $300 worth of books for classes. Then, yes, this means you also have $300 to buy some of the software you need, right up until you spend that $300 on books. Once you buy the books than you have $0 left for software that you would have previously used in the computer lab. Heck, some students don't have enough for the books and instead will borrow them from the library so they don't have to give up eating for a couple weeks to keep up with classes.
And I disagree with the AC below, it is about the money. Especially considering that you will at least be able to take the books with you at the end of your college career or trade them in at a discount to help pay for the next semester, the discounted software cannot be resold and generally comes with a clause requiring you to stop using it when you end your college career.
Laptops are not as handy for reading as paper (thus the market for e-book readers) and scribbling in the margins is still going to be just a bit easier and natural then a comparable method on a laptop (at least for anyone who has been grading papers their entire career).
The student has to handle one paper, the professor or teacher has to handle 20-40 papers x number of classes and they are likely going to sit in several different places while trying to grade all of them.
I used to do my taxes myself every year, then along came websites that can do it for me (I'm lazy).
This is what I see: We have a treasury secretary who found the tax code confusing and, instead of seeking assistance (like a paid professional) he dumped the whole thing in a dark corner somewhere and refused to face it.
Between the tactical error of dumping into the dark corner and the inability to either request assistance or improve his knowledge I don't think I'd even want to hire him to manage a local fast food restaurant because these types of acitons tend to be even stronger in the workplace...
This full admin lazy installation thing drives me nuts.
I've worked in IT for a while and I am always severely annoyed when people automatically assume that receiving a permission denied dialod or two means that an application requires full administrative rights to the system.
They will generally go on about how life would have been easier if only that one application had let them do things the right way when, in reality, had that one application provided better notes (or had someone had called the vendor) on the subdirectories, external DLLs, or registry entries that needed permissions, then it would have been the next program that came along that required administrative rights. Of course the next admin would have come along and faced with an absolute lack of documentation from the previous admin, would have ended up giving anyone admin access all over again.
Many companies are taking advantage of better packing (euither by changing their packaging, tweaking their loading method, etc) as a method to cut down on costs. The ability to cram two extra boxes on a truck can often mean selling two extra boxes that you originally couldn't or reducing the shipping costs per order by reducing the amount of leftover head space in a shipping container by stacking more efficiently.
I have gotten used to $4-$6 DRM-free eBooks. My wife got me a Kindle and I started looking through Amazon prices...while I will likely buy a book or two from them, I think the majority of my books will still come from other places (Glad it has a USB cable). I mean come on, $9-$10 for books that are $7 for the paperback?
Having worked in a large position in a large company does not make you more qualified to work in a small company. If anything, the reverse may be true. Small companies have a much smaller margin for error and a much larger reaction to the individual's skill level or abilities. Additionally the role someone has in a larger company is generally narrower than the same role in a smaller company. Someone that does very well at a specialized subset of skills in a large company may end up completely mediocre in a small company that requires the full range of skills.
Same in NC. Heck, the only time I even have to go in the DMV anymore is for new things and license renewals. Duplicate licenses, registration fees, etc are all online.
Yes, they really should have just lobbed hundreds of missiles randomly over the border and been done with it. That would have been much less indiscriminate.
And so far their "wanton indiscriminate killing" seems to have gotten pretty lucky, in that more Hamas militia members have been killed then civilians...either that indicates a much larger Hamas force in the area (ie, high 'militia' to civilian ratio) than Hamas, the media, or Israel has indicated (and also that Hamas for once is not claiming any number of it's members are civilians) or it's just pure dumb luck that the "wanton indiscriminate killing" only penetrated far enough into the country to see 500 people and a high number of those were militia.
Because that cuts the margin. Gamers stress machines, so they would have to choose to either continue loading in the extremely crappy motherboard and RAM they use and deal with the complaints as the system keeled over in no time, or they would have to upgrade to decent motherboard and RAM combinations reducing their margin even further. Add to that the fact that 3 years of gaming on mid-range equipment is going to have a much more drastic effect than 3 years of MS office on a low range and your talking higher costs in the warranty and support arena too...
Not at all. I know plenty of people at my current workplace who couldn't even name the equipment on our production floor if they saw it (we have only have 3-4 steps in our main production process). Not too long ago I overheard a coworker ask someone to draw them a map to get to a cabinet on the opposite side of the production floor because they couldn't follow the verbal instructions (we only have 3 main areas, and they are rectangular so "middle of far wall of X area" should have been enough).
In a manufacturing plant your operations staff, R&D, Quality, and logistics staff generally have a good idea (some better than others) how things work. Maintenance and engineering also have an idea, but it's a little more removed (they allow "you should be doing it this way" to color their perspective). IT is the only group (unless their is also a Plant Automation group and Process Improvement group) that has to understand a good portion of all of these perspectives (at least, a good IT group). And most of your finance, sales, marketing, and HR departments will not know, though one or two members may as part of their job description and several others may have some knowledge due to curiosity.
The same is true in many businesses (though the functions of operations, quality, logistics, etc may be different in places such as a software publisher, etc).
However I would not say this is the function of a "Software Engineering". Outside of IT groups the major role that has to have this understanding is Software Architects/Designers. Your standard "Software Engineer" at a software development company has to have some understanding of the architectural vision or design and an idea what they are trying to accomplish, but that is where it generally ends. They do not work in the environment they are building software for and do not spend the same level of time supporting and learning the process they are trying to improve.
Looking at it from a different point of view, there are also degrees for Biology, Liberal Arts, etc, but I was still required to take a number of classes outside of CS in order to be "well-rounded". I would go so far as to say that a business class requirement would not be a bad replacement for one of the many electives that are required today. That and the couple I took were disturbingly easy:)
If the login is myname@yourautherver.com then you now know what system to send an authentication request to and the username for the account. Whether you use DNS records as the original poster is suggesting to create something akin to an MX record or not, all it takes now is a standard URL or socket and protocol to authenticate the given user against the server they signed up on. If it happens that DNS is out the window for some reason and you want to subdomain the authentication server, then just hand out id's that look like yourname@auth.yourserver.com
And the addition of DNS records is not something we require every potential end user to do, it's something we do for the authentication server. If you have been creating yuor own MX records for every person you intend to send an email to, you can probably take a break for a while...
On the other hand, the number of game physics positions out there is drastically overshadowed by the number of business-related development ones. Would it not make more sense to take additional business classes to better understand the audience, target user group, and purpose behind the a business asking for development? What makes Calculus more useful to a CS grad then, say, advanced Bio classes? Does one set better support an understanding of CS than another?
Note: This is not necessarily my own viewpoint, just playing devil's advocate:)
It's homogenized Walmart-ware that doesn't like you to go off the beaten path. That's both a good thing and a bad thing.
That's actually been one of my main problems with MS tech all along. Every application, language, technology, whatever that I have used has had this magic bar. Doing anything under the magic bar is simpler, on average, then it's competitors. However once you try to exceed the magic bar you end up in a position likened to being kicked in the head, repeatedly, by steel toe boots. The first 99% of what you were working on took a few days, the one function/feature that isn't working the way your expecting it to will actually end up costing you twice as long to get past.
Oh, and not using the click-and-drag method will actually raise the bar a bit. I've found that if you forgo those methods early you will manage to get a lot farther before the tool tries to force bad practices on you or leaves you ni a dead end trying to figure out how to make a function work the way it should.
Offtopic: When did the quote button start spitting out extra, incorrect HTML as part of it's duties?
To back up the GPP on flakiness: All of the suites I have worked with have at least one or two flaky components. However this does not mean all of their components were flaky, I just remember the flaky ones more. Listed are the two I have spent the most time with:
Wonderware: I'd love to know why Alarm Manager, DT Analyst, certain IAS components, and some of the DI objects flake out. I cannot even find a log entry when these things fail, but eventually we'll bite the bullet and get someone in to take a good hard look and fix them as best as possible (or upgrade them to newer versions). InSQL 8 had some flakiness early on and the interface is still a little cruddy in 9.
OsiSoft: It's been a while but the flakiest app I used with them was ACE. There was also an issue in several other apps that Osi didn't know about that we tracked down to a particular version of ProcessBook - it would overwrite a critical registry key and leave it without read/write permissions, which would kill a multitude of other apps. Overall I recall less flakiness with their products than WWs.
My issue is not so much the occasional flaky app (especially since these are usually early versions) but just the general lag between process software and the rest of the world. But the reason behind this (and the part of the cost) is that it takes a great deal of time to make majors changes to some of these apps. I get frustrated with these suites at time also, then I compare the issues of a v2 product with v20 of a normal enterprise app and realize that while I may have an app flaking out occasionally, the v20 enterprisy app still has serious design and functional issues after 20 major revisions. Yeah the interface may be dated, the APIs may suck a bit, but it still beats the hell out of most enterprise software out there.
The biggest issue I have seen across the board with industrial suites is not the core software, but what happens to the core software when you let lose certain types of engineers on them with development tools. The absolute biggest problems I have seen with all of these packages have been user inflicted.
Adding a dash of water helps release the flavor of a whiskey. And I don't know of any good ones in the $25 range, my range is more the $40-$90 (Aberlour, Oban, The Macallan, Lagavulin, Dalwhinnie, The Balvenie, Laphroiag, etc). In the $25 dollar range you might even find me adding ice...
Slashdot Mirror
I was going to suggest ipMonitor (which Solarwinds acquired a few years ago). They just changed the licensing so it's $1000-$2000 for an unlimited number of monitors. The system has a basic mapping engine, NOC environment, sound/email/pager/exe/etc alarms, complex SNMP, WMI, and SQL alarms for everything fmor basic up/down to user experience types of alarms, etc. I never had to look into distributed pollers and I don't think it supports that. The biggest con I ran into was that the reporting engine seemed slow and there was no way to get at the raw data to export it out to other tools.
Other pros include wizards (enter an IP to scan and it will suggest a list of monitors for that device by scanning for SNMP, WMI, etc), ability to create dependancy chains so that you would receive alerts from 5 devices when their shared switch to the backbone went down (you would just receive the ones for the switch), ability to create "Smart" groups which were basically dynamic groups that include all devices/monitors that meet a set of search criteria, etc.
Not sure how it scales, but we had this running against 100+ servers and network switches from a little virtual server w/ maybe 1GB of RAM and it didn't seem to be hurting. We also used Cacti for another perspective into traffic flows an such to give us another dimesnion to use when troubleshooting (monitoring + flows + logs).
You mean tech-high concentration like RTP? I don't know the numbers for RTP proper, but I've seen numbers for Raleigh that said overall average income of $57k and overall average house cost of $175k.
As others have already commented, Agile is one option but unless you want to be the guy in the corner pounding on screws with a hammer, I'd suggest reading a little about several methodologies and then dig deeper into one methodology when you have a situation that fits.
Remember, the OP is in a single person environment, that alone is going to make daily team meetings a little challenging (or at least a little odd to watch).
Other options include Critical Path, which can be executed with a waterfall model but could just as easily outline an iterative model or a TDD model.
A methodology that I like is Critical Chain Project Management, which follows lean manufacturing principles rather than lean product development (like most Agile methods).
In any case, my comment to the OP would be to learn some basic project management topics from books then spend an afternoon on wikipedia learning the basics about a number of different methods. Then choose one to statr with and see how it works. Don't eb afraid to throw it away and choose a different one for your next project, they all build experience and breadth of knowledge.
Unfortunately, while I personally like the XMPP protocol and think it would normally be an excellent solution, I think you have uncovered the biggest flaw. Preventing the clients from talking to the outside world is going to be nearly impossible unless you keep them on a network that doesn't route to the outside world.
For instance, GTalk uses SSL over port 443 so if you want any type of secure web transactions with the outside world then your also going to be allowing secure chatting. Even if you go through and block obvious XMPP hosts that are using non-standard ports (443, 80, etc) it will require ongoing attention as other sites start their own services.
I believe "logic" is the answer your looking for, not observable evidence.
If the world ends, then it no longer exists.
If the world exists, then it could not have ended.
Therefore the world can only end once because afterwards it will no longer exist to end a second time.
So to the original authors point, the world will continue not ending until the day it actually ends, at which point it won't be around to continue not ending any further (or to end a second time).
I believe the only observation that his/her logic rests on is the observation that the world has not ended yet, and therefore the day it does end has not occurred yet.
Hmm, you may need some more practice at math...
If you have enough to purchase $300 worth of books for classes. Then, yes, this means you also have $300 to buy some of the software you need, right up until you spend that $300 on books. Once you buy the books than you have $0 left for software that you would have previously used in the computer lab. Heck, some students don't have enough for the books and instead will borrow them from the library so they don't have to give up eating for a couple weeks to keep up with classes.
And I disagree with the AC below, it is about the money. Especially considering that you will at least be able to take the books with you at the end of your college career or trade them in at a discount to help pay for the next semester, the discounted software cannot be resold and generally comes with a clause requiring you to stop using it when you end your college career.
Laptops are not as handy for reading as paper (thus the market for e-book readers) and scribbling in the margins is still going to be just a bit easier and natural then a comparable method on a laptop (at least for anyone who has been grading papers their entire career).
The student has to handle one paper, the professor or teacher has to handle 20-40 papers x number of classes and they are likely going to sit in several different places while trying to grade all of them.
There's a difference between refund and redistribution...but either way someone will be relieved
I used to do my taxes myself every year, then along came websites that can do it for me (I'm lazy).
This is what I see:
We have a treasury secretary who found the tax code confusing and, instead of seeking assistance (like a paid professional) he dumped the whole thing in a dark corner somewhere and refused to face it.
Between the tactical error of dumping into the dark corner and the inability to either request assistance or improve his knowledge I don't think I'd even want to hire him to manage a local fast food restaurant because these types of acitons tend to be even stronger in the workplace...
This full admin lazy installation thing drives me nuts.
I've worked in IT for a while and I am always severely annoyed when people automatically assume that receiving a permission denied dialod or two means that an application requires full administrative rights to the system.
They will generally go on about how life would have been easier if only that one application had let them do things the right way when, in reality, had that one application provided better notes (or had someone had called the vendor) on the subdirectories, external DLLs, or registry entries that needed permissions, then it would have been the next program that came along that required administrative rights. Of course the next admin would have come along and faced with an absolute lack of documentation from the previous admin, would have ended up giving anyone admin access all over again.
Many companies are taking advantage of better packing (euither by changing their packaging, tweaking their loading method, etc) as a method to cut down on costs. The ability to cram two extra boxes on a truck can often mean selling two extra boxes that you originally couldn't or reducing the shipping costs per order by reducing the amount of leftover head space in a shipping container by stacking more efficiently.
Florida is a banana or it wouldn't smell like bananas.
Oh sorry, i was trying out this new logic of yours...
Let's clarify, use DRM and pay more for ebooks.
I have gotten used to $4-$6 DRM-free eBooks. My wife got me a Kindle and I started looking through Amazon prices...while I will likely buy a book or two from them, I think the majority of my books will still come from other places (Glad it has a USB cable).
I mean come on, $9-$10 for books that are $7 for the paperback?
Having worked in a large position in a large company does not make you more qualified to work in a small company. If anything, the reverse may be true.
Small companies have a much smaller margin for error and a much larger reaction to the individual's skill level or abilities. Additionally the role someone has in a larger company is generally narrower than the same role in a smaller company. Someone that does very well at a specialized subset of skills in a large company may end up completely mediocre in a small company that requires the full range of skills.
If only I had mod points and the ability to simultaneously post a +1 Insightful and +1 Owned :)
Same in NC. Heck, the only time I even have to go in the DMV anymore is for new things and license renewals. Duplicate licenses, registration fees, etc are all online.
Yes, they really should have just lobbed hundreds of missiles randomly over the border and been done with it. That would have been much less indiscriminate.
And so far their "wanton indiscriminate killing" seems to have gotten pretty lucky, in that more Hamas militia members have been killed then civilians...either that indicates a much larger Hamas force in the area (ie, high 'militia' to civilian ratio) than Hamas, the media, or Israel has indicated (and also that Hamas for once is not claiming any number of it's members are civilians) or it's just pure dumb luck that the "wanton indiscriminate killing" only penetrated far enough into the country to see 500 people and a high number of those were militia.
Because that cuts the margin. Gamers stress machines, so they would have to choose to either continue loading in the extremely crappy motherboard and RAM they use and deal with the complaints as the system keeled over in no time, or they would have to upgrade to decent motherboard and RAM combinations reducing their margin even further. Add to that the fact that 3 years of gaming on mid-range equipment is going to have a much more drastic effect than 3 years of MS office on a low range and your talking higher costs in the warranty and support arena too...
Not at all. I know plenty of people at my current workplace who couldn't even name the equipment on our production floor if they saw it (we have only have 3-4 steps in our main production process). Not too long ago I overheard a coworker ask someone to draw them a map to get to a cabinet on the opposite side of the production floor because they couldn't follow the verbal instructions (we only have 3 main areas, and they are rectangular so "middle of far wall of X area" should have been enough).
In a manufacturing plant your operations staff, R&D, Quality, and logistics staff generally have a good idea (some better than others) how things work. Maintenance and engineering also have an idea, but it's a little more removed (they allow "you should be doing it this way" to color their perspective). IT is the only group (unless their is also a Plant Automation group and Process Improvement group) that has to understand a good portion of all of these perspectives (at least, a good IT group). And most of your finance, sales, marketing, and HR departments will not know, though one or two members may as part of their job description and several others may have some knowledge due to curiosity.
The same is true in many businesses (though the functions of operations, quality, logistics, etc may be different in places such as a software publisher, etc).
However I would not say this is the function of a "Software Engineering". Outside of IT groups the major role that has to have this understanding is Software Architects/Designers. Your standard "Software Engineer" at a software development company has to have some understanding of the architectural vision or design and an idea what they are trying to accomplish, but that is where it generally ends. They do not work in the environment they are building software for and do not spend the same level of time supporting and learning the process they are trying to improve.
Looking at it from a different point of view, there are also degrees for Biology, Liberal Arts, etc, but I was still required to take a number of classes outside of CS in order to be "well-rounded". I would go so far as to say that a business class requirement would not be a bad replacement for one of the many electives that are required today. :)
That and the couple I took were disturbingly easy
Fail.
If the login is myname@yourautherver.com then you now know what system to send an authentication request to and the username for the account. Whether you use DNS records as the original poster is suggesting to create something akin to an MX record or not, all it takes now is a standard URL or socket and protocol to authenticate the given user against the server they signed up on. If it happens that DNS is out the window for some reason and you want to subdomain the authentication server, then just hand out id's that look like yourname@auth.yourserver.com
And the addition of DNS records is not something we require every potential end user to do, it's something we do for the authentication server. If you have been creating yuor own MX records for every person you intend to send an email to, you can probably take a break for a while...
On the other hand, the number of game physics positions out there is drastically overshadowed by the number of business-related development ones. Would it not make more sense to take additional business classes to better understand the audience, target user group, and purpose behind the a business asking for development?
What makes Calculus more useful to a CS grad then, say, advanced Bio classes? Does one set better support an understanding of CS than another?
Note: This is not necessarily my own viewpoint, just playing devil's advocate :)
That's actually been one of my main problems with MS tech all along. Every application, language, technology, whatever that I have used has had this magic bar. Doing anything under the magic bar is simpler, on average, then it's competitors. However once you try to exceed the magic bar you end up in a position likened to being kicked in the head, repeatedly, by steel toe boots. The first 99% of what you were working on took a few days, the one function/feature that isn't working the way your expecting it to will actually end up costing you twice as long to get past.
Oh, and not using the click-and-drag method will actually raise the bar a bit. I've found that if you forgo those methods early you will manage to get a lot farther before the tool tries to force bad practices on you or leaves you ni a dead end trying to figure out how to make a function work the way it should.
Offtopic: When did the quote button start spitting out extra, incorrect HTML as part of it's duties?
To back up the GPP on flakiness: All of the suites I have worked with have at least one or two flaky components. However this does not mean all of their components were flaky, I just remember the flaky ones more. Listed are the two I have spent the most time with:
Wonderware: I'd love to know why Alarm Manager, DT Analyst, certain IAS components, and some of the DI objects flake out. I cannot even find a log entry when these things fail, but eventually we'll bite the bullet and get someone in to take a good hard look and fix them as best as possible (or upgrade them to newer versions). InSQL 8 had some flakiness early on and the interface is still a little cruddy in 9.
OsiSoft: It's been a while but the flakiest app I used with them was ACE. There was also an issue in several other apps that Osi didn't know about that we tracked down to a particular version of ProcessBook - it would overwrite a critical registry key and leave it without read/write permissions, which would kill a multitude of other apps. Overall I recall less flakiness with their products than WWs.
My issue is not so much the occasional flaky app (especially since these are usually early versions) but just the general lag between process software and the rest of the world. But the reason behind this (and the part of the cost) is that it takes a great deal of time to make majors changes to some of these apps. I get frustrated with these suites at time also, then I compare the issues of a v2 product with v20 of a normal enterprise app and realize that while I may have an app flaking out occasionally, the v20 enterprisy app still has serious design and functional issues after 20 major revisions. Yeah the interface may be dated, the APIs may suck a bit, but it still beats the hell out of most enterprise software out there.
The biggest issue I have seen across the board with industrial suites is not the core software, but what happens to the core software when you let lose certain types of engineers on them with development tools. The absolute biggest problems I have seen with all of these packages have been user inflicted.
Adding a dash of water helps release the flavor of a whiskey. And I don't know of any good ones in the $25 range, my range is more the $40-$90 (Aberlour, Oban, The Macallan, Lagavulin, Dalwhinnie, The Balvenie, Laphroiag, etc). In the $25 dollar range you might even find me adding ice...