One of the choices people are saying would've been better is removing the requirement that it be for California businesses only, but by doing that they would have had a huge number of California residents who sold a business anywhere filing amended returns as far back as possible to get the no-longer-limited credits.
If it has a decent dual-mode color/b&w screen readable in direct light without backlighting, I'll shop at Wal-Mart for the first time in years, because this will immediately become an eInk killer.
Sure she'd like to get the money, but there's also the issue that if the company is going to book revenue in a particular way they have to deal with ALL of the things that go along with that. Companies paying commission don't get to say on one side "this was a great sale and we're going to compensate our executives on a great sales year" and on the other "that's an ok lease agreement you got, too bad commission on those is so low."
That's worse than the old Dilbert where the secretary "neglected" to put anything between the announcements of miserable numbers and an increased United Way push.
For email, it's actually really simple. What he sees in email headers (From, Subject, etc.) is the equivalent of the return address written in the top left corner of an envelope. There's absolutely nothing keeping you from putting false information there, and if he doesn't believe you ask him when's the last time he had to present identification to send a letter. What you're showing him instead is kind of like inspecting the cancellation mark on the stamp to determine that while the return address may say the White House, the letter was actually mailed from Portland, Oregon.
To give him an impression of the need to update, there are a few things to point out, and hopefully at least one will get through.
* First, among the most dangerous sites on the web these days are church websites - they're created as a volunteer effort by someone who may not even still be with the church (or who graduated HS and moved on in life). They're unmaintained. If they're infected, it may be a long time before someone even notices. In contrast, the "skeevy" sites like porn have a financial incentive to make sure their sites are safe.
* Second, once upon a time malware was written by spotty-faced geeks competing with each other for reputation. Those days are gone and have been gone for 20 years. These days malware is written by professional virus authors who do it for a living.
* Finally, show him the picture from http://www.deependresearch.org/2012/11/common-exploit-kits-2012-poster.html which shows a bunch of *commercially available* malware kits used to create new viruses and some of the security holes they target.
Another constraint that she may be considering is power efficiency - there may be tradeoffs between minimal wait time, etc. and power consumption, and the electricity cost for lifting an elevator car many times a day is not going to be negligible.
So far they've only done a Phase 1 trial which is to prove that it's not harmful, and the researchers called it "hideously expensive" at $1 million for 10 patients. If it shows clinical promise in Phase 2 and beyond, that price is likely to drop quite a bit and quite frankly the available MS treatments are also very expensive - if a single treatment is $100,000 but works for 5+ years, it may still be cheaper than what's currently available.
More information in an NBC News article: http://vitals.nbcnews.com/_news/2012/11/18/15246299-new-approach-could-treat-ms-other-autoimmune-diseases?lite
And the original article (for those willing to cough up $32 for a single article or with a subscription to the Nature Biotechnology journal): http://www.nature.com/nbt/journal/vaop/ncurrent/full/nbt.2434.html
I've settled on these as a pretty good pen that I'm also not going to mind losing. Good lines, not too thick on the fine, lasts well, inexpensive enough.
This is a core element of the early parts of Charles Stross' book Accelerando. (available online and in various ebook formats at http://www.antipope.org/charlie/blog-static/fiction/accelerando/accelerando-intro.html under a Creative Commons license of some sort).
The protagonist for the early part of the book is Manfred Macx, a "Venture Altruist" - he's not just an Open Source guy, he's an Open Ideas guy. There's some question about how much of Macx's personality (particularly the public-facing parts) is in the meat and how much is in the array of personal software agents that he interfaces with through his smart glasses.
I'd say that they don't currently have any real differentiation from the other products out there, at least not in a good way.
I'm using an HTC phone right now and I'm pretty pleased with it, but I bought it when it had already been discontinued after only being available for less than a year. Why did I buy the HTC Amaze 4G instead of the other options (HTC One S, Galaxy S II, etc.)? It has a camera button, and after my previous Samsung phone I really wanted that. It also has a replaceable battery and MicroSD card.
Grandstreams are inexpensive, work fine and are generally simple to configure, but for some of the low-end ones I've seen phone system upgrades that simply dropped the older low-end devices (e.g. BLF strobing stopped working and nobody was bothering to determine why). I know of one site that's using Yealink phones that they're pretty happy with so far (only a few months in). I've heard a variety of complaints about another (Aastra?) being a real headache to configure.
For email, there are many options one of which is Kolab (http://www.kolab.org) which is apparently better known in Europe. They're working up to their 3.0 release which includes some fairly major structural changes, including a switch from Horde to RoundCube for webmail, etc. Open source with commercial support also available, and unlike Scalix et al it's a live project.
I'm awaiting delivery of a new ThinkPad, and I expect that the second upgrade (after a 4 GB SODIMM) is going to be a mSATA SSD. It's a trade-off in that I can't do internal WWAN, but I wasn't going to be doing that anyway.
With a small one I could just configure it as a cache drive, but with a larger one I'll likely end up using that for the KUbuntu install and just leave Windows and storage on the HDD.
Crashplan is likely to be a reasonable option, including the free version, particularly if you're working from an office that is separate from your home, or if you have a friend, relative, or given the size of your daily data increase perhaps just a neighbor who's willing to accept a machine for you. Notes on that at the end.
What you'll want to do is set up a backup machine with plenty of disk space. This will likely need to be a single monster volume and you probably want to ensure that it's expandable; I can't help with the details of that but others here certainly can. What are the good expandable RAID options?
The backup machine will receive the initial backup sitting next to the source machine, but will then be moved offsite. Given the volume of data you described, creating a "seed" backup and just moving that would likely be more hassle than it's worth.
All machines involved should have Gigabit Ethernet, and I strongly recommend investing in a Gigabit network switch or better, an ABGN router with Gigabit ports. If you can find one, I like the WNDR3700v2 (that v2 is REALLY IMPORTANT and hard to find these days) running OpenWRT, but there are plenty of other options. You're still going to need an Internet connection for the machines to identify each other (I believe this uses Crashplan's servers even if you're not backing up to them).
Once you have all that set up, you're just going to install Crashplan on both machines, run your backup, and move the backup machine to its new home. Crashplan assigns unique IDs to machines and uses those for coordinating the backups between them, so the fact that the backup server has moved won't cause a problem.
Finally for the note about "just a neighbor" - if you're creating tens or hundreds of gigs of new data per month, just hosting your offsite backup may be a problem for many people (think bandwidth caps), and transmitting it may be just as big a problem for you. If you have or can find a reasonably close neighbor that you'll trust with an encrypted copy of your data and who's willing to host your machine, I'm going to strongly suggest that you set up the machine with a good wireless connector and have it in their house but on your wireless. Ideally this person would not be in the same building you're in (assuming condo/apartment). Wireless-N at 2.4GHz is likely to be your best bet for range, though it's more prone to interference than 5.5GHz.
The premise is a future in which copyright has expanded even further, and they're on the brink of â€oein perpetuity." Along with automated testing for matches within the entire existing corpus of art and music, the race is rapidly approaching the point where all combinations playing to humans are already known and protected. Forever.
The one time I was laid off (knowing it was coming for months - closing an entire facility, plus I got extended a couple times and had turned down an offer to move to Dayton, Ohio), I was working on wrapping up a project up to the very last day. The last parts were documenting, etc. but when I walked out the door I had my personal laptop that I'd been using for some development work and testing.
What did I do with the company information on that laptop? I zipped it all up, burned it to a CD along with an index/directory and notes on what might be of interest in case there was anything like homegrown test tools that wasn't on my main system, and mailed it to them. What did I get for all this? Thanks for being so great about everything, which kind of confused me - they'd offered to keep me on if I was willing to move and I refused, and I wasn't going to screw the people I'd been working with for years.
If you dislike the people you work with enough to screw them when you leave, you're in the wrong place (mentally, physically, whatever) already.
As it turned out, I ended up doing some fairly substantial hourly consulting for a different division of the same company a few years later, and I suspect that had I pouted my way out the door it wouldn't have happened. I didn't end up needing any of my old coworkers as references (jumped into freelance work with some other former employees), but I have no doubt that I'd have been able to get good references with no difficulties.
If 720p (1366x768) and a 15" screen is acceptable, $1500 is more than triple what you should be paying. Treat the machine as a commodity, expect it to get dropped/lost/etc. Set it up with automatic backups (e.g. Crashplan, Carbonite, Backblaze, etc.) and good encryption for when it's lost/stolen (if it's a single-user machine this may be easier since there are no cross-account encryption issues for backups) and save the money to get a better machine in a year or two. If you want CPU but don't care about graphics, get an Intel chip; if you're going to be playing games on an inexpensive laptop get an AMD A4 or A6 processor for the built-in graphics (comparable graphics on Intel requires an added graphics card).
As far as the sites go, I sent a nastygram over to Asus not too long ago noting that since clearly Marketing had completely subdued anyone with technical knowledge, I was sure that any laptop from them would continue to look mahvelous even after components started to fail. I also criticized the complete lack of any way to search by specifications other than manually opening each product's page.
First, if your boss thinks the programming is "monkey work" then you're already in trouble right there.
With only two of you total, and only one person even potentially capable of reviewing outsourced work, you personally are going to spend all of your time attempting to integrate not-quite-compatible or not-quite-complete or worst not-quite-right pieces that you get from outside. If what you're outsourcing is self-contained pieces you may have more luck, but even there unless you're simply contracting some things out to a small group of outside folks then you're going to see a lot of variation. Finding qualified contractors is always an issue, particularly if you're in the financial constraints I suspect based on the shift from a team to two people.
I can't speak well to the domestic vs. international outsourcing question, but I will say that A) your boss who's outsourcing the monkey work is going to want to put it out as cheap as possible and B) if you're outsourcing UI/content/etc. then there's going to be a lot of cultural stuff that may be subtly wrong in ways that will hurt your products.
Without knowing more about your business this may seem presumptuous, but are you focusing on the right thing? Do you have a good framework for building educational titles/apps/games/etc.? If so, can you improve it further so you can outsource development of complete packages using your framework? Who are your customers, and are they the right customers for what you have and can build?
In this case they're showing graphs of values from two separate runs. This is fine, there's no error information to be shown, it's not aggregated numbers as in a poll, etc. I'm just suggesting that if their results are that variable (even assuming the two graphs shown are extremes) then aggregated data (with standard deviations shown) might be more useful.
In the two sample runs they show, the Innovator does well in one and the "Milker" with multiple redundant apps does well in the other. The "Optimizer" who improves their best app comes in second in both, and I'd wonder if that holds over a larger set of simulations.
I suspect that what might be more interesting is the standard deviation of ending positions over many runs.
You don't give nearly enough detail here for people to be able to help much. Are you talking about an app for mobile devices where it'll cost a few dollars, business software for small/midsize companies, or (potentially) enterprise-level business software?
For an app, you could make it open source but sell it for a buck or two - most app purchasers will happily just get it through the relevant app store; those who will care about the open source nature will hopefully be willing to throw an (insignificant) couple of bucks to you particularly if you mention that (and the convenience of getting it that way vs. downloading/loading separately); those who could purchase but are too cheap to part with less than the cost of a Starbucks coffee are the ones who're probably more likely to pirate anyway. The biggest danger here is someone else lifting your code, rebranding and selling it themselves. (e.g. JMRI and Jacobsen v. Katzer)
For small business software, is it a turnkey app or something that requires setup? I have no great suggestions for turnkey apps unless you're offering customization/extension to clients; otherwise you're selling setup, configuration, support. Your business model here is that some of your customers COULD do this (or hire someone else to do it), but it's worth their time to have you do it. Further, whether sales of support are viable depends on the software and what it does.
For enterprise-scale, your selling point is that it will cost less for them to purchase the software and services from you than to dedicate resources to learn it. If learning, configuration, etc. are going to take someone a month plus some problems for users while things get straightened out, the cost in staff time is huge - a skilled IT person could easily be $10k/month with benefits, and if there are 500 users that lose 2 hours of productivity each for a month or two, you could be looking at six figures of total cost to do it in-house instead of getting the professional services from you to do it up front.
I used to work on software that sold (installed & configured) for $70-100k and higher. At one point early on I looked at what we were doing/selling and thought "Why are companies paying for this? We're not doing anything revolutionary here!" but it didn't take long to realize that while our larger customers COULD build something comparable in-house, it could well take a skilled programmer a year to do so assuming they had one available and idle. Basically, it was worth it for our customers to use our product instead of doing it themselves, and even at the prices we were charging the ROI was such that I wouldn't be surprised if we actually fudged numbers to make it look *worse* in some cases (I wasn't in sales, but I always thought going in and claiming a 3-6 month break-even ROI would be questionable as in "are you saying that our processes are THAT bad?")
I was going to post that, but it's hard to beat it coming from the original developer.
On a possibly-not-approved-by-said-developer note, if you want to play it you can find ways to do so that involve Atari ST emulators and bootleg content. The actual title of the game was "Sundog: Frozen Legacy"
Slashdot Mirror
Established US legal doctrine says that drones in US airspace are perfectly legal as long as they're being operated by foreign militaries.
Interesting. Does this mean that before too long there's going to be a nice glut of Samsung laptops being sold as refurbs? Replace, reflash, resell?
One of the choices people are saying would've been better is removing the requirement that it be for California businesses only, but by doing that they would have had a huge number of California residents who sold a business anywhere filing amended returns as far back as possible to get the no-longer-limited credits.
Next, marketing to alcohol manufacturers.
If it has a decent dual-mode color/b&w screen readable in direct light without backlighting, I'll shop at Wal-Mart for the first time in years, because this will immediately become an eInk killer.
Sure she'd like to get the money, but there's also the issue that if the company is going to book revenue in a particular way they have to deal with ALL of the things that go along with that. Companies paying commission don't get to say on one side "this was a great sale and we're going to compensate our executives on a great sales year" and on the other "that's an ok lease agreement you got, too bad commission on those is so low."
That's worse than the old Dilbert where the secretary "neglected" to put anything between the announcements of miserable numbers and an increased United Way push.
For email, it's actually really simple. What he sees in email headers (From, Subject, etc.) is the equivalent of the return address written in the top left corner of an envelope. There's absolutely nothing keeping you from putting false information there, and if he doesn't believe you ask him when's the last time he had to present identification to send a letter. What you're showing him instead is kind of like inspecting the cancellation mark on the stamp to determine that while the return address may say the White House, the letter was actually mailed from Portland, Oregon.
To give him an impression of the need to update, there are a few things to point out, and hopefully at least one will get through.
* First, among the most dangerous sites on the web these days are church websites - they're created as a volunteer effort by someone who may not even still be with the church (or who graduated HS and moved on in life). They're unmaintained. If they're infected, it may be a long time before someone even notices. In contrast, the "skeevy" sites like porn have a financial incentive to make sure their sites are safe.
* Second, once upon a time malware was written by spotty-faced geeks competing with each other for reputation. Those days are gone and have been gone for 20 years. These days malware is written by professional virus authors who do it for a living.
* Finally, show him the picture from http://www.deependresearch.org/2012/11/common-exploit-kits-2012-poster.html which shows a bunch of *commercially available* malware kits used to create new viruses and some of the security holes they target.
Another constraint that she may be considering is power efficiency - there may be tradeoffs between minimal wait time, etc. and power consumption, and the electricity cost for lifting an elevator car many times a day is not going to be negligible.
So far they've only done a Phase 1 trial which is to prove that it's not harmful, and the researchers called it "hideously expensive" at $1 million for 10 patients. If it shows clinical promise in Phase 2 and beyond, that price is likely to drop quite a bit and quite frankly the available MS treatments are also very expensive - if a single treatment is $100,000 but works for 5+ years, it may still be cheaper than what's currently available.
More information in an NBC News article: http://vitals.nbcnews.com/_news/2012/11/18/15246299-new-approach-could-treat-ms-other-autoimmune-diseases?lite
And the original article (for those willing to cough up $32 for a single article or with a subscription to the Nature Biotechnology journal): http://www.nature.com/nbt/journal/vaop/ncurrent/full/nbt.2434.html
I've settled on these as a pretty good pen that I'm also not going to mind losing. Good lines, not too thick on the fine, lasts well, inexpensive enough.
This is a core element of the early parts of Charles Stross' book Accelerando. (available online and in various ebook formats at http://www.antipope.org/charlie/blog-static/fiction/accelerando/accelerando-intro.html under a Creative Commons license of some sort).
The protagonist for the early part of the book is Manfred Macx, a "Venture Altruist" - he's not just an Open Source guy, he's an Open Ideas guy. There's some question about how much of Macx's personality (particularly the public-facing parts) is in the meat and how much is in the array of personal software agents that he interfaces with through his smart glasses.
I'd say that they don't currently have any real differentiation from the other products out there, at least not in a good way.
I'm using an HTC phone right now and I'm pretty pleased with it, but I bought it when it had already been discontinued after only being available for less than a year. Why did I buy the HTC Amaze 4G instead of the other options (HTC One S, Galaxy S II, etc.)? It has a camera button, and after my previous Samsung phone I really wanted that. It also has a replaceable battery and MicroSD card.
Grandstreams are inexpensive, work fine and are generally simple to configure, but for some of the low-end ones I've seen phone system upgrades that simply dropped the older low-end devices (e.g. BLF strobing stopped working and nobody was bothering to determine why). I know of one site that's using Yealink phones that they're pretty happy with so far (only a few months in). I've heard a variety of complaints about another (Aastra?) being a real headache to configure.
For email, there are many options one of which is Kolab (http://www.kolab.org) which is apparently better known in Europe. They're working up to their 3.0 release which includes some fairly major structural changes, including a switch from Horde to RoundCube for webmail, etc. Open source with commercial support also available, and unlike Scalix et al it's a live project.
I'm awaiting delivery of a new ThinkPad, and I expect that the second upgrade (after a 4 GB SODIMM) is going to be a mSATA SSD. It's a trade-off in that I can't do internal WWAN, but I wasn't going to be doing that anyway.
With a small one I could just configure it as a cache drive, but with a larger one I'll likely end up using that for the KUbuntu install and just leave Windows and storage on the HDD.
I'd love to see Assange go somewhere that's seeking to extradite Bush and/or Cheney and offer a swap.
Crashplan is likely to be a reasonable option, including the free version, particularly if you're working from an office that is separate from your home, or if you have a friend, relative, or given the size of your daily data increase perhaps just a neighbor who's willing to accept a machine for you. Notes on that at the end.
What you'll want to do is set up a backup machine with plenty of disk space. This will likely need to be a single monster volume and you probably want to ensure that it's expandable; I can't help with the details of that but others here certainly can. What are the good expandable RAID options?
The backup machine will receive the initial backup sitting next to the source machine, but will then be moved offsite. Given the volume of data you described, creating a "seed" backup and just moving that would likely be more hassle than it's worth.
All machines involved should have Gigabit Ethernet, and I strongly recommend investing in a Gigabit network switch or better, an ABGN router with Gigabit ports. If you can find one, I like the WNDR3700v2 (that v2 is REALLY IMPORTANT and hard to find these days) running OpenWRT, but there are plenty of other options. You're still going to need an Internet connection for the machines to identify each other (I believe this uses Crashplan's servers even if you're not backing up to them).
Once you have all that set up, you're just going to install Crashplan on both machines, run your backup, and move the backup machine to its new home. Crashplan assigns unique IDs to machines and uses those for coordinating the backups between them, so the fact that the backup server has moved won't cause a problem.
Finally for the note about "just a neighbor" - if you're creating tens or hundreds of gigs of new data per month, just hosting your offsite backup may be a problem for many people (think bandwidth caps), and transmitting it may be just as big a problem for you. If you have or can find a reasonably close neighbor that you'll trust with an encrypted copy of your data and who's willing to host your machine, I'm going to strongly suggest that you set up the machine with a good wireless connector and have it in their house but on your wireless. Ideally this person would not be in the same building you're in (assuming condo/apartment). Wireless-N at 2.4GHz is likely to be your best bet for range, though it's more prone to interference than 5.5GHz.
The premise is a future in which copyright has expanded even further, and they're on the brink of â€oein perpetuity." Along with automated testing for matches within the entire existing corpus of art and music, the race is rapidly approaching the point where all combinations playing to humans are already known and protected. Forever.
The one time I was laid off (knowing it was coming for months - closing an entire facility, plus I got extended a couple times and had turned down an offer to move to Dayton, Ohio), I was working on wrapping up a project up to the very last day. The last parts were documenting, etc. but when I walked out the door I had my personal laptop that I'd been using for some development work and testing.
What did I do with the company information on that laptop? I zipped it all up, burned it to a CD along with an index/directory and notes on what might be of interest in case there was anything like homegrown test tools that wasn't on my main system, and mailed it to them. What did I get for all this? Thanks for being so great about everything, which kind of confused me - they'd offered to keep me on if I was willing to move and I refused, and I wasn't going to screw the people I'd been working with for years.
If you dislike the people you work with enough to screw them when you leave, you're in the wrong place (mentally, physically, whatever) already.
As it turned out, I ended up doing some fairly substantial hourly consulting for a different division of the same company a few years later, and I suspect that had I pouted my way out the door it wouldn't have happened. I didn't end up needing any of my old coworkers as references (jumped into freelance work with some other former employees), but I have no doubt that I'd have been able to get good references with no difficulties.
If 720p (1366x768) and a 15" screen is acceptable, $1500 is more than triple what you should be paying. Treat the machine as a commodity, expect it to get dropped/lost/etc. Set it up with automatic backups (e.g. Crashplan, Carbonite, Backblaze, etc.) and good encryption for when it's lost/stolen (if it's a single-user machine this may be easier since there are no cross-account encryption issues for backups) and save the money to get a better machine in a year or two. If you want CPU but don't care about graphics, get an Intel chip; if you're going to be playing games on an inexpensive laptop get an AMD A4 or A6 processor for the built-in graphics (comparable graphics on Intel requires an added graphics card).
As far as the sites go, I sent a nastygram over to Asus not too long ago noting that since clearly Marketing had completely subdued anyone with technical knowledge, I was sure that any laptop from them would continue to look mahvelous even after components started to fail. I also criticized the complete lack of any way to search by specifications other than manually opening each product's page.
First, if your boss thinks the programming is "monkey work" then you're already in trouble right there.
With only two of you total, and only one person even potentially capable of reviewing outsourced work, you personally are going to spend all of your time attempting to integrate not-quite-compatible or not-quite-complete or worst not-quite-right pieces that you get from outside. If what you're outsourcing is self-contained pieces you may have more luck, but even there unless you're simply contracting some things out to a small group of outside folks then you're going to see a lot of variation. Finding qualified contractors is always an issue, particularly if you're in the financial constraints I suspect based on the shift from a team to two people.
I can't speak well to the domestic vs. international outsourcing question, but I will say that A) your boss who's outsourcing the monkey work is going to want to put it out as cheap as possible and B) if you're outsourcing UI/content/etc. then there's going to be a lot of cultural stuff that may be subtly wrong in ways that will hurt your products.
Without knowing more about your business this may seem presumptuous, but are you focusing on the right thing? Do you have a good framework for building educational titles/apps/games/etc.? If so, can you improve it further so you can outsource development of complete packages using your framework? Who are your customers, and are they the right customers for what you have and can build?
In this case they're showing graphs of values from two separate runs. This is fine, there's no error information to be shown, it's not aggregated numbers as in a poll, etc. I'm just suggesting that if their results are that variable (even assuming the two graphs shown are extremes) then aggregated data (with standard deviations shown) might be more useful.
In the two sample runs they show, the Innovator does well in one and the "Milker" with multiple redundant apps does well in the other. The "Optimizer" who improves their best app comes in second in both, and I'd wonder if that holds over a larger set of simulations.
I suspect that what might be more interesting is the standard deviation of ending positions over many runs.
I believe this has ended, but at one point Pegasus Mail was supported largely by the sale of manuals.
You don't give nearly enough detail here for people to be able to help much. Are you talking about an app for mobile devices where it'll cost a few dollars, business software for small/midsize companies, or (potentially) enterprise-level business software?
For an app, you could make it open source but sell it for a buck or two - most app purchasers will happily just get it through the relevant app store; those who will care about the open source nature will hopefully be willing to throw an (insignificant) couple of bucks to you particularly if you mention that (and the convenience of getting it that way vs. downloading/loading separately); those who could purchase but are too cheap to part with less than the cost of a Starbucks coffee are the ones who're probably more likely to pirate anyway. The biggest danger here is someone else lifting your code, rebranding and selling it themselves. (e.g. JMRI and Jacobsen v. Katzer)
For small business software, is it a turnkey app or something that requires setup? I have no great suggestions for turnkey apps unless you're offering customization/extension to clients; otherwise you're selling setup, configuration, support. Your business model here is that some of your customers COULD do this (or hire someone else to do it), but it's worth their time to have you do it. Further, whether sales of support are viable depends on the software and what it does.
For enterprise-scale, your selling point is that it will cost less for them to purchase the software and services from you than to dedicate resources to learn it. If learning, configuration, etc. are going to take someone a month plus some problems for users while things get straightened out, the cost in staff time is huge - a skilled IT person could easily be $10k/month with benefits, and if there are 500 users that lose 2 hours of productivity each for a month or two, you could be looking at six figures of total cost to do it in-house instead of getting the professional services from you to do it up front.
I used to work on software that sold (installed & configured) for $70-100k and higher. At one point early on I looked at what we were doing/selling and thought "Why are companies paying for this? We're not doing anything revolutionary here!" but it didn't take long to realize that while our larger customers COULD build something comparable in-house, it could well take a skilled programmer a year to do so assuming they had one available and idle. Basically, it was worth it for our customers to use our product instead of doing it themselves, and even at the prices we were charging the ROI was such that I wouldn't be surprised if we actually fudged numbers to make it look *worse* in some cases (I wasn't in sales, but I always thought going in and claiming a 3-6 month break-even ROI would be questionable as in "are you saying that our processes are THAT bad?")
I was going to post that, but it's hard to beat it coming from the original developer.
On a possibly-not-approved-by-said-developer note, if you want to play it you can find ways to do so that involve Atari ST emulators and bootleg content. The actual title of the game was "Sundog: Frozen Legacy"