Fog Creek Software
Discussion Board




Money saved by using open source can be used to em

Many people who promote open source say that the money saved by using open source can be used to hire programmers.

In their opinion, if a company doesn't buy Microsoft products, and replaces them with open source software, the money saved is enough to employ other programmers.


I think that this is false.

Let's do the math a little bit.


An average programmer's salary is 60 K (US $).

Depending on what you want to do, the cost of commercial software is much lower than that.

If you do the math, you will see that it's true.

Of course, that is, if you include reasonable priced software (which is still extremely powerful), and not Oracle UltraExpensiveDB 12i and IBM MoreExpensiveThanGodProduct version 16.


Let's to a bit of math:


Let's say you have 50 employees. All of them, right now, use Windows XP, and you have to upgrade the OS every 3 years.

So.. since Windows XP retail costs about $500, the cost of an OS from Microsoft, for the company is 50*$500 = $25000.

But, this cost is paid every 3 years (you don't have to upgrade every year). So, in fact, you are spending per year 8300 US$, for the OS.


Your internal team of 5 programmers uses Visual C#.

No problem. Visual C# costs about $200, and you have to upgrade it yearly, to get some productivity gains.

So.. $200*5 = $1000, per year.


Also, there is the cost of Windows 2003 Server for 50 users.. about 2500 US$.

And, you upgrade it every 2 years - that is, for Windows 2003, you pay 1250 US$.


So.. company costs, for year, for using Microsoft products:

8300 $ + 1000 $ + 1250 $ = 10550 $ per year


Of course, I forgot to include a few things in my calculation.

Also, I didn't include substantial savings that can be obtained by buying OEM licenses (which is practical, since a 3 years hardware upgrade cycle is very reasonable).


So.. can you pay a programmer with $10 K per year?

No!


I belive that the idea that a company can save enough money to employ additional programmers, by switching to open source, is, in most cases, FALSE.

You can see this by simply doing the math - adding up the cost, per year, of the commercial software products the company is using, etc.


If a company uses software that is truly extremely expensive, then indeed, the company can switch to open source and hire additional programmers using the savings.

But this approach usually has a large problem that may not be obvious: companies that use very expensive software do so for a reason. That reason is sometimes the fact that the more expensive tool is a more productive tool.

In this case, by switching to another product (which happens to be open source), the company loses some software development productivity.


Anyway, the conclusion is:

Savings obtained by switching to open source are usually much lower than the money needed to employ new programmers.

If something like this is proposed, do the math, and see how much the savings are, per year.

Before proposing somebody a switch to open source based on this reason, do the math, and see if you are right.

John
Thursday, January 01, 2004

"Savings obtained by switching to open source are usually much lower than the money needed to employ new programmers."

A 50 person company is hardly the benchmark of IT cost centers -- how about you redo the math for a financial company or telecommunications company (or government agency) with 10s of thousands of client workstations. For these organizations the payments to Microsoft yearly often end up in the millions of dollars.

Even for your 50 person company, lets be a little more reasonable here:

-It's doubtful that you'll have one single server to server the entire company. In fact if you follow Microsoft's own guidelines, you'll have a file server, a AD domain server, a web server, an email server, and a database server. How about adding the cost of five copies of Server 2003 with 50 CAL each ($2798/server, or $13980). Indeed one could state with reasonable truthfulness that Microsoft solutions require significantly more hardware (I'm not saying that it _actually_ requires more hardware, but rather that the norm with MS software is to throw much more hardware at it).

-Now add the cost of Exchange ($699 + 50 CALS @ $67 ea, so $4049) as obviously you need email and it's the natural Microsoft choice

-Add the cost of SQL Server, as it's the basis for DNA and .NET apps, so add $146/ user ($7300, or $4999 if a single processor will server your needs). If you plan on hosting your own web app to the public and maybe you draw from a database in a part of it,

-Every user needs a copy of Office, as this is the foundation of everything Microsoft. Let's cheap out and get the standard edition at $399 ea, or $19950.

There's easily another $40,000 or more over your quote. Add to this the fact that Microsoft may drop by uninvited any day and audit your license compliance. Note that purchasing something worth $40,000 at most organizations costs actually double or more because of the process, approval, management and maintenance of the spending.

BTW: The above is pretty much exactly the setup I saw first hand at one ~50 person firm.

Am I saying that it isn't worth every penny? Not even remotely -- In most cases it is. However there is a case to be made that the cost to run a Microsoft shop are measurable.

Dennis Forbes
Thursday, January 01, 2004

I'm not proposing anything...

I agree with your math but not your conclusions. I suspect that open source could become the next "fashion" among IT managements, as offshoring has been for the last couple of years. The appeal, as in the offshoring vogue, will be the appearance of savings on the surface.  Of course, as you imply, the savings gained by adopting open source may be wiped out and replaced with much higher payroll costs.

IE: I suspect that many companies that are being sticker-shocked by new product upgrades and whipsawed by continual patches and reinstalls will perceive that they can break out of the cycle with LAMP servers and OpenOffice.

However: open source has a very long way to go toward standardization, and entrepreneurs can fill this gap locally. I think those businesses that try to staff internally in skills they lack will crash and burn.

Here is a hypothetical example: it would not be beyond the skill of certain individual consultants to create standardized installations of servers and workstation environments that they would install, maintain and upgrade for their clients, for a total cost of ownership far less than commercial software.

A real example of how this could be done is as follows. I know a guy who once had a fairly lucrative consulting practice of developing green-screen Unix dumb terminal based distributed business applications using the ancient Curses terminal library. He did well throughout the 1990's buildup toward Windows and later, internet based "thin clients". He stuck with a conservative, simple, easy to maintain, almost impossible to break "template" for his applications. No J2EE, no ActiveX, no NIMDA, no hassles. Not even C++ - he only knows C. His clients loved it because they didn't have to worry about Madge in payroll getting a virus. Everything was "guaranteed" to work. His only sin was in not promoting his business properly.

Bored Bystander
Thursday, January 01, 2004

The saving is not in the software costs of the company that is writing the software; it is in the software costs of the client. Say I'm producing a database solution for a client with 50 computers. As Dennis has pointed out the client will have a software  cost including CAL's in the tens of thousands of dollars over a three year period. That is money that is freed up to pay the consultant who is providing the solution.

Obviously the saving is not enough for a 50 person shop to employ a programmer full time, but 50 person shops don't need full time programmers. The saving is sufficient to ensure that a competitive bid can be given; for smallish web apps, that maybe require a week or two's work, the saving involved in Open Source is much greater; in fact many of these would simply be uneconomic with commercial software.

Now often there are enough savings to be made through productivity gains using commercial softwre to obviate this - usng Access as a development tool can lead to increases in productivity that more than compensate for the cost - but the argument about Open Source saving money for programmers came about because this forum has a large number of seriously intellectually challenged posters who seem to think that they wil be better off being replaced by a commercial program costing thousands of dollars than by a free one. Presumably they find the second humiliating, like the aristocrats of yore who insisted on being beheaded instead of being hung like commoners.

Stephen Jones
Thursday, January 01, 2004

The cost depends on how you select the software.

Yes, MS Exchange is expensive, but there are a lot of other choices on the Windows plaform, some of which are less expensive.

For example:

http://www.serverwatch.com/mailservers.html

Users don't have to use MS Office. There are other office packages on the market, some commercial, some free.

You don't NEED to be an all-MS shop in order to use some MS technologies that are extremely productive.


Also - about your calculation which includes MS Office, SQL Server, and several other servers: you need to calculate the cost PER YEAR, and see if that cost per year is equal or greater than a programmer's salary.

John
Thursday, January 01, 2004

I know a 7 people non-IT company which has a full time programmer.

The programmer maintains the company's internal apps which play an important role in the company.

John
Thursday, January 01, 2004

Stephen, the misunderstanding about the effects of open source that you comment on are not typical on JOS and, in a different recent thread, came from an oss fan, not an opponent. (By the way, verifying general impressions of oss fans.)

I think John's point is a good one. In every case where I ever undertook contract software development, they already had their software infrastructure. They did not need to buy any new software so pricing of software was no barrier. I think that would be a typical scenario.

Companies that can't afford to buy basic business software are not going to pay much for developers.

Extend the analogy to other business costs. Do companies renting offices in cheap locations pay more for developers? No, they don't.

me
Thursday, January 01, 2004

----"I know a 7 people non-IT company which has a full time programmer.

The programmer maintains the company's internal apps which play an important role in the company. "----

In the health service in the UK you try and have one sysadmin support person for every 250 users. A seven user app shouldn't need maintaining! The company is either grossly inefficient in this respect or uses its programs so much that it is prepared to pay for somebody to be on call even though he might only do two or three hours work a week, or a combination of the two.

I agree with the last poster, in many cases the infrastructure is already there. The analogy to rent is stupid; if a company is paying a load in rent it probably has to pay more for developers because it is in an area with a high cost of living, and presumably it is paying a load in rent because it needs to target high-powered clients nearby, so it has more money. But if you take a look at the few companies that survived the dotcom bust you will find that they did so because they controlled costs.

What proportion of savings on Open Source are going to be passed on in extra work for developers is moot, but it is clear that developers are not going to get more work because the company pays extra for commercial software.

Stephen Jones
Thursday, January 01, 2004

"So.. can you pay a programmer with $10 K per year?"

Yes, if it's an open source developer:)

OUch
Thursday, January 01, 2004

In many cases small shops will pay developers by project, not hire them full time, because software development is not a core activity for them. So, if there is more money available, companies may be more willing to outsource more software projects in a contract basis. No, you won't pay one year salary with $10K, but you will pay for quite a few small jobs.

uncronopio
Thursday, January 01, 2004

"A 50 person company is hardly the benchmark of IT cost centers "


Something like 98% of small businesses in the US have less than $100K in revenue per year.


Thursday, January 01, 2004

A couple years ago, I talked to the president of a company that was based around a giant defense department contract that it had won implementing a major CCC (Command, Control, Communications) system. The system was supposed to be delivered in 3 years but they had been working on it for 10 due to extensions because of new requirements. He told me that their bids are audited before being accepted and they have to justify their low bids by proving they pay their developers the lowest salaries in industry. He said they would never consider paying more than $17,000 for a starting developer - that was the absolute maximum. I said, "At that rate, you'll only get the guys who absolutely can't work anyone else and likely have the lowest productivity." He responded, "Yep, that's the idea."

Dennis Atkins
Thursday, January 01, 2004

<I>Something like 98% of small businesses in the US have less than $100K in revenue per year.</I>

98% of small businesses are probably single person consultancy. They are hardly who we're talking about here.

Dennis Forbes
Thursday, January 01, 2004

(BTW: I realized that if you purchased your W2K3 Server CALs "per seat", you would only have to buy 50 per seat CALs, and could then access the 5 servers with no additional licenses on each)

Dennis Forbes
Friday, January 02, 2004

----"No, you won't pay one year salary with $10K, but you will pay for quite a few small jobs."------

Thanks uncropio. That's exactly what the poster in the other thread that provoked this one was saying. It was those small jobs that were putting the bread on his table.

Stephen Jones
Friday, January 02, 2004

I embrace and support free software and open source, and I too think that the "less money for software means more money for programmers" doesn't really correlate with reality.

Yes, open source and free software _will_ and _does_ drive programmer salaries down. And that's a good thing overall, because salaries are still inflated (And yes, I earn my living doing programming work).  Keeping salaries artificially high, as some people suggest, results in work going out to cheaper places. Avoiding free software works towards keeping salaries artificially high.

Unless you are sure that you will always be only a producer, and with some market monopoly, you _want_ your market to be efficient. Constantly being on both sides of the fence (software producer and software consumer), I know I do. YMMV, of course.

I recently heard of an interesting phenomenon in real estate markets - they often crash rather than decline slowly; And that's a result of sellers keeping prices artificially high (not believing that their property has, in fact, lost value -- until they must, for some reason, convert it to cash). Don't know about you guys, but I'd rather have continuous information about the real value of my property and skills -- with which I can adapt -- rather than wake up one morning and find out that the value is zero, and has been that way for some time.

Oh, and John & Dennis - you forgot to factor in liccense administration costs. Making sure you have all the required licenses takes time; Being audited takes _lots_ of time (and no matter how compliant you think you are, there's a good chance you'll be found non complying - read [ http://news.com.com/2008-1082_3-5065859.html ] if you haven't yet - it may also cost _lots_ of money).

From a capitalistic viewpoint, Open source / Free software is, more often than not, the right way for the consumer to go. Yep, it will hurt a few capitalists, but that does not make it a non-capitalistic paradigm. It seems, to me anyway, that the vendors and producers of closed source software will adapt.

Ori Berger
Friday, January 02, 2004

If everybody would think like you, then I would aggree with you.

But other professions, like doctors and lawyers, strongly protect their professions.

So.. why should we be suckers, and not protect others?

James
Friday, January 02, 2004

The point that less COTS software cost leads to more money for development is one I agree with, but all the argumentation in this discussion so far is insane.


- Nobody pays full retail pricing for business.

You get a volume licence deal, or upgrade the software together with your harware. This is >significantly< cheaper than those prices quoted above.

- The configuration is sub-optimal.

You would typically find SBS (including Exchange and SQL Server) for this business type which costs >significantly< less than the quoted config.

- Most small businesses I know upgrade only every 4-5 years now, making the software purchase even cheaper.

So I guess point 1 would be: That microsoft stuff is far less expensive than you guys indicate.

On to point 2:
That OSS installation is not going to be "free", but >more< expensive in licensing.

Typically it is going to be RedHat Enterprise Linux. Undoubtedly they will have some volume licencing to offer as well, but I guess the end result will still be more expensive. And yes, RedHat will audit you premises as well (with even less advance notice), so that "licence compliance" costing is a mute point.

but point 3: Is that a bad thing for IT people?

My feeling is that when the product cost goes down, the budget for all the services and complements around it does not stay the same or expand, but >shrinks<!
Every time a nice big piece of Iron is replaced by commodity PC's, every time some expensive Oracle or SAP solution gets replaced buy less expensive (OSS or MS or whatever) software, the "care and feeding" army surrounding that expensive centerpiece dissolves, and the new "replacement" system has to run with far less, not more, and cheaper, not better salaried  attendants/developers.

Just me (Sir to you)
Friday, January 02, 2004

of course that had to be:

"The point that less software cost ~DOES NOT LEAD~ to more money for development is one I agree with"

Just me (Sir to you)
Friday, January 02, 2004

Memo to self:

I should not start rants before my first cup of coffee.

Just me (Sir to you)
Friday, January 02, 2004

Dear Justme,
                    So there are less IT people around now tnan when everything ran on IBM mainframes?

                      You are probably right if you are just looking at one company. But the decline in the cost of running the IT infrastructure caused by technological improvement means that the cost to entry is lower and more companies enter. We've been seeing this happening for the last forty years.

                      And you should be saying that lower costs of software to not NECESSARILY lead to more money for developers. You'd think from reading some people in this thread that higher off the shelf software costs mean more money for custom developers, which is nuts.

Stephen Jones
Friday, January 02, 2004

And are you serioulsy recommending small business server for companies with 50 - 500 seats, which is what we are talkng about here. SBS is normally used for small offices with 5-15 computers.

You are correct that the costs are less than full retail, but volume discounts for business rarely go above 30%. The real saving are for consumers with OEM licenses or educational insititutions.

Stephen Jones
Friday, January 02, 2004

> Many people who promote open source say that the money saved by using open source can be used to hire programmers.

My take is that this is only correct if heavy emphasis is placed on the word "can".

The business will use whatever money it saves on what it thinks will be most advantageous for itself, period. They would be foolish to do otherwise.

Portabella
Friday, January 02, 2004

Hey can I whip this dead horse too? 
- I have to say this a very uncommon argument.  In fact, I have only heard that OSS can be used to add employees once before.  Perhaps, the better argument is it will offset savings for your bonus, raise, or job.  But really, has anyone ever seen a concrete case where OSS has added or removed a position?  I doubt it, because a "position" is filled or removed based on 1000 different criteria most having to do with skills and business need.
- Using some of logic I have seen here, if you want to save jobs, then don't buy software.  Think how many programmers your company will employ writing an OS, compiler, debugger, text editor, calculator, word processor, spread sheet,...  at least until you file for bankruptcy.  (I wonder if this would be the same group that says "welcome to the global economy" every-time someone speaks of off shoring.) 
- OSS is not going away until MS can lock down the hardware.  While a "sealed box" is part of the long term MS strategy, they took too long.  It is unlikely to ever happen now that non-OSS companies have billions at stake in an open box.
- MS is worried because of market saturation, so they want to change their licensing.  Sales people will tell you, if you are about to change your sales pitch, everyone starts looking everywhere again.  So, people are looking.  This is not good for any "established" company, as new comers often sweeten the deal to entice customers.  This is true of non-OSS too. (Think of Toyota in 1980s with GM)
- "OSS is not free."  I keep hearing people say this like there is someone who does not know that.  Going to the bathroom isn't free.  Walking to the coffee machine isn't free.  However, relative to other things, it is cheap.  Even the government does not use lowest cost is always the winner.  If that is the sole criteria for selecting anything by your company, be afraid -- be very afraid.
- OSS is not as good as MS, (or proprietary) software.  Again, this is news?  MS Notepad is not as good as EditPad, but is the relevant? Further, would not the same be true of Volkswagen relative to Porsche?  Yet, Volkswagen does a pretty good business.  How is that possible?  Because sometimes, many times, being the best is more than most of us need. 

So everyone is right, right?  So can't we all just get along?  The days of "no one ever got fired for choosing IBM" were replaced with ""no one ever got fired for choosing Microsoft."  Today, you will get fired for not having a sound business reason, beyond the company name, for selecting a product.  If a manager is about to sign for 10,000 copies of MS Office, they better have at least done a gap analysis on OpenOffice and any other office suite. 

Hopefully, this horse can be buried soon, it is starting to smell. ;)

MSHack
Friday, January 02, 2004

Switching to open source saves very little money per year.

Why? Because most software is upgraded every 3 years or so... so you have to divide the cost of the software by 3.

You have to calculate savings per year, in order to see if with that money, a company could employ another developer.

John
Friday, January 02, 2004

Good morning MSHack

"Perhaps, the better argument is it will offset savings for your bonus, raise, or job.  But really, has anyone ever seen a concrete case where OSS has added or removed a position?"

When TCO comparisons are done between Linux and Windows solution, they are always generally a wash -- the reason is that the analysis always presumes that the Linux side will require more, or more expensive, administration, thereby negating the extra cost of Windows. If the TCO of Linux and Windows are remotely comparable (as is claimed frequently to deride Linux), then clearly the IT staff of the organization are yielding more dollars, and Microsoft is receiving less.

"Using some of logic I have seen here, if you want to save jobs, then don't buy software.  Think how many programmers your company will employ writing an OS, compiler, debugger, text editor, calculator, word processor, spread sheet,...  at least until you file for bankruptcy. "

This is, I'm sorry to say, horribly dumb. No one is saying that you should write software yourself, but rather that some savings can be yielded via the use of comparable OSS -- If your staff is fully functional and happy with OpenOffice (which I think is a turd, but regardless), then doesn't that make sense?

Dennis Forbes
Friday, January 02, 2004

Dennis - I agree.  I should have had the sacasm meter turned on for my "build everything" comment.

MSHack
Friday, January 02, 2004

:-)

Dennis Forbes
Friday, January 02, 2004

WindowsXP is $500 and Visual C# is $200? That seems a little backwards.

Spleenor
Friday, January 02, 2004

Stephen,

the original posts talk about a 50 seat scenario, which is well within the SBS target.
I take your point about penetrating different segments of the market by basic cost reduction leading to overall IT workforce expansion. You are right there. Best example was Wintel vs. the old guard (IBM, SUN, ...), which put warpdrive on the IT growth machine. I am just not seeing OSS delivering a lower cost point so far.

Just me (Sir to you)
Friday, January 02, 2004

A lot depends on what the fifty seats are doing? If they are all running Exchange, and a large number of them are accessing an MS SQL app, and then there is the college intranet, and of course the company has its own web site, and all documents are kept centrally, then I wouldn't recommend SBS.

We're opening a new college this September, and I have sent out a spec that will have separate PDC, MS SQL server, Exchange Server, and File Server, Intranet server and an internet modem or router. As the educational discount is in the region of 75% I will probably go for an all MS solution there, but would certainly be thinking of Samba and Apache otherwise. And the computer labs will all have Samba servers eventually, though initially they will be P2P.

The reason for going for the five separate servers is firstly security, secondly simplicity, thirdly redundancy, and fourthly the fact that I can use fairly cheap IDE single processor bottom of the range servers in every case (though because of the importance of the data on the MS SQL server I will be recommending SCSI RAID 0), and for SQL and Exchange can go for per processor licenses. Our present college, which is considerably larger, runs a couple of quadruple Xeon Compaq servers. Overkill, and in fact they are running at about 20% utilization.

Stephen Jones
Friday, January 02, 2004

"...though because of the importance of the data on the MS SQL server I will be recommending SCSI RAID 0"

Just to clarify, you really mean RAID 1 (or RAID 10) here, right?

Dennis Forbes
Friday, January 02, 2004

Yep RAID 1.

I did get it right in the spec though!

Stephen Jones
Friday, January 02, 2004

Stephen,

In the mainframe era, all capable and interested IT people could get a job they liked. These days, costs of equipment are lower, but there are disproportionately more workers, so the prospects for individual workers are worse.

So if we're going to make comparisons with the mainframe era, the comparison should be the rate of IT employment.

Ori,  when you introduce capitalist ideals, you make the common mistake of open source advocates in stating that a benefit for the buyer as is also a benefit for the seller. Every other industry knows that's not the case.

As to people adapting, I think it's actually the buyers who will adapt, rather than commercial developers. As time goes on, the limitations of open source in the market will become more apparent, and that sector will be seen as the low-end amateur sector.

tree
Friday, January 02, 2004

"In the mainframe era, all capable and interested IT people could get a job they liked. These days, costs of equipment are lower, but there are disproportionately more workers, so the prospects for individual workers are worse."

This doesn't really make sense. In the past, computers were very expensive and programmers much less expensive compared to the price of the computer. For each computer in the world, there were tens of thousands of people.

Now there is one computer for each person, or even more and the computers cost almost nothing.

According to the commoditize your complement theory, programmers should be worth more now that computers are basically free.

I think your logic is the reverse of the way things really are.

Tony Chang
Friday, January 02, 2004

It's an interesting point but the fact remains that programmers ten and 20 years ago enjoyed much better job prospects and relative incomes.

If we're trying to apply Commoditise Your Complements (CYC) theory to this, then we have to understand that we have not been the only players in the game, and that we ourselves have been commoditized.

That is, we ourselves our complements to other parties, and those other parties have been busy commoditizing us.

tree
Friday, January 02, 2004

John, I am sure I am not the first to point this out, but in the interest of clarity I think it would be wiser to consider the cost per deployment unit (total cost of ownership) rather than just the software licenses involved in using open source. The cost per deployment will be a bigger value than just software licenses due to support costs and other administrative costs. There are those who insist employments can be gained by reducing total cost of ownership, but I think most people realize it's not so simple. In the end most people want to be in a situation where: 1) the cost of building the software profitably can be spread around a large enough body of customers; and 2) the cost of having a solution is far out-weighted by the problem it actually solves. It's possible to achieve both (very obvious) goals using OSS or proprietary software, how you want to go about doing it is really up to you.

Li-fan Chen
Saturday, January 03, 2004

If the domain of knowledge for both the software vendor and the customer's benefit can be grown organically and protected using OSS, so be it, on the other hand, if it most be proprietary and protected, so be it. For a really extreme example, if you purchased a quarter of a million dollar installation of a NT5-based short-run printing press, chances are the training, maintanence, administrative, retirement/migration, and customization cost would render the NT5 licenses laughably small.

Li-fan Chen
Saturday, January 03, 2004

I doubt that programmers had better salaries twenty years ago than they did four years ago!

The reason there are more programmers now is over supply; this has been partly caused by the drop in computer prices, so that learning programming is more accessible. Nothing whatsoever to do with Open Source.

Stephen Jones
Saturday, January 03, 2004

Li-fan Chen, surely the point is that the value of the expertise in the software _can't_ usefully be protected if the source code is revealed?

tree
Saturday, January 03, 2004

tree, I specifically mentioned that I consider this from both sides of the fence. I get paid developing software that isn't getting sold, but rather is used internally; I have to buy software products and services often to get my job done.

I suspect that's true of many programmers. Definitely not all - but consider how much money you'd have to spend _per year_ on software and software services if you wend independent, if you had to buy anything and everything you've used - then multiply by 10 to get a rough idea (The mere existence of free alternatives keeps prices "reasonable" in some sense - This is very visible in industries for which now free alternatives exist).

Therefore, the "good for the consumer" part is just as important to me as the "good of the producer". A free and efficient market is eventually good for everyone except those currently profiting off the market's inefficiency.

Ori Berger
Sunday, January 04, 2004

s/wend/went/
s/now free/no free/

Ori Berger
Sunday, January 04, 2004

Suppose the reverse happened -- open source software is no longer available, causing the proprietary players to seriously raise their prices due to less competition, and forcing you to pay for stuff that was free before.

You as an employer now evaluate the new costs of all the licenses needed for a developer -- version control tool, client and server OS licenses, database per-seat licenses, office suite, compiler/IDE, modeling tools, third-party libraries, intranet web server licenses, mid-tier app server per-seat licenses.  Say that the total price increase has now added 15% to the total cost of employing a developer.  If you had plans for hiring additional developers, that increase will almost certainly have a negative effect on how many you are willing to hire, how fast you are willing to hire them, or how much you are willing to pay them.

T. Norman
Sunday, January 04, 2004

"causing the proprietary players to seriously raise their prices due to less competition"

So that's why th e best IDE I ever used, Turbo Pascal, costs $79 a decade before anyone ever heard of open source. Now it all makes sense!

later
Sunday, January 04, 2004

Open source software was around long before Turbo Pascal.  The BSDs, for example.

Second, how much did other IDEs cost at the time?

T. Norman
Sunday, January 04, 2004

"An average programmer's salary is 60 K (US $)."

You're kidding right?  That's all??  Boy, I thought I was hurt when the job market tanked, but that's awful.  If that's all I could make, I'd find another line of work.  Now I can guess why most programmers are so mediocre.

v
Sunday, January 04, 2004

v, are you being sarcastic?

--
Sunday, January 04, 2004

Ori and T Norman, why do you presume alleged cost increases would impact developers only? People only hire developers if they need them, so that's not going to change.

In the fanciful event that software prices could be raised significantly, then:

1) maybe companies would demand that lawyers and accountants reduced their fees and salaries, while keeping all their developer, and

2) the lucky software companies would be hiring left, right and centre to build more and better products. So there would be oddles of interesting work.

But, most importantly, software costs are tiny in the budget for companies, and are much, much lower than staff costs, real estate costs, management costs and other costs of doing business.

tree
Sunday, January 04, 2004

So Turbo Pascal cost $79 because BSD was an free open source competitor. Gotcha!

(Didn't realize BSD ran on the early PCs - I learn something new on this board everyday!)

later
Sunday, January 04, 2004

1. Strawman arguments will get you nowhere.  Address the relevant points being discussed, not things that were never said or implied.

2. Turbo Pascal is one piece of software from a company that sucked and still sucks at various aspects of marketing including pricing, so it isn't a good representatition of what compilers cost at the time.

T. Norman
Sunday, January 04, 2004

"v, are you being sarcastic?"

No.  Most I know make half again more, which is still a big drop from three years ago.

v
Tuesday, January 06, 2004

*  Recent Topics

*  Fog Creek Home