So last quarter RH's income was 22m while MS's was 6B+.
That's profit, not revenue but anyway. MS's income depends almost
entirely on a government-granted monopoly, whereas Red Hat is
thriving even though they allow (in fact want) people to
copy their software.
Right, so from your point of view (you as a Russian citizen) it's mostly good that your government is "weak and slow" because
they aren't enforcing "rights" (ie monopolies) which would just benefit foreign companies. If you want s/China/Russia/ in my great-great-grandparent post.
I don't see any government monopoly behind commercial software development (could you elaborate on this, if you please?)
Copyright is a government-granted monopoly. It's not a force of nature. We all pay the police etc. to enforce it, and in return
the copyright holder benefits.
Why would you burn company resources on an in house project if a benefit can not be gained from it? If during the design phase of your project, it is discovered that the project is not worth more than the cost of development (hourly wages, etc) then the development should not happen in the first place. ANY code written in house must help the company turn a profit or otherwise create an advantage.
Software gets written in-house for many reasons which don't always come down to cost. For example, suitable
software might not exist (where by 'not exist' I include things like - it's available but doesn't run on our platforms,
or it's too difficult to purchase), or our developers just went ahead and wrote it, or we inherited it from another
company we bought. You often can't accurately know how much software really cost you to write, nor
how much monetary benefit you get, nor the true cost of alternatives that you didn't try out.
Nevertheless, putting that difficulty aside, you can see that some software really gives you a competitive
advantage. Take Google -- they have their own system management & deployment software, written by an
in-house team. They also wrote their own filesystem.
For Google, with 10^6 or 10^7 servers this is a real advantage over competitors. Their filesystem enabled
gmail and their management system enables them to manage their huge numbers of servers at a fraction
of the cost of what was thought possible previously. Major competitive advantage over say MS and Yahoo.
But Google, Microsoft and Yahoo also use software to manage their payrolls, reordering of stationary supplies,
and conference room scheduling. Is that software really the defining feature of Google? Do Google have
a huge advantage because they are able to schedule conference rooms more easily than their competitors? Or would they all
be better off just sharing that cost and not having a big IT cost sink (whether it is because the respective
companies wrote this software or they buy it in). Before answering, don't forget to include the cost
of this software in the equation: conference room scheduling software isn't all "competitive upside", unless
you somehow found a way to write/purchase this software for nothing.
And if it gives you an advantage, it will most likely give another company the same advantage.
Sure, in cutting-edge inventory control software (for Toyota), or if you've found the magic bullet
to develop software for nothing.
I don't know how many times I have to say this, but he's not talking about software which gives a competitive advantage. If your desktop remote management scripts give you such a big advantage over your competitors, then fine, keep them secret! But does your company also invent their own hammers because off-the-shelf hammers don't give you a secret advantage? Do you
refuse to use off-the-shelf CPUs in your computers and instead create your own? Maybe you do, if you're the NSA. Probably you don't.
Ok, but if my company sells the software, how can it benefit from that magnification?
There are two points here.
Firstly Jim isn't talking about packaged software
for sale. He clearly states that 95% of the software written in the world
is written inside companies, for companies themselves to use. Much of that
gives the company a competitive advantage, but also a lot of it
doesn't (think: toilet roll resupply spreadsheets, software to update
desktops remotely, certain payroll and expenses software, etc.). It's
this second type of non-competitive, internal software which Jim
is talking about.
Secondly, and the bigger point, your commercial software relies on
a strange government monopoly to exist, and really it only works
provided that everyone in the world agrees to police this monopoly
(at our expense and your gain). Maybe that will continue, maybe
places like, say, China won't see this as such a good deal for
them in the future and instead they'll just copy your software and
they won't see a particular advantage in them paying to maintain a monopoly
for your benefit. Red Hat as a company instead relies on support --
what the people who work for Red Hat know and can do -- and doesn't
require the expensive government monopoly.
But you can use and develop and add on to open source programs *without* giving anything back.
Sure you can, but this isn't a good long-term strategy.
Let's say you fork
Apache to add some super private feature. Now the development of Apache
continues, but that's OK because your private version is better (for you). However
after a while the 'public' Apache starts to diverge from your private copy,
getting bug fixes, loads of new features and security fixes. This happens
far faster than you can keep up because Apache has 100 developers for every
one of your own.
Now you're in a dilemma - do you keep your old, hard-to-maintain insecure private version? Or do you
contribute the new feature back to Apache and thus get all the other features
/ bug fixes that you need?
And this is, of course, without talking about the legal side of things
w.r.t. GPL'd software (not that the GPL forces you to contribute changes
back, unless you are distributing binaries).
One company pays for initial development. That money is gone and spent. They're not getting it back, ever, but every competitor will get that initial development for free. No matter how much community help is provided, that company that gives away code ALWAYS spends more than competitors who get their hands on the code,
That's complete nonsense. A couple of obvious counterexamples:
Rob McCool wrote the original Apache
webserver (it was called NCSA httpd at the time), and it was a very simple
HTTP/0.9-only web server, barely configurable, hardly any features. Apache
today is 100s of times larger and more featureful, has probably been ported
to more server platforms than any other server software, runs more than half
of the websites on the net, has a huge sphere of talent, books, tips, web pages, etc.
around it. Rob McCool did not write all this himself (in fact I don't think
he's even actively involved).
Example 2: Linus Torvalds releases Linux 0.1 on the web. It's pretty crappy,
it can just about boot on a basic single processor i386-based machine from a floppy, and has
drivers for just a few pieces of hardware. Today Linux is millions of lines
of code, has drivers for everything you can imagine, runs on everything and is
very efficient.
Linus's original work has been magnified maybe 10^4 - 10^6 times.
You write some crappy software to manage payroll for your 10 employee
company. It's barely more than a few scripts really. If you keep
developing that software on your own, it may one day become a few slightly better
documented scripts that can manage a 20 employee company. Released and
with contributions from ten other developers, it could become a hugely
powerful payroll suite that scales up and down, ported to everything,
masses of features etc.
Well, what's an advantage? How does a company that pays Joe Blow to write something, then give all of that code to competitors who did not have to pay Joe Blow, possibly benefit? It makes no business sense, whatsoever, other than PR.
It makes complete business sense because there are likely to be 10 or
100 times as many developers outside your company as in it, who would
be willing to contribute to improve the code. Alternatively you
could join an existing project, contribute Joe Blow (your 1 internal
developer) and gain 10x or 100x the development effort back because
of the other developers and companies involved.
In other words, it magnifies your own contribution. This, by the way,
is exactly Red Hat's business model: There are more packages just in
Fedora than there are employees at Red Hat, and I'm including
all the office staff, cleaners, HR, etc in that number. Believe me,
each Red Hat developer employee has their contributions magnified 10s or 100s of times.
If I'm the CEO of a big-ass Insurance Company, Bank, Airline or Widget Manufacturer and I just invested a bajillion hours of developer time into creating software that gives me an advantage over my competition, why would it be in my best interest to give my code away?
Because it's not this software that he's talking about. Obviously software
which gives you a competitive advantage is your lifeblood and you should
not give it away. It's all the other stuff that people should collaborate
on, stuff like your toilet-roll reordering spreadsheets and custom
desktop remote management scripts and (to a lesser extent) business apps like
payroll and inventory.
Share their stuff with others? Give other companies an advantage that WE paid for? NEVER!
And certainly for key business-driving software that attitude is right.
I'm sure FedEx have a large amount of routing software which they wrote
themselves and it may be their most important asset. However, FedEx's
expenses software or stationary supplies reordering software ain't so
critical to the business, and it's exactly this sort of thing which (if a company wrote
themselves) then the company should be encouraged to collaborate on.
Without patents, the result will be predictable: most people will keep their algorithms a closely guarded secret.
You can't keep an algorithm secret since it's so simple to disassemble code.
In the same way that patents don't help in pharmaceuticals these days either
since mass spectroscopy makes it (relatively) simple to work out what a drug
is composed of.
The idea that patents protect us against those who would keep
recipes secret belongs in the age of the alchemist.
Now Google comes in, indexes your site, and the search-in-search feature starts to take away from a good deal of the traffic that is searching for specific content on your site. Although google's results do link to your site, the ads you would have served on search result pages are now no longer paying you as much.
My heart bleeds for you. Oh wait, actually it doesn't.
You can (and have been able to for years)
tell Google to restrict adverts against your trademark,
either to a group of approved companies or indeed just to no adverts but your own company
(example). If you bothered to
read the article you'll see that Google will even remove the search-within-search feature
for your site if you ask them.
Google doesn't owe you a living, but in this case they give
you the tools and the ability to control adverts against
your site.
(Happy memories flood back...)
I knew about *PROT, it was the other people who didn't:-)
In those pre-internet days there was so little sharing of
information that each school probably learned the hacking
and the protection techniques separately, or the latter by carefully
reading manuals.
Econet... a good example of why you shouldn't design a network with
zero security for use by schoolchildren.
Amongst its many flaws: You could spoof any machine on the network just
by POKE-ing a single address (the machine's address was a single byte, I guess
they never expected more than 256 machines on a single shared segment).
I think the command was ?3362 = <node>
You could send text messages to anyone on the network. But get this: the messages
were injected into the remote system via the keyboard driver. That's right: You could
TYPE REMOTELY ON ANYONE'S KEYBOARD
over the network! What finally got me thrown out of the computing labs at school
at age 14 was writing a program which typed on all the keyboards in the lab at the same
time, controlling the whole lab from a single machine.
Another good one was the quota system used by the file server. It didn't
store total/available, as any sensible system would. Instead each user had
a single quota value (free space). The only problem was you could also write
to anyone else's file, eg. appending data to a file owned by another user.
When you did the append, your own quota would be diminished. But when the
other user deleted the file, *their* quota would be increased. I wrote
several trojan games which other people ran that surreptitiously appended
to a file owned by me. Then by deleting this file, I could steal other people's
quota and sell it back to them later.
"Functional" languages, actually a real misnomer. Really you should look at what are currently
classed as "research" languages
which are interesting because they have advanced features which make programming easier and safer.
The fear of cyber-warfare has climbed Whitehall's agenda since last year's attack on the Baltic nation of Estonia, in which Russian hackers swamped state servers with millions of electronic messages until they collapsed. The Estonian defence and foreign ministries and major banks were paralysed,
Except that these were done by some Estonian script kiddies, so it wasn't "CYBERWARFARE!!!11@@!"
Take the free software movement as an example... the movement isn't ruled by anyone, the society of human individuals (programmers) can license their work any way they like, but they _choose_ to push for freedom on to others.
... Thus showing that you understand neither anarchy nor the free software movement.
Free software is the perfect example of governance. It arises from the grassroots, the workers writing the
software, but it is there nevertheless. Take a look at projects like Debian. "Debian" is essentially
defined by its huge policy document, a body
of law which defines what does and what does not get distributed. Debian even has a constitution,
a leader, elected annually, a cabinet (technical committee), law enforcement (the security guys) etc.
And it's not just Debian. FSF, Fedora, FreeBSD all have similar organisations.
So even if you publish your own software on your own private website eventually you'll have to conform
(or your software will be packaged to conform) to all this law, if you ever want it in a major distribution.
I figure that's a rate of 2,800 per day, or 116 per hour. Nearly two spam messages, every minute, 24x7.... and most of it consists of duplicates. Why are the spammers doing this? Unless they are paid per message they send, I don't see it improving their chances of getting a message past filters.
The spam is being sent by a botnet of indeterminate size, and not always in direct communication back to their "masters". Sending
emails, even duplicates, costs nothing and is better than having to know the size of your botnet or be in constant communication
with the individual bots.
I don't know why the parent was modded troll. I used to be very
active on Wikipedia, but gave up after getting one two many things
I'd worked hard on deleted by power-tripping admins.
Slashdot Mirror
So last quarter RH's income was 22m while MS's was 6B+.
That's profit, not revenue but anyway. MS's income depends almost entirely on a government-granted monopoly, whereas Red Hat is thriving even though they allow (in fact want) people to copy their software.
Rich.
Right, so from your point of view (you as a Russian citizen) it's mostly good that your government is "weak and slow" because they aren't enforcing "rights" (ie monopolies) which would just benefit foreign companies. If you want s/China/Russia/ in my great-great-grandparent post.
Rich
See my reply here.
I don't see any government monopoly behind commercial software development (could you elaborate on this, if you please?)
Copyright is a government-granted monopoly. It's not a force of nature. We all pay the police etc. to enforce it, and in return the copyright holder benefits.
Rich.
Why would you burn company resources on an in house project if a benefit can not be gained from it? If during the design phase of your project, it is discovered that the project is not worth more than the cost of development (hourly wages, etc) then the development should not happen in the first place. ANY code written in house must help the company turn a profit or otherwise create an advantage.
Software gets written in-house for many reasons which don't always come down to cost. For example, suitable software might not exist (where by 'not exist' I include things like - it's available but doesn't run on our platforms, or it's too difficult to purchase), or our developers just went ahead and wrote it, or we inherited it from another company we bought. You often can't accurately know how much software really cost you to write, nor how much monetary benefit you get, nor the true cost of alternatives that you didn't try out.
Nevertheless, putting that difficulty aside, you can see that some software really gives you a competitive advantage. Take Google -- they have their own system management & deployment software, written by an in-house team. They also wrote their own filesystem. For Google, with 10^6 or 10^7 servers this is a real advantage over competitors. Their filesystem enabled gmail and their management system enables them to manage their huge numbers of servers at a fraction of the cost of what was thought possible previously. Major competitive advantage over say MS and Yahoo.
But Google, Microsoft and Yahoo also use software to manage their payrolls, reordering of stationary supplies, and conference room scheduling. Is that software really the defining feature of Google? Do Google have a huge advantage because they are able to schedule conference rooms more easily than their competitors? Or would they all be better off just sharing that cost and not having a big IT cost sink (whether it is because the respective companies wrote this software or they buy it in). Before answering, don't forget to include the cost of this software in the equation: conference room scheduling software isn't all "competitive upside", unless you somehow found a way to write/purchase this software for nothing.
And if it gives you an advantage, it will most likely give another company the same advantage.
Sure, in cutting-edge inventory control software (for Toyota), or if you've found the magic bullet to develop software for nothing.
Rich.
I don't know how many times I have to say this, but he's not talking about software which gives a competitive advantage. If your desktop remote management scripts give you such a big advantage over your competitors, then fine, keep them secret! But does your company also invent their own hammers because off-the-shelf hammers don't give you a secret advantage? Do you refuse to use off-the-shelf CPUs in your computers and instead create your own? Maybe you do, if you're the NSA. Probably you don't.
Rich.
Ok, but if my company sells the software, how can it benefit from that magnification?
There are two points here.
Firstly Jim isn't talking about packaged software for sale. He clearly states that 95% of the software written in the world is written inside companies, for companies themselves to use. Much of that gives the company a competitive advantage, but also a lot of it doesn't (think: toilet roll resupply spreadsheets, software to update desktops remotely, certain payroll and expenses software, etc.). It's this second type of non-competitive, internal software which Jim is talking about.
Secondly, and the bigger point, your commercial software relies on a strange government monopoly to exist, and really it only works provided that everyone in the world agrees to police this monopoly (at our expense and your gain). Maybe that will continue, maybe places like, say, China won't see this as such a good deal for them in the future and instead they'll just copy your software and they won't see a particular advantage in them paying to maintain a monopoly for your benefit. Red Hat as a company instead relies on support -- what the people who work for Red Hat know and can do -- and doesn't require the expensive government monopoly.
Rich.
But you can use and develop and add on to open source programs *without* giving anything back.
Sure you can, but this isn't a good long-term strategy.
Let's say you fork Apache to add some super private feature. Now the development of Apache continues, but that's OK because your private version is better (for you). However after a while the 'public' Apache starts to diverge from your private copy, getting bug fixes, loads of new features and security fixes. This happens far faster than you can keep up because Apache has 100 developers for every one of your own.
Now you're in a dilemma - do you keep your old, hard-to-maintain insecure private version? Or do you contribute the new feature back to Apache and thus get all the other features / bug fixes that you need?
And this is, of course, without talking about the legal side of things w.r.t. GPL'd software (not that the GPL forces you to contribute changes back, unless you are distributing binaries).
Rich.
One company pays for initial development. That money is gone and spent. They're not getting it back, ever, but every competitor will get that initial development for free. No matter how much community help is provided, that company that gives away code ALWAYS spends more than competitors who get their hands on the code,
That's complete nonsense. A couple of obvious counterexamples:
Rob McCool wrote the original Apache webserver (it was called NCSA httpd at the time), and it was a very simple HTTP/0.9-only web server, barely configurable, hardly any features. Apache today is 100s of times larger and more featureful, has probably been ported to more server platforms than any other server software, runs more than half of the websites on the net, has a huge sphere of talent, books, tips, web pages, etc. around it. Rob McCool did not write all this himself (in fact I don't think he's even actively involved).
Example 2: Linus Torvalds releases Linux 0.1 on the web. It's pretty crappy, it can just about boot on a basic single processor i386-based machine from a floppy, and has drivers for just a few pieces of hardware. Today Linux is millions of lines of code, has drivers for everything you can imagine, runs on everything and is very efficient. Linus's original work has been magnified maybe 10^4 - 10^6 times.
You write some crappy software to manage payroll for your 10 employee company. It's barely more than a few scripts really. If you keep developing that software on your own, it may one day become a few slightly better documented scripts that can manage a 20 employee company. Released and with contributions from ten other developers, it could become a hugely powerful payroll suite that scales up and down, ported to everything, masses of features etc.
Rich.
Well, what's an advantage? How does a company that pays Joe Blow to write something, then give all of that code to competitors who did not have to pay Joe Blow, possibly benefit? It makes no business sense, whatsoever, other than PR.
It makes complete business sense because there are likely to be 10 or 100 times as many developers outside your company as in it, who would be willing to contribute to improve the code. Alternatively you could join an existing project, contribute Joe Blow (your 1 internal developer) and gain 10x or 100x the development effort back because of the other developers and companies involved.
In other words, it magnifies your own contribution. This, by the way, is exactly Red Hat's business model: There are more packages just in Fedora than there are employees at Red Hat, and I'm including all the office staff, cleaners, HR, etc in that number. Believe me, each Red Hat developer employee has their contributions magnified 10s or 100s of times.
Rich.
If I'm the CEO of a big-ass Insurance Company, Bank, Airline or Widget Manufacturer and I just invested a bajillion hours of developer time into creating software that gives me an advantage over my competition, why would it be in my best interest to give my code away?
Because it's not this software that he's talking about. Obviously software which gives you a competitive advantage is your lifeblood and you should not give it away. It's all the other stuff that people should collaborate on, stuff like your toilet-roll reordering spreadsheets and custom desktop remote management scripts and (to a lesser extent) business apps like payroll and inventory.
Rich.
Share their stuff with others? Give other companies an advantage that WE paid for? NEVER!
And certainly for key business-driving software that attitude is right. I'm sure FedEx have a large amount of routing software which they wrote themselves and it may be their most important asset. However, FedEx's expenses software or stationary supplies reordering software ain't so critical to the business, and it's exactly this sort of thing which (if a company wrote themselves) then the company should be encouraged to collaborate on.
Rich.
Without patents, the result will be predictable: most people will keep their algorithms a closely guarded secret.
You can't keep an algorithm secret since it's so simple to disassemble code.
In the same way that patents don't help in pharmaceuticals these days either since mass spectroscopy makes it (relatively) simple to work out what a drug is composed of.
The idea that patents protect us against those who would keep recipes secret belongs in the age of the alchemist.
Rich.
Now Google comes in, indexes your site, and the search-in-search feature starts to take away from a good deal of the traffic that is searching for specific content on your site. Although google's results do link to your site, the ads you would have served on search result pages are now no longer paying you as much.
My heart bleeds for you. Oh wait, actually it doesn't.
You can (and have been able to for years) tell Google to restrict adverts against your trademark, either to a group of approved companies or indeed just to no adverts but your own company (example). If you bothered to read the article you'll see that Google will even remove the search-within-search feature for your site if you ask them.
Google doesn't owe you a living, but in this case they give you the tools and the ability to control adverts against your site.
Rich.
The scary thing is if we were kids doing it nowadays, they'd probably call the police, take our DNA and put us in prison.
Rich.
I guess no one told you about *PROT
(Happy memories flood back ...)
I knew about *PROT, it was the other people who didn't :-)
In those pre-internet days there was so little sharing of information that each school probably learned the hacking and the protection techniques separately, or the latter by carefully reading manuals.
Rich.
Econet ... a good example of why you shouldn't design a network with
zero security for use by schoolchildren.
Amongst its many flaws: You could spoof any machine on the network just by POKE-ing a single address (the machine's address was a single byte, I guess they never expected more than 256 machines on a single shared segment). I think the command was ?3362 = <node>
You could send text messages to anyone on the network. But get this: the messages were injected into the remote system via the keyboard driver. That's right: You could TYPE REMOTELY ON ANYONE'S KEYBOARD over the network! What finally got me thrown out of the computing labs at school at age 14 was writing a program which typed on all the keyboards in the lab at the same time, controlling the whole lab from a single machine.
Another good one was the quota system used by the file server. It didn't store total/available, as any sensible system would. Instead each user had a single quota value (free space). The only problem was you could also write to anyone else's file, eg. appending data to a file owned by another user. When you did the append, your own quota would be diminished. But when the other user deleted the file, *their* quota would be increased. I wrote several trojan games which other people ran that surreptitiously appended to a file owned by me. Then by deleting this file, I could steal other people's quota and sell it back to them later.
Ah, misspent youth ...
Rich.
Obligatory:
My FORTH tutorial
Rich.
"Functional" languages, actually a real misnomer. Really you should look at what are currently classed as "research" languages which are interesting because they have advanced features which make programming easier and safer.
My list would include:
Rich.
From TFA:
The fear of cyber-warfare has climbed Whitehall's agenda since last year's attack on the Baltic nation of Estonia, in which Russian hackers swamped state servers with millions of electronic messages until they collapsed. The Estonian defence and foreign ministries and major banks were paralysed,
Except that these were done by some Estonian script kiddies, so it wasn't "CYBERWARFARE!!!11@@!"
Rich.
Take the free software movement as an example... the movement isn't ruled by anyone, the society of human individuals (programmers) can license their work any way they like, but they _choose_ to push for freedom on to others.
Free software is the perfect example of governance. It arises from the grassroots, the workers writing the software, but it is there nevertheless. Take a look at projects like Debian. "Debian" is essentially defined by its huge policy document, a body of law which defines what does and what does not get distributed. Debian even has a constitution, a leader, elected annually, a cabinet (technical committee), law enforcement (the security guys) etc.
And it's not just Debian. FSF, Fedora, FreeBSD all have similar organisations.
So even if you publish your own software on your own private website eventually you'll have to conform (or your software will be packaged to conform) to all this law, if you ever want it in a major distribution.
Rich.
I figure that's a rate of 2,800 per day, or 116 per hour. Nearly two spam messages, every minute, 24x7.... and most of it consists of duplicates. Why are the spammers doing this? Unless they are paid per message they send, I don't see it improving their chances of getting a message past filters.
The spam is being sent by a botnet of indeterminate size, and not always in direct communication back to their "masters". Sending emails, even duplicates, costs nothing and is better than having to know the size of your botnet or be in constant communication with the individual bots.
Rich.
This is a sales pitch, there's nothing new in that article. Google is just fishing for more business for postini...
You mean TFA is just a sophisticated form of spam :-)
Rich.
I don't know why the parent was modded troll. I used to be very active on Wikipedia, but gave up after getting one two many things I'd worked hard on deleted by power-tripping admins.
Rich.
They seem to be the #1 first world / major power on that list.
Rich.