1. Software costs next to nothing to produce. Why is it so easy to have startups in this sector? What is the overhead of a software/web company? It is almost all intellectual. Ask Microsoft. Pure profit.
2. It is difficult for examiners to find prior art because there is no "official" channel for "publishing" software research that they recognise. There is much know-how that is common knowledge among developers but is never published in any paper, textbook, standard, RFC, etc. The "database" for searching all these sources may end being a web search engine. That is not ideal.
1. Drugs are very expensive to produce. Not only the research, the equipment, the staff, but compliance with the regulatory regime. There are no protections against bad software to protect users (developers operate in a wild west type atmosphere), but we have many rules about medicines. These are designed to protect patients. Can you imagine an unregulated pharma industry? Profits take precedence over safety. It would be very dangerous for patients.
2. There is a littany of journals that public chemical, biological and medical research and they can all be searched easily in combined databases, to which most all pharmaceutical patent owners subscribe. It is not too difficult for examiners to find prior art and establish publication dates.
> 1. Software costs next to nothing to produce. Why is it so easy to have startups in this sector? What is the overhead of a software/web company? It is almost all intellectual. Ask Microsoft. Pure profit.
I'd argue software is extremelty expensive to produce, on the basis that distributing software is so insignificantly expensive that it could be called free. Relative to the costs of distribution, paying developers to write an application is extraordinarily expensive. So is maintaining that code, providing support, and advertising the product. I'd "guess" startups are so abundant in the software sector because distribution is so cheap, not because development is. Entering any other market's distribution channels on an international scale is prohibitively expensive for almost anyone but an already successful mid-sized company.
> 2. It is difficult for examiners to find prior art because there is no "official" channel for "publishing" software research that they recognise. There is much know-how that is common knowledge among developers but is never published in any paper, textbook, standard, RFC, etc. The "database" for searching all these sources may end being a web search engine. That is not ideal.
Even more fundamentally, it is hard to find prior art because software is information / knowledge and traditionally (in the context of math at least) we never had patents on it. If you consider the physical properties of software, it is just the presence of absence of charge over billions of indicants. Does the interface that exactly duplicates the ios home screen but has a binary representation that is nothing like the disassembly of it represent infringement, or the instant messenger whose disassembly looks 95% like the ios home screen? What happens when a music track and an android map are 90% binary identical?
> 1. Drugs are very expensive to produce. Not only the research, the equipment, the staff, but compliance with the regulatory regime. There are no protections against bad software to protect users (developers operate in a wild west type atmosphere), but we have many rules about medicines. These are designed to protect patients. Can you imagine an unregulated pharma industry? Profits take precedence over safety. It would be very dangerous for patients.
Absolutely you need to regulate the pharma industry somewhat, but I would also argue it is a consumers responsibility to know what they are putting in their bodies. We expect people to be able to judge their own alcohol consumption, how are prescriptions any different? If you make knowledge easily and readily available on what drugs do and their effects, consumers could make informed decisions. A lot of the fears in the phara industry is in how close-boxed knowledge is - you assume your doctor knows best, and don't investigate the effects of drugs you consume. I'd call it foolish to not understand the risks of what you are consuming.
> 2. There is a littany of journals that public chemical, biological and medical research and they can all be searched easily in combined databases, to which most all pharmaceutical patent owners subscribe. It is not too difficult for examiners to find prior art and establish publication dates.
That doesn't mean that a government granted monopoly on drugs is a good thing. When the expenses of drug research (which end up significantly being the rigorous 3 phase testing drugs need to pass to be allowed to be sold to end users) are so high, and profit motive is involved, drug costs spiral out of control. For Americans like myself without health insurance, we just can't afford to buy any of these drugs even with a prescription because they can cost upwards of $200 or more a bottle. The fact the actual drugs only cost at most a few dollars to produce per bottle, so the rest is profit to recoup research costs, doesn't help the fact that the good is cheap to make and is artificially expensive to recoup tremendous research costs. One (not me, but it is a theoretical) might argue that the R&D of a to-market drug could be just paid in full to whoever developed it, and that the formula itself could be released public domain. Taxes go up, drugs are cheap.
The key is that software development is an unavoidable sunk cost. Once it has been paid, it can be ignored in terms of considering per-unit pricing.
The actual marginal cost of each unit of software sold is pretty close to zero. This is what makes it so fantastically profitable compared to almost anything else.
The same is true of most drugs. Marginal cost is near zero compared to development cost.
In neither case do you necessarily have fantastic profits although if you can sell at scale at a good price it becomes very likely but if you have a drug for a niche market or software in a crowded market it is quite possible to make a loss.
This is neither an argument for or against the patent abolition proposition which I am torn on. I can definitely imagine with alternative arrangements for Pharma we could be better off than currently but I believe that there must be something possible that is better than both what we have now and better than nothing.
I think in half a century we won't have the backwards investment funding model in industries where units are extremely cheap but R&D id expensive. Instead you would have something akin to crowdfunding with guarantees and insurance against failed investments by those that want the end product, so the motive is not profit, but end product. That people who want a thing fund the creation of the thing, instead of people who want money funding an expensive up front thing that is then unnaturally restricted with patents and copyright to try to profit off the unit sales, when in many cases (pharma) they are near free, or in the case of software (past the first copy being distributed) are absolutely free (if those that have the software willingly share it).
PageKite looks interesting. Although when I read the list under "without PageKite, you would have to:" it seems easier than PageKite. Crazy I know. :) (The catch I think is getting that static reachable IP. Most people do not have that. But I think there are workarounds.)
Question: Why the base64 encoding instead of a simple gzipped tar file? Also why instruct people to feed untrusted stuff from the internet directly into bash via curl? That is definitely not safe as a matter of policy.
But I like the idea of PacketKite. Nicely done.
Also curious about Dalpay. Are they a good alternative to PayPal?
The main point of PageKite is converting a networking/sysadmin problem, into a software problem. This has fundamental implications for scaling it out (software copying scales infinitely, sysadmins don't).
I ended up with the weird merged python file (see: http://pagekite.net/wiki/Floss/PyBreeder/) because I wanted the continued perceived simplicity of a single .py, but I ended up wanting to refactor my code into submodules. So I made PyBreeder to reverse the refactoring, basically. It is a little silly and will probably go away over time. Regarding the bash installer, I don't see much of a difference between trusting my installer or trusting pagekite itself. Both came from the network. The bash installer is at least easier to review than "install.exe" or the shell scripts embedded in an RPM or .deb. My goal was to get people up and running in literally two commands, this was the only cross-distro/cross-OS way to accomplish that.
Dalpay are OK, but they have their warts. I'd rather deal with a small local company than a monster like PayPal, and being in Iceland my options for actually receiving funds from PayPal and the likes were actually rather limited.
I love this HTML.* If only all the web could follow this simple way to convey information. CONTENT. Heck, that's what makes HN interesting. It's the lack of web developer fluff. Just text. CONTENT.
*But why the reference to tags.js? Javascript is not used.
That's assuming the user has js enabled. What I like about this is that it's just one line. No inline script (that would be just bloat for the user with js turned off.)
IMDB was originally crowsourced. Started on Usenet.
Can anyone assert ownership over and derive commercial benefit from a data source that traces its origins to volunteer contributors? What would be legally required to do that?
2. It is difficult for examiners to find prior art because there is no "official" channel for "publishing" software research that they recognise. There is much know-how that is common knowledge among developers but is never published in any paper, textbook, standard, RFC, etc. The "database" for searching all these sources may end being a web search engine. That is not ideal.
1. Drugs are very expensive to produce. Not only the research, the equipment, the staff, but compliance with the regulatory regime. There are no protections against bad software to protect users (developers operate in a wild west type atmosphere), but we have many rules about medicines. These are designed to protect patients. Can you imagine an unregulated pharma industry? Profits take precedence over safety. It would be very dangerous for patients.
2. There is a littany of journals that public chemical, biological and medical research and they can all be searched easily in combined databases, to which most all pharmaceutical patent owners subscribe. It is not too difficult for examiners to find prior art and establish publication dates.
These are my thoughts.