User: Password:
|
|
Subscribe / Log in / New account

why software is different

why software is different

Posted Nov 15, 2012 8:37 UTC (Thu) by simlo (guest, #10866)
In reply to: why software is different by dlang
Parent article: Phipps: Stop patent mischief by curbing patent enforcement

The current problems with software patents are:

1) They are granted on small, low cost inventions, where others "invent" the same thing. The later should indicate that they are simply not novel and inventive enough to be patentable.

2) Tons of them are hidden inside a product, so you don't stand a chance of knowing what is patented and what is not.

3) Software patents are simply not needed to drive the software business: Without patents the products are copyrighted and that is incentive enough to keep producing software. There is no need to have a patent market around software as opposed to the drugs industry.

But what strikes me as a difference is that a in the drug industry a product more or less corresponds to one patent. That patent will only block copying of that single product - a specific treatment - not a lot of other product which happens to use the same technology.

Are we back a demanding a more or less one-to-one relation between products and patents? In software the small incremental inventions can not carry a product and shall therefore not be a patentable. In the drug industry a discovered drug is product and therefore patentable. But if one invents some algorithm of say making picture manipulation, and can make a business case around selling it as say plugings for Photoshop or Gimp, why should he not get the same patent protection as the drug company?


(Log in to post comments)

why software is different

Posted Nov 25, 2012 2:08 UTC (Sun) by steffen780 (guest, #68142) [Link]

> But what strikes me as a difference is that a in the drug industry a product more or less corresponds to one patent. That patent will only block copying of that single product - a specific treatment - not a lot of other product which happens to use the same technology.

Completely inaccurate, and in any case I don't see the relevance. Coming up with the treatments isn't even the issue in a lot of cases - delivering the treatment without killing the patient (or causing other unacceptable damage to him/her) is. Why should copyright on treatments not be sufficient?
And whilst arguing for drug patents keep in mind that drug companies are more than just notorious. It is standard procedure for them to publish only successful trials, but not the unsuccessful ones; I won't even start on pure BS like "restless leg syndrome". They spend (much) more on marketing than on R&D - and they don't get patents on marketing, why should they get patents on a much smaller cost factor?

why software is different

Posted Nov 25, 2012 2:11 UTC (Sun) by steffen780 (guest, #68142) [Link]

Just a small addendum, I am actually in favour of concentrating on software patents. There the ratio of damage compared to the virtually complete lack of any benefit appears to be highest. But patents are a terrible idea in general.

why software is different

Posted Nov 25, 2012 2:23 UTC (Sun) by Cyberax (✭ supporter ✭, #52523) [Link]

> Completely inaccurate, and in any case I don't see the relevance. Coming up with the treatments isn't even the issue in a lot of cases - delivering the treatment without killing the patient (or causing other unacceptable damage to him/her) is. Why should copyright on treatments not be sufficient?
What is the difference between a 'drug' and a 'treatment'?

>It is standard procedure for them to publish only successful trials, but not the unsuccessful ones
Pure BS.

Every drug before being granted an FDA approval must pass through public trials. Their results are, well, public ( http://clinicaltrials.gov/ ).

Companies are free to do their own post-approval trials for their own purposes, and they are not forced to publish their results. But so what?

why software is different

Posted Nov 25, 2012 4:35 UTC (Sun) by viro (subscriber, #7872) [Link]

That's one hell of a serious accusation. Selectively publishing only successful trials is an outright fraud, and very likely a literally deadly one, at that. Do you have specific examples? Details, please - not on the "everyone knows" or "such and such quack webshite says" level. This is not to say that fraud doesn't happen or that it's effectively punished (hell, Andy Wakefield is still out of jail - the bastard has gotten away with merely losing his UK license and this fraud has lead to hundreds of dead bodies), but if some company has been caught at what you describe, I'd rather be aware of that. To stay away from any stuff they'd developed, if nothing else. Verifiable proof, please.

why software is different

Posted Nov 25, 2012 11:35 UTC (Sun) by dark (guest, #8483) [Link]

Bad Science has a couple of summary articles with links to the data: http://www.badscience.net/category/hiding-data/

He's looking primarily at the UK regulatory situation but the problem is worldwide. I'm afraid it's systemic, not just "some company".

why software is different

Posted Nov 25, 2012 11:53 UTC (Sun) by Cyberax (✭ supporter ✭, #52523) [Link]

That's a bit different. Companies may creatively bend post-approval trials (for new drug uses) and/or not be sufficiently transparent.

But that can happen only _after_ the drug approval. FDA takes a very dim view on any data manipulation in approval trials. In fact, that's the reason these trials cost hundreds of millions of dollars.

And FDA (unlike USPTO) is decidedly NOT a "rubber-stamping" agency. FDA delights in denying approval for drugs on which companies could have spent billions of dollars.

why software is different

Posted Nov 26, 2012 15:10 UTC (Mon) by etienne (guest, #25256) [Link]

I do not know anything about this subject, but have seen not long ago on a BBC TV documentary some strange practices to approve drugs in India, and if I remember well it was not only for drugs to be sold inside India.
Only links I can find right now:
http://www.bbc.co.uk/news/magazine-20136654
http://www.reuters.com/article/2012/05/10/drugs-india-bri...
http://www.bbc.co.uk/news/world-asia-india-18018158

why software is different

Posted Nov 26, 2012 17:25 UTC (Mon) by bronson (subscriber, #4806) [Link]

Here's a good article showing how data manipulation isn't even needed. The companies stage the tests so the desired results are just significant and any heart problems or complications are hidden in the noise.

http://www.washingtonpost.com/business/economy/as-drug-in...

And it very much happens pre-approval.

why software is different

Posted Nov 26, 2012 17:45 UTC (Mon) by Cyberax (✭ supporter ✭, #52523) [Link]

Link doesn't work. But I very much doubt that wishful thinking of drug companies makes much difference in FDA approval. More likely, they just tend to overlook drug complications and exaggerate efficacy.

There were even a couple cases where the company was amazed that drug don't work after the controlled first or second phases of trials.

why software is different

Posted Nov 26, 2012 20:30 UTC (Mon) by bronson (subscriber, #4806) [Link]

Another attempt: http://www.washingtonpost.com/business/economy/as-drug-in... Definitely worth a read.

And if that link doesn't, work, the Washington Post might as well put all its content behind a paywall because I sure can't figure it out. Maybe they need a "Send a free link" button...?

why software is different

Posted Nov 27, 2012 0:45 UTC (Tue) by Cyberax (✭ supporter ✭, #52523) [Link]

Avandia is a very well known case. Media portrays it as if underhanded executives in GSK overlooked possible risks during its development.

But in reality similar risks are inherent in most of drug development, and there simply won't be any new drugs on the market if ALL risks were required to be zero. The additional risk of avandia is significant, but not really that great compared to risks of alternative medications (they are really not good enough), that's why the FDA let it remain on the market.

If you follow the drug industry then there are much more hilarious stories. Like finding out that your drug actually makes things worse for patients after the Phase II of clinical trials (at first company executives thought that the control group and the treatment group were mixed up).

Drug development is HARD.

why software is different

Posted Nov 27, 2012 11:03 UTC (Tue) by hummassa (subscriber, #307) [Link]

> But in reality similar risks are inherent in most of drug development, and there simply won't be any new drugs on the market if ALL risks were required to be zero.

Yes! and let's not forget that "no new drugs on the market" incurs in a casualty penalty also... Even if a new drug kills or maims 1% of the patients, if it also saves the lives of 10% of them, what should be the decision?

why software is different

Posted Nov 27, 2012 15:11 UTC (Tue) by mathstuf (subscriber, #69389) [Link]

It means that it's useless for 89.1% of the patients. The target of the drug better have a high fatality rate (to justify the 1% "bad stuff" rate) and low false positive (to justify the 9.9% efficacy plus the cost) to justify its use. Numbers should always be taken in context :) .

why software is different

Posted Dec 12, 2012 10:50 UTC (Wed) by hummassa (subscriber, #307) [Link]

A useless drug (as opposed to a killing drug) is a good drug.

if you have 1000 patients, half of them sick and the other half misdiagnosed:

* with the "useless" drug: 10 of them will die (because of the side effects of the drug, will kill 1% of the patients).

* without the "useless" drug: 50 of them will die (because the drug would have saved 10% of the really sick patients).

why software is different

Posted Dec 12, 2012 10:58 UTC (Wed) by hummassa (subscriber, #307) [Link]

I should go without saying that I mean "N people will die" DURING the treatment. All of them will die, of course.

why software is different

Posted Dec 12, 2012 10:59 UTC (Wed) by hummassa (subscriber, #307) [Link]

Again: IT should go without saying...

why software is different

Posted Dec 12, 2012 11:59 UTC (Wed) by mpr22 (subscriber, #60784) [Link]

We haven't in this example been told what the disease's fatality rate under the current standard-of-care treatment is, or even if there is a current standard-of-care treatment other than "nutrients, fluids, painkillers, and rest".

why software is different

Posted Nov 25, 2012 13:55 UTC (Sun) by paulj (subscriber, #341) [Link]

Someone else pointed at Ben Goldacre's "Bad Science" page. Note he has a book "Bad Pharma", a large chunk of which is about selective publication practices of the pharmaceutical industry. Well worth reading.


Copyright © 2017, Eklektix, Inc.
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds