... and the UK and Germany and France and Spain and Finland and Sweden and Greece and Denmark and Luxembourg and Belgium and Portugal and the Netherlands and Ireland and more coming soon
On the one hand we have the EPC and the Guidelines which define the process and how the Examiner's should work. The process itself is followed in the sense that the EPO obeys the deadlines in the EPC and issues the correct official communications, apart from the occasional balls up. The issue is whether examiners are letting through cases they shouldn't.
I cannot think of an example, speaking as a patent attorney, of a case where the examiner seemed to allow a case just to get it off their desks. If we only consider the patentability of software as such, dodgy grants are more likely to be the result of a lack of relevant technical knowledge making arguments regarding the presence of technical effect seem more persuasive than they should. Also, since examination is an administrative not a legal process, the applicant tends to be given the benefit of the doubt.
It is now very easy to monitor European patent application publications. Anyone with an interest in a software product should set some time aside to look at the newly published applications. If any are not new or obvious, the prior art that shows this can be sent to the EPO, anonymously even, with a few paragraphs explaining the relevance of the prior art. If the invention seems to be just software, you can write to the EPO and explain why the invention merely uses software to solve a software problem, i.e. why it does not make a technical contribution.
Even if the examiner takes no action, the observations can have knock-on effects. For example, the patentee may not be allowed to amend the patent later to overcome your prior art because they allowed the patent to be granted knowing it to be bad. Prior art which is also relevant prior art for a corresponding US application will need to be notified to the USPTO by the applicant.
It is unlikely in the extreme that patents for software-implemented inventions which are technical in character, i.e. not-software as such, will be excluded from patentability. At one extreme, there are patent for televisions that employ software to process the television signals (patentable). At the nother extreme there are content management frameworks for web sites (sticking my neck out - unpatentable unless there is some effect such as a reduction in resource requirments for the same performance). In the middle are data compression and transmission systems and methods. Compressing video data can be performed using software and a PC or using custom hardware. If the custom hardware is patentable, what justification is there for not granting a patent covering the software + PC implementation? They both represent systems that operate according to an underlying algorithm.
There are other practical problems that are highlighted by cases where hardware and software implementations are possible. If a patent application describes a system comprising hardware than converts signal A into signal B and hardware for convering signal B into signal C, the main claim might be:-
A signal converter comprising means for convering a first signal into a second signal and means for converting the second signal into a third signal.
In Europe, this claim would be infringed by a hardware only device, software + a PC and by a hybrid embedded processor device. It would be difficult to frame a law that distinguished between these three examples.
So I am certain the EU will give the EPO the right to award software patents. The only way to stop this is by periodically checking what [provisional patents] applications are [awarded] published and bombard the EPO with negative comments and prior arts.
This is why applications are published and why the EPC has specific provision (Art. 115) for this. There are no fees for filing third party observations.
The EPO is self-financed. They do not receive any money from the EU states.
Almost true, it receives a cut of patent renewals fees from the contracting states (Art. 39 EPC).
The EPO makes money from patent royalties.
Nonsense.
The EPO does not have much money.
Irrelevant as long as its running costs are covered.
The EPO is ready to accept anything to make money as long as nobody complains.
Unsubstantiated fantasy
Profit!
But no shareholders so what reason is there for the EPO to maximise profits?
The way it works at the EPO is the following: Someone submits something to get a patent.
If it is not completely stupid, a provisional patent is awarded, even if obvious prior art exists!!!
If there has been no negative comments after one year, the patent is awarded.
WRONG - the poster was not only an Anonymous Coward, he is an anonymous know nothing.
Without the word "European", the headline is ambiguous because is could be suggesting that the EU is moving towards harmonisation with the US system (although hell will freeze over first).
This is all making a mountain out of a molehill. The situation under the European Patent Convention, which set the patentability criteria for the EPO and the national patent systems of the EU and other European states, is currently that a technical contribution is required for a software-implemented invention to be patentable. Its just that the EPC doesn't actually use these words.
The reasoning that leads from the wording of the EPC to the requirement for a technical contribution is summarised here.
<RANT>
"Software Patent" is not a useful term. Some inventions are implemented using embedded processors controlled by software, for example most aspects of mobile phones. Other inventions relate to CPU task schedulling and others relate to organising and displaying financial information. All of these involve software but are not equivalent in terms of their engineeringness. The debate about "software patents" is so much wind without some appreciation of the different sorts of invention that may involve software and the different economic considerations. Patents are about encouraging economic development by offering the hope of protection for investment and alway have been. They are not primarily about some guy in his garage.
The analysis needs to consider the economics of R&D and cost/benefit for consumers. At present, the line is draw on the boundary of "hard engineering" and excludes business methods and inventions that relate only to the way something is coded.</RANT>
Ok then, I'll combine all three together: I'll use those my accounts that doesn't have any of my personal ID *AND* I'll publish the code under GPL on Sourceforge/Savanna-like servers colocated in software-patent-free countries. So now what?
Firstly, there is the issue of identifying you. This may or may not be possible. This just a matter of getting evidence rather than legal principle. Someone may even grass you up.
Secondly, you cannot legally relase the code under the GPL, or any other licence, into jurisdiction where a relevant patent is in force.
Thirdly, if someone downloads the software into a jurisdiction where there is a relevant patent, you risk being used as an importer or various other more subtle grounds which vary from country to country.
In some circumstances, you can be sued in one country for infringement of a patent in another.
Yes, that's right, I forgot to mention in my previous post: US is the country with the worst patent laws. In many other countries software (code and algorithms) is math and as such it cannot be patented. File and message formats as well as protocols are languages and as such they cannot be patented.
Your understanding of what is and is not patentable outside the US is a little naive. Software isn't patentable as such. However, if an invention has a technical character, which can include encryption and data compression, the fact that it will be put into practice using software does not prevent it being patented. Another example of a patentable software-implemented invention would be a new CPU scheduling method.
It is not reasonable to simply ban any patent for inventions relating to software because microprocessors are the fundamental buildng block of virtually all modern control systems from DVD players to mobile phones to chemical plants.
1. If I would publish the code under GPL and then people will improve and enhance it, and then they will fork ii, still under GPL - what can you do about it?
Sue you for all the lost sales of my product arising from your distribution of your infringing version. I may not get much but it sure as hell is going to make your life a misery - pour encourager les autres
2. Sourceforge doesn't require from me any legal ID: I can host the project under any name I want.
I'll sue SourceForge instead
3. Americans still forget that Internet is not a WAN of USA - it's international. If people around the world will publish the implemetation, that is in infringement of US patent, what is a patent holder going to do about it? Send US Army there?
Fancy that, I'm a big corporation and I've got a patent in your country too. Alternatively, be very careful about where you go on vacation. IP lawyers use private detectives.
Remember PGP? Well, there was no any infringement of the pattented process, but the way how the code has left USA was certainly a violation of US export rules. And what has US govt done about to prevent, stop or revert it?
I just want you to stop and give me money. The US government wanted to prevent it happening in the first place.
The situation with GSM is that the main player cross-licence each other which reduces the net licence fees per handset/base station/etc. If you are not already in there with a stack of patents, the licences required present a significant barrier to entry into the market.
The nature of a mobile phone system means that effective implementation involves careful hardware design as much as software design. Consequently, you do not generally find people knocking up GSM handsets as a hobby. Also, as transmitting apparatus, government type approval would required and would cost both arms and both legs.
The web on the other hand, can be fully implemented, not necessarily optimally, using commodity hardware. Consequently, the hobbyist or small business person can imnplement any protocol with an investment of primarily time rather than cash. Thus, the implementation of web technologies and GSM are economically distinct.
Put another way, GSM has these characteristics:
high cost of hardware development
patent cross-licences provide relatively cheap, in royalty terms, access to market for companies leading development (compensation for doing the development)
patent licences provide high cost of entry to the market for companies not leading development.
The Web at present has the following characteristics:
low hardware costs (remember we are talking about the Web not the Internet as a whole)
absence of need for non-free patent licences encourages entry into market, leading to expansion of market for client and server software to the potential benefit of all
copyright provides protection against actual copying of the software as opposed to the protocols that the software uses to communicate
the presence of patents could be used to enforce standards compliance, i.e. licence only free/available where standard is implemented correctly
These differences do not dictate the policy that W3C should follow. However, for the commercial players, the issue is whether their propects are better with unrestricted competition in an expanding market or restricted competition in a more limited market. For W3C, presumably, this issue is a mixture of what would provide the most effective web and some philosophical goals.
This is only possible in the US and then only because of a quirk in the definition of prior art in 35 USC 102 which excludes use outside of the US from the prior art. This means that if a description of traditional remedy from outside the US has not been described in a printed publication, it can be patented in the US. Direct taking of the traditional remedy would fall foul of the patent application not being made by the inventor. However, if the claimed invention is a variant of the traditional remedy, the fact of the foreign prior use cannot be used as the basis for rejection of the patent application on the grounds of obviousness.
Most patent systems define the prior art as what was available to the public (i.e. anyone, anywhere) before the filing date of the patent application or the date of invention. In such a case the use of the traditional remedy would render unpatentable any obvious variants or developments of it.
OK here are some real numbers for a US-originating European patent application with 20 pages of text and 20 claims (somewhat on the low side when compared with US practice but multiple dependencies are positively encouraged in Europe not penalised as in the USPTO)
Filing: $2200
Examination: $1800
Designation fees* (all states): $800
Dealing with objections by Examiner: $4000
Maintenance fees (say): $2000
Grant and validation:
United Kingdom: $300
Austria: $1700
Belgium: $900
Cyprus: $1700
Denmark: $3600
Finland: $3300
France: $2300
Germany: $2500
Greece: $2250
Ireland: $700
Italy: $2100
Latvia: $1600
Lithuania: $2100
Luxembourg: $700
Monaco: $750
Netherlands: $2300
Portugal: $2700
Romania: $1500
Slovenia: $2000
Spain: $2000
Sweden: $3800
Switzerland: $1000
Turkey: $1900
Sub-Total: $43,700
Official Grant + Printing Fees: $1500 Grand Total: $56,000 (plus US patent attorney's time and mark up.
Most of the validation costs are for translations and some countries appear low because they share a language with another country, e.g. French is used in France, Switzerland and Belgium
If you are not a pharmaceutical company, you would probably be looking for patents in the UK, France, Germany and perhaps Scandinavian countries, Italy, Spain and the Netherlands. Assuming the UK, France and Germany would give a cost per country of about $5200 per country.
Honestly now, you have never watched any BBC output?
But you benefit from the second order effect where the commercial channels are driven to keep there standards up, to some degree, to avoid losing viewers to the Beeb. Look at the relative viewing figures for the World Cup (I don't care whether you like football.)
If a patentee issues software, implementing the patented invention, under the GPL. There must be an implicit patent licence that runs in parallel with the GPL because without the parallel patent licence no one can take advantage of the GPL. It has to be assumed that when the software was issued under the GPL, it was intended that people would be able to use the software as envisioned in the GPL.
However, the patent licence would not cover independently created software implementing the invention. This has two consequences.
Firstly proprietory software would infringe the patent, possibly a good thing.
Secondly, alternative, independent open source implementations would infringe the patent and require a licence. This would enable the patentee to say to Linus, if you want this feature in the kernel, you have to use our code or a derivative thereof, irrespective of whether another implementation was better.
The features of an independent claim, e.g. claim 1, are ANDed. If claim 1 turns out to be bad, you AND in the features of one or more dependent claims, e.g. claim 2, which refer back to the independent claim.
Dependent claims, i.e. those that refer to preceding claim can be thought of as fallback positions.
Patents must be held by individuals, not corporate entities
The ecomonics of building factories means that corporations are going to want to have the exclusive benefit of a patent, even if they do not own it as such. What would happen if an engineer changed jobs, would the patent rights go with the engineer and, if so, why would anyone pay engineers to invent?
Only physical objects and processes may be patented.
(Corrolary) No patent shall be granted for algorithms or business processes
Naturally occuring results of processes may not be patented (ex: DNA)
You are in the mainstream with these. The US is pretty much out on its own in the degree to which it allows patents for these. See for example Art. 52 European Patent Office and this mauling of Dell by the UK Patent Office.
A working implementation of the patented process must be provided (upon request of USPTO)
Not really possible if the patent application is for a nuclear power station. However, part of the patent deal is that the description is sufficient to teach other skilled people how to work the invention. If the Examiner is suspicious he should raise an objection on this ground but there may be other ways of proving the invention, e.g. using computer models.
The USPTO must conduct a good faith search for any prior art
I think that this is a little hard on the USPTO, which is pretty robust in its searches compared with some other patent offices. If there is a quality issue, it will be down to resources and individual staff competency rather than bad faith.
You're being a little parochial, the original issue was the improvement of the system. I pointed out that publication of applications went some way to providing the public review. It is irrelevant how the US currently deals with this issue. The point is that there are models which provide an improvement and, although not perfect because of the AIPA exclusion, even the US is moving the the right direction.
It has always been possible to put prior art before the USPTO, although difficult to identify the case to which it relates, and the examiner is bound by a public policy requirement not to knowingly grant bad patents to take relevant prior art into account if it comes to his attention.
Re: Public Review
on
Fair IP Laws?
·
· Score: 3, Insightful
Once a patent application is published, anyone can send prior art to the relevant patent office. This is one of the reasons that patent applications are generally published.
Slashdot Mirror
... and the UK and Germany and France and Spain and Finland and Sweden and Greece and Denmark and Luxembourg and Belgium and Portugal and the Netherlands and Ireland and more coming soon
This is why he is safe in Italy
What was published before 17 Oct 1993? Are copies available?
On the one hand we have the EPC and the Guidelines which define the process and how the Examiner's should work. The process itself is followed in the sense that the EPO obeys the deadlines in the EPC and issues the correct official communications, apart from the occasional balls up. The issue is whether examiners are letting through cases they shouldn't.
I cannot think of an example, speaking as a patent attorney, of a case where the examiner seemed to allow a case just to get it off their desks. If we only consider the patentability of software as such, dodgy grants are more likely to be the result of a lack of relevant technical knowledge making arguments regarding the presence of technical effect seem more persuasive than they should. Also, since examination is an administrative not a legal process, the applicant tends to be given the benefit of the doubt.
It is now very easy to monitor European patent application publications. Anyone with an interest in a software product should set some time aside to look at the newly published applications. If any are not new or obvious, the prior art that shows this can be sent to the EPO, anonymously even, with a few paragraphs explaining the relevance of the prior art. If the invention seems to be just software, you can write to the EPO and explain why the invention merely uses software to solve a software problem, i.e. why it does not make a technical contribution.
Even if the examiner takes no action, the observations can have knock-on effects. For example, the patentee may not be allowed to amend the patent later to overcome your prior art because they allowed the patent to be granted knowing it to be bad. Prior art which is also relevant prior art for a corresponding US application will need to be notified to the USPTO by the applicant.
It is unlikely in the extreme that patents for software-implemented inventions which are technical in character, i.e. not-software as such, will be excluded from patentability. At one extreme, there are patent for televisions that employ software to process the television signals (patentable). At the nother extreme there are content management frameworks for web sites (sticking my neck out - unpatentable unless there is some effect such as a reduction in resource requirments for the same performance). In the middle are data compression and transmission systems and methods. Compressing video data can be performed using software and a PC or using custom hardware. If the custom hardware is patentable, what justification is there for not granting a patent covering the software + PC implementation? They both represent systems that operate according to an underlying algorithm.
There are other practical problems that are highlighted by cases where hardware and software implementations are possible. If a patent application describes a system comprising hardware than converts signal A into signal B and hardware for convering signal B into signal C, the main claim might be:-
A signal converter comprising means for convering a first signal into a second signal and means for converting the second signal into a third signal.
In Europe, this claim would be infringed by a hardware only device, software + a PC and by a hybrid embedded processor device. It would be difficult to frame a law that distinguished between these three examples.
So I am certain the EU will give the EPO the right to award software patents. The only way to stop this is by periodically checking what [provisional patents] applications are [awarded] published and bombard the EPO with negative comments and prior arts.
This is why applications are published and why the EPC has specific provision (Art. 115) for this. There are no fees for filing third party observations.
http://annual-report.european-patent-office.org/re port/site/en/chapter_6_financial_report.html
Reasons why Anonymous Coward is wrong:-
Almost true, it receives a cut of patent renewals fees from the contracting states (Art. 39 EPC).
Nonsense.
Irrelevant as long as its running costs are covered.
Unsubstantiated fantasy
But no shareholders so what reason is there for the EPO to maximise profits?
The way it works at the EPO is the following: Someone submits something to get a patent. If it is not completely stupid, a provisional patent is awarded, even if obvious prior art exists!!! If there has been no negative comments after one year, the patent is awarded.
WRONG - the poster was not only an Anonymous Coward, he is an anonymous know nothing.
Without the word "European", the headline is ambiguous because is could be suggesting that the EU is moving towards harmonisation with the US system (although hell will freeze over first).
Still feel clever?
This is all making a mountain out of a molehill. The situation under the European Patent Convention, which set the patentability criteria for the EPO and the national patent systems of the EU and other European states, is currently that a technical contribution is required for a software-implemented invention to be patentable. Its just that the EPC doesn't actually use these words.
The reasoning that leads from the wording of the EPC to the requirement for a technical contribution is summarised here.
<RANT> "Software Patent" is not a useful term. Some inventions are implemented using embedded processors controlled by software, for example most aspects of mobile phones. Other inventions relate to CPU task schedulling and others relate to organising and displaying financial information. All of these involve software but are not equivalent in terms of their engineeringness. The debate about "software patents" is so much wind without some appreciation of the different sorts of invention that may involve software and the different economic considerations. Patents are about encouraging economic development by offering the hope of protection for investment and alway have been. They are not primarily about some guy in his garage.
The analysis needs to consider the economics of R&D and cost/benefit for consumers. At present, the line is draw on the boundary of "hard engineering" and excludes business methods and inventions that relate only to the way something is coded.</RANT>
Umm, crumbs in difficult-to-remove places. I bet her fans will be wanting to lap this up.
Firstly, there is the issue of identifying you. This may or may not be possible. This just a matter of getting evidence rather than legal principle. Someone may even grass you up.
Secondly, you cannot legally relase the code under the GPL, or any other licence, into jurisdiction where a relevant patent is in force.
Thirdly, if someone downloads the software into a jurisdiction where there is a relevant patent, you risk being used as an importer or various other more subtle grounds which vary from country to country.
In some circumstances, you can be sued in one country for infringement of a patent in another.
Your understanding of what is and is not patentable outside the US is a little naive. Software isn't patentable as such. However, if an invention has a technical character, which can include encryption and data compression, the fact that it will be put into practice using software does not prevent it being patented. Another example of a patentable software-implemented invention would be a new CPU scheduling method.
It is not reasonable to simply ban any patent for inventions relating to software because microprocessors are the fundamental buildng block of virtually all modern control systems from DVD players to mobile phones to chemical plants.
Sue you for all the lost sales of my product arising from your distribution of your infringing version. I may not get much but it sure as hell is going to make your life a misery - pour encourager les autres
I'll sue SourceForge instead
Fancy that, I'm a big corporation and I've got a patent in your country too. Alternatively, be very careful about where you go on vacation. IP lawyers use private detectives.
I just want you to stop and give me money. The US government wanted to prevent it happening in the first place.
The situation with GSM is that the main player cross-licence each other which reduces the net licence fees per handset/base station/etc. If you are not already in there with a stack of patents, the licences required present a significant barrier to entry into the market.
The nature of a mobile phone system means that effective implementation involves careful hardware design as much as software design. Consequently, you do not generally find people knocking up GSM handsets as a hobby. Also, as transmitting apparatus, government type approval would required and would cost both arms and both legs.
The web on the other hand, can be fully implemented, not necessarily optimally, using commodity hardware. Consequently, the hobbyist or small business person can imnplement any protocol with an investment of primarily time rather than cash. Thus, the implementation of web technologies and GSM are economically distinct.
Put another way, GSM has these characteristics:
- high cost of hardware development
- patent cross-licences provide relatively cheap, in royalty terms, access to market for companies leading development (compensation for doing the development)
- patent licences provide high cost of entry to the market for companies not leading development.
The Web at present has the following characteristics:These differences do not dictate the policy that W3C should follow. However, for the commercial players, the issue is whether their propects are better with unrestricted competition in an expanding market or restricted competition in a more limited market. For W3C, presumably, this issue is a mixture of what would provide the most effective web and some philosophical goals.
This is only possible in the US and then only because of a quirk in the definition of prior art in 35 USC 102 which excludes use outside of the US from the prior art. This means that if a description of traditional remedy from outside the US has not been described in a printed publication, it can be patented in the US. Direct taking of the traditional remedy would fall foul of the patent application not being made by the inventor. However, if the claimed invention is a variant of the traditional remedy, the fact of the foreign prior use cannot be used as the basis for rejection of the patent application on the grounds of obviousness.
Most patent systems define the prior art as what was available to the public (i.e. anyone, anywhere) before the filing date of the patent application or the date of invention. In such a case the use of the traditional remedy would render unpatentable any obvious variants or developments of it.
OK here are some real numbers for a US-originating European patent application with 20 pages of text and 20 claims (somewhat on the low side when compared with US practice but multiple dependencies are positively encouraged in Europe not penalised as in the USPTO)
Filing: $2200
Examination: $1800
Designation fees* (all states): $800
Dealing with objections by Examiner: $4000
Maintenance fees (say): $2000
Grant and validation:
- United Kingdom: $300
- Austria: $1700
- Belgium: $900
- Cyprus: $1700
- Denmark: $3600
- Finland: $3300
- France: $2300
- Germany: $2500
- Greece: $2250
- Ireland: $700
- Italy: $2100
- Latvia: $1600
- Lithuania: $2100
- Luxembourg: $700
- Monaco: $750
- Netherlands: $2300
- Portugal: $2700
- Romania: $1500
- Slovenia: $2000
- Spain: $2000
- Sweden: $3800
- Switzerland: $1000
- Turkey: $1900
Sub-Total: $43,700Official Grant + Printing Fees: $1500
Grand Total: $56,000 (plus US patent attorney's time and mark up.
Most of the validation costs are for translations and some countries appear low because they share a language with another country, e.g. French is used in France, Switzerland and Belgium
If you are not a pharmaceutical company, you would probably be looking for patents in the UK, France, Germany and perhaps Scandinavian countries, Italy, Spain and the Netherlands. Assuming the UK, France and Germany would give a cost per country of about $5200 per country.
* seven buys all states
The London Transport Routemaster bus. Still in service after 49 years.
The difference, since the BBC TV version, is that we now have the technology to produce these images.
Honestly now, you have never watched any BBC output?
But you benefit from the second order effect where the commercial channels are driven to keep there standards up, to some degree, to avoid losing viewers to the Beeb. Look at the relative viewing figures for the World Cup (I don't care whether you like football.)
It just makes me want to hug the TV licence fee ~$160 per year. This gives me 8 TV channels with no ad breaks - whole uninterrupted movies.
Just imagine a whole evening's viewing without anything allegedly washing whiter.
The UK TV licence, you can't justify it but by gum it works!
If a patentee issues software, implementing the patented invention, under the GPL. There must be an implicit patent licence that runs in parallel with the GPL because without the parallel patent licence no one can take advantage of the GPL. It has to be assumed that when the software was issued under the GPL, it was intended that people would be able to use the software as envisioned in the GPL.
However, the patent licence would not cover independently created software implementing the invention. This has two consequences.
Firstly proprietory software would infringe the patent, possibly a good thing.
Secondly, alternative, independent open source implementations would infringe the patent and require a licence. This would enable the patentee to say to Linus, if you want this feature in the kernel, you have to use our code or a derivative thereof, irrespective of whether another implementation was better.
The features of an independent claim, e.g. claim 1, are ANDed. If claim 1 turns out to be bad, you AND in the features of one or more dependent claims, e.g. claim 2, which refer back to the independent claim.
Dependent claims, i.e. those that refer to preceding claim can be thought of as fallback positions.
Patents must be held by individuals, not corporate entities
The ecomonics of building factories means that corporations are going to want to have the exclusive benefit of a patent, even if they do not own it as such. What would happen if an engineer changed jobs, would the patent rights go with the engineer and, if so, why would anyone pay engineers to invent?
Naturally occuring results of processes may not be patented (ex: DNA)
You are in the mainstream with these. The US is pretty much out on its own in the degree to which it allows patents for these. See for example Art. 52 European Patent Office and this mauling of Dell by the UK Patent Office.
A working implementation of the patented process must be provided (upon request of USPTO)
Not really possible if the patent application is for a nuclear power station. However, part of the patent deal is that the description is sufficient to teach other skilled people how to work the invention. If the Examiner is suspicious he should raise an objection on this ground but there may be other ways of proving the invention, e.g. using computer models.
The USPTO must conduct a good faith search for any prior art
I think that this is a little hard on the USPTO, which is pretty robust in its searches compared with some other patent offices. If there is a quality issue, it will be down to resources and individual staff competency rather than bad faith.
You're being a little parochial, the original issue was the improvement of the system. I pointed out that publication of applications went some way to providing the public review. It is irrelevant how the US currently deals with this issue. The point is that there are models which provide an improvement and, although not perfect because of the AIPA exclusion, even the US is moving the the right direction.
It has always been possible to put prior art before the USPTO, although difficult to identify the case to which it relates, and the examiner is bound by a public policy requirement not to knowingly grant bad patents to take relevant prior art into account if it comes to his attention.
Once a patent application is published, anyone can send prior art to the relevant patent office. This is one of the reasons that patent applications are generally published.
However, you've got to provide the evidence.