32-bits vs 64-bits -- thoughts, issues, etc.

aragon said:
I know all that. What I meant is that they don't support open source in the manner of providing open source drivers or hardware documentation so that the community can develop open source drivers. All their "support" is in the form of closed up binaries.
Click to expand...

ahh
yah, i getcha....

I still think nvidia drivers work better on linux than the intel or ati drivers, but yah, there ARE both open and closed source binaries for ati

so i totally see what you mean.
 
>I still think nvidia drivers work better on linux than the intel or ati drivers

They suck less. Anyway I don't like the business model of nVidia with legacy drivers and so on, it's a only mess in open source operating systems.
 
to be fair the linux nvidia driver is QUITE good. ESPECIALLY for hd video

hopefully the amd64 nvidia driver will be as good when it comes out
 
wonslung said:
nvidia DOES support opensource
Click to expand...

Could you be so kind to provide a link to the documentation for their hardware. I didn't think so. How would you be able to write a professional graphical application for their hardware if you do not know specifications. You wouldn't. You just need that binary blob crap to play video games. I know, I know...

5 years old Tezro graphics station made by SGI which you can probably get on Ebay for 5K is 100 times better than anything NVidia will be able to produce in next 10 years. There are people around here who did more than just playing moronic games with their computers.

For the record NVidia binary blob drivers are huge security risks.
 
If i was playing games with my computer i wouldn't be using freebsd. If you had read the rest of the thread you'd know that we just had a slight misunderstanding, no need to get snippy.

And to be fair, my Nvidia cards do fine. I mainly use them for hd movies and 2d stuff anyways but i am not unhappy with them in the least.

and why the hell would i pay 5k for a used tezro? seems crazy.
 
wonslung said:
and why the hell would i pay 5k for a used tezro? seems crazy.
Click to expand...

In order to make special effects for the newest Star War movie I am directing which you will be able to watch in HDD using your crappy NVidia video card;)
 
wonslung said:
If i was playing games with my computer i wouldn't be using freebsd. If you had read the rest of the thread you'd know that we just had a slight misunderstanding, no need to get snippy.

And to be fair, my Nvidia cards do fine. I mainly use them for hd movies and 2d stuff anyways but i am not unhappy with them in the least.

and why the hell would i pay 5k for a used tezro? seems crazy.
Click to expand...

If I would play such games, I wouldn't even try Linux. And instead of using Windows I would buy some Playstation. I'm using open source to get rid of companies like nVidia. I wouldn't even support a blob in the kernel, but well that's just my opinion.
 
i don't see the big problem with closed source stuff if it's not horrible. I can see why it would be nice if it was completely open but at the same time i don't see it as being "wrong" or "evil" for being closed.
 
wonslung said:
i don't see the big problem with closed source stuff if it's not horrible. I can see why it would be nice if it was completely open but at the same time i don't see it as being "wrong" or "evil" for being closed.
Click to expand...

How do you know that isn't horrible if you can't see the code.
How do you know that NVidia binary blob driver is not just collecting private information about you and that NVidia people are just waiting for an opportunity to get all money from your bank accounts?

There is nothing wrong with charging for code. You want pay out for your work investment. There is nothing wrong with licensing your software under any license you like even the one like GPLv3 which expect you to become a communist before you can use their code. However, there are so many things that are wrong with the close source concept. Close source concept if fundamentally wrong!


I happen to be a mathematician. Could you imagine me stating theorem in one of my papers and then saying that the proof is closed source and I can not show it to you? That is just laughable.

There is not such a thing as close source scientific/engineering
activity. Close source in computer science is code word for lack of competence, poor work, and hidden agenda.It is as simple as that.
 
Oko said:
How do you know that isn't horrible if you can not see the code.
How do you know that NVidia binary blob driver is not just collecting private information about you and that NVidia people just wait for an opportunity to get all money from your bank accounts?

There is nothing wrong with charging for code. You want pay out for your work investment. There is nothing wrong with licensing your software under any license you like even the one like GPLv3 which expect you to become a communist before you can use their code. However, there are so many things that are wrong with the close source concept. Close source concept if fundamentally wrong!


I happen to be a mathematician. Could you imagine me stating theorem in one of my papers and than saying that the proof is closed source and I can not show it to you? That is just laughable.

There is not such a thing as close source scientific/engineering
activity. Close source in computer science is code word for lack of competence, poor work, and hidden agenda.It is as simple as that.
Click to expand...

i guess i'm just not that paranoid. If it works for what i need it to do, i am not worried. Your math analogy is an apples to oranges comparison though.


I am not a programmer, so even if i COULD see the code i'd have no idea what i was looking at and would still have to rely on another group or person to tell me it's not "horrible"

But you know what i meant when i said it's not horrible anyways, you are just trying to turn this into an argument. I am a fan of opensource but i do NOT believe all closed source projects are evil, terrible or wrong. I could just as easily make an argument for why the BSD license isn't as fair/good/right as GPL3. I don't BELIEVE that but you see what i'm saying. I personally can see valid reasons for closed source and open source projects. I applaud companies and originations that are opensource but I do not condemn them just for being closed either.

ANYWAYS we're getting WAY off topic here. It's honestly really silly to turn my comment into this. If you want to start another thread about closed source vs open source, i'll be glad to explain my position but i'm through arguing it in this thread, which is supposed to be about 32 vs 64 bit FreeBSD
 
Oko said:
There is not such a thing as close source scientific/engineering activity.
Click to expand...

I don't agree with this. In fact, unless there is a patent filed (and hence it is disclosed) many scientific and engineering activities are entirely closed. My area is in biotech instrumentation; other than the patents there's often nothing published about those instruments at all. You certainly don't get the source code to the programs that operate the instruments. Same for traditional biotech production: that is all a trade secret.

Now I grant that there is a huge academic and quasi academic community that publishes their results in the open literature, and I'm starting to do so with some interesting novelties that are not patentable. But it really is a stretch to claim that there isn't closed-source science or engineering.
 
DrJ said:
My area is in biotech instrumentation; other than the patents there's often nothing published about those instruments at all. You certainly don't get the source code to the programs that operate the instruments. Same for traditional biotech production: that is all a trade secret.
Click to expand...

I guess our opinions are influenced by the fields in which we work.

Computer assisted proof of mathematics theorem aided by a close source program or even worse non scrutinized algorithm is just unpublishable in a serious Mathematics journals. I am not saying that you will not be able to find journal which will accept such paper. I am just saying that I could not care less for such papers and journals.

Let me put more bluntly. Any biology paper which claims new results using bio-tech instruments for which you can not provide specification and code IMHO is not admissible peace of scientific work.

There is a fantastic paper on that topic by Fields Medalist Sergei Okounkov published couple years ago in Notices of American Mathematical Society.

I guess an analogy in C.S. would be a having a binary blob on your firewall machine. People are doing it but ask serious people like Theo what they think about it.

Sorry Dutch for this way off topic comment.

Oko

P.S. Doc, you also mentioned the topic of open publishing which is completely different thing. Some people, me included, believe that knowledge should be available at no cost to anybody. In particular, I am appalled by text-books which are sold for $100 out of which usually only $5 goes to author of the book who probably also prepared the book in electronic format. The price of printing is probably no more than $2-$3 which means that 90% of the price of a text book is just profit for publisher. That topic is far more controversial and is really subject to debate.
 
Oko said:
Let me put more bluntly. Any biology paper which claims new results using bio-tech instruments for which you can not provide specification and code IMHO is not admissible peace of scientific work.
Click to expand...

I wouldn't go that far. You must of course run standards and get known results, and keep the gizmo calibrated. It must also operate on sound technical principals. And the specifications are known. The code that runs them uniformly is closed. All of them. With your position there would be no publication in chemistry or biology. None.

P.S. Doc, you also mentioned the topic of open publishing which is completely different thing.
Click to expand...
I meant something a bit different. Publishing in the open literature is a term of art that means it has gone through the traditional peer-review process, and is available to the public. It counts as a public disclosure for patent purposes.

Some people, me included, believe that knowledge should be available at no cost to anybody. In particular, I am appalled by text-books which are sold for $100 out of which usually only $5 goes to author of the book who probably also prepared the book in electronic format. The price of printing is probably no more than $2-$3 which means that 90% of the price of a text book is just profit for publisher. That topic is far more controversial and is really subject to debate.
Click to expand...

Dutch, if you want to break this into another thread, that's fine with me.

Publishing is very expensive. I can assure you that 90% of the cost of a textbook is not straight profit. There is the advance to the author, the distribution costs, junking titles that don't sell, editorial review and a whole host of other expenses. It just is not that profitable a business.

It is even worse at the journals, though. Those are incredibly expensive, and they don't even pay their peer reviewers (ask me how I know). Unlike an academic institution, where this expense is buried in a Prof's salary, I essentially bear the cost of reviewing out of my own pocket. I review for journals that do not even allow me complimenary access to the underlying journal! The Royal Society is the worst in this regard.
 
DrJ said:
Those are incredibly expensive, and they don't even pay their peer reviewers (ask me how I know). Unlike an academic institution, where this expense is buried in a Prof's salary, I essentially bear the cost of reviewing out of my own pocket. I review for journals that do not even allow me complimenary access to the underlying journal! The Royal Society is the worst in this regard.
Click to expand...

Well, I do know what are you talking about. I said that open publishing is much more controversial idea;). I got $50 for reviewing 300 pages textbook. With Journals the situation in mathematics is probably better than in other fields. Of course you will get $0 for being reviewer but at least you get a copy of Journals.

My friends who are astronomers have to pay out from their grants to the publishers of some Journals to publish their accepted and per-reviewed papers. That is just the standard in their business.

I guess it really depends a lot from field to field. The cost of doing mathematics (where I work) is just buried in our salaries which are largely paid by tax payers and tuition money. It is as simple as that.
 
Oko said:
I guess it really depends a lot from field to field. The cost of doing mathematics (where I work) is just buried in our salaries which are largely paid by tax payers and tuition money. It is as simple as that.
Click to expand...

It is more that you work at an academic institution, and service (for which this counts) is part of the academic expectation. I have to eat what I kill (as it were) so if I'm not paid for something, then I am not paid. There is no salary that covers it.
 
I think we're done with the 32-bit/64-bit topic, and I believe the open/closed source dichotomy (when it comes to software, at least) has been done to death already. I'll leave this thread open for some closing arguments, gentlemen, but that'll be it ;)
 
i'm a big fan of both the 32 bit and 64 bit FreeBSD

I tend to use the 64 bit more though because of ZFS.

I think if you don't mind using linux-compat for flash, and don't need 3d video or wine 64 bit freebsd is fine.
 
Has anybody done video encoding editing on both platforms (i386 as well as AMD 64)?

Can anybody tell me if the performance difference is significant? Does the 64-bit version utilize the CPU's features better?
 
harishankar said:
Has anybody done video encoding editing on both platforms (i386 as well as AMD 64)?

Can anybody tell me if the performance difference is significant? Does the 64-bit version utilize the CPU's features better?
Click to expand...

Of course not, buy some multicore cpu and find software which utilizes it.
 
kungfujesus said:
8 ... contained the features necessary for nvidia to code a 64 bit driver.
Click to expand...

Not all of them have been incorporated yet. Ed Zander at nVidia speculates it will be a few weeks yet before that happens. Then nVidia has to complete its driver (and much of it works now).
 
DrJ said:
Good luck with that. I'd bet an alpha version of the nVidia driver is out in two months, maybe three. 3D on ATI I'd guess is a year or two off. Personally it does not matter to me, but I have found the nVidia driver to be very good.
Click to expand...
The ATI drivers might be ready sooner than you think. :)
 
You must log in or register to reply here.
Back
Top