Investigative journalism
- Thread starter Reverend
- Start date
- Status
Yes, if you think the charts nVidia gives you are wrong (and you have information that suggests that they are pretty significantly wrong) don't print the charts. Or is there some deal where you can only write an article based on the information they give you if you include the charts unaltered?
Its a matter of he said/she said.
Why not put all the information out there and mark it as so? That way you don't get accused of being biased? (hahaha)
Why not put all the information out there and mark it as so? That way you don't get accused of being biased? (hahaha)
Because I want NVIDIA to know they don't know all there is to know about the R300 and that their "paper" is incorrect in specific instances? Because I want other websites that may be similarly briefed by NVIDIA based on the same papers I have to know that NVIDIA may be incorrect/mistaken in the cases I highlighted?GetStuff said:
Doomtrooper
Veteran
What I think people want to see..
http://www.tech-report.com/etc/2002q3/nextgen-gpus/index.x?pg=1
The information provided to Reverend was from a competing IHV that had a slide show presentation on how their solution is better, I really don't think Nvidia is going to say alot of good things about the R300..do you ??
I think the Tech-Report article did a excellent job of being neutral.
http://www.tech-report.com/etc/2002q3/nextgen-gpus/index.x?pg=1
The information provided to Reverend was from a competing IHV that had a slide show presentation on how their solution is better, I really don't think Nvidia is going to say alot of good things about the R300..do you ??
I think the Tech-Report article did a excellent job of being neutral.
Joe DeFuria
Legend
Sure. Not much else you can do. Of course, every single web-site will end up posting the same info, so if you don't add at least your own commentary, there won't be anything to differentiate your article from others, and it becomes therefore less valuable.
Again though, if part of the NDA'd materials includes comparisons to competitor products as a means of defining their own product, then extra care should be taken to ensure everything is factually correct.
IHV's can do comparisons in one of two ways: compare to their previous gen product, or compare to the competitor's product. If they compare to their own gen, there's not much investigating you need to do. If they compare to the competition (especially in this case where the competition product is not available yet to the public), then I would like to see "official" comment from the competitor IHV...or at the very least a statement that you requested official comment and will post the response when / if given.
Doomtrooper
Veteran
Qroach said:
Thats nice...
If you fail to see the difference in the Tech Report article vs. the NV30 Tech Voodooextreme article you are blind.
Tech report Author:
That is the key here as Reverend with his insults yesterday was keen to point out, he has contacts @ Nvidia only and is proud of it.
Thats fine and if the Nv30 article didn't specifically target the R300 in almost every page I wouldn't have a issue...but when they talk about superiority in certain instances it would have been nice to find a reply from the other IHV.
Reverends quote:
Since the article is about NV30..thats cool only hearing from one IHV..but when they start doing 'mine is better than yours charts' the other IHV should be involved IMO.
Doomtrooper said:
The tech-report article was rather long winded IMO. I found around 90% of the material to be irrelevant to NV30 vs R300. Granted, there isn't that much info available (especially on NV30), but that doesn't make it good journalism to fluff up the article with a bunch of unrelated information, does it?
It's much easier to sound neutral (or boring?) when all your doing is reciting the short history of 3d graphics....
I found Rev's article a much better read...
Can we get away from talking about specifics and talk about the general subject Rev is trying to get at, please.
DT,
The only one here that's blind is yourself DT. Here's what i saw in revs article.
1. It's an article ABOUT a briefing on CineFX capable hardware.
2. Nvidia provided charts showing difference between their planned hardware and other competing hardware out there.
3. Some of the information was wrong and Rev got corrections from someone that WORKS at ATI.
4. Rev posted what he had been told (truth) regarding these features below each area that was not correct.
Looks to me like Rev actually DID his best to get correct information by going to someone that WORKS at ATI. The only differnce here is that he didn't get the information directly from ATI PR, but he DID get the information from a engineer at ATI. What's wrong with that? Nothing. Get your eyes checked.
Really, so I guess next time ATI use little charts showing how much faster the 9700 is ('mine is better than yours charts') compared to the Geforce 4, they should get Nvidia invovled? BAH, get real.
Rev made a fine article. Stop complaining.
The only one here that's blind is yourself DT. Here's what i saw in revs article.
1. It's an article ABOUT a briefing on CineFX capable hardware.
2. Nvidia provided charts showing difference between their planned hardware and other competing hardware out there.
3. Some of the information was wrong and Rev got corrections from someone that WORKS at ATI.
4. Rev posted what he had been told (truth) regarding these features below each area that was not correct.
Looks to me like Rev actually DID his best to get correct information by going to someone that WORKS at ATI. The only differnce here is that he didn't get the information directly from ATI PR, but he DID get the information from a engineer at ATI. What's wrong with that? Nothing. Get your eyes checked.
Really, so I guess next time ATI use little charts showing how much faster the 9700 is ('mine is better than yours charts') compared to the Geforce 4, they should get Nvidia invovled? BAH, get real.
As far as I'm concerned you deserved those "insults". You're acting way out of line.
Rev made a fine article. Stop complaining.
Fuz said:
LOL, nice try Fuz!
IMO Rev did the best job he could do with the limited resources provided. At every juncture NVIDIA claimed something Rev though was wrong he refuted it in black and white. That takes a lot of guts in itself as you risk pissing off your contacts.
Regarding the actual topic I believe there are plenty of good sites that do very thorough 'benchmark reviews' for example anandtech.com and tomshardware.com.
The next wave of articles come from other sites such as beyond3d.com and techreport.com where they try to make sense of some of the numbers instead of posting such comments like 'here we see the GF4 is faster than the Radeon 8500 in Quake 3' --> benchmark follows... etc etc etc
When you see a review in a magazine you do NOT see 7 pages of benchmarks with one or two captions. You see a lot of editorial, some background information and hypothesis and then proving the hypothesis with maybe one or two benchmarks. With online media you cannot do this. But there has to be a VARIETY of benchmarks, one of the best sites for this is simhq.com and aceshardware.com. Otherwise it seems like the reviewer played with the 'thing' being reviewed for a couple of days and ran some benchmarks. These are just more padded previews to me. If you are going to review something then you need to live with it for a while so that you notice any interesting nuances. It should be like a lasting relationship, not a one night stand.
Now the difference between an online journalist and one of print journalism is simply, accountability, objectivity and the ability to form that long lasting realtionship.
It is impossible for a human to be unbiased as the information he selects is a limiting choice and therefore other information has to be discarded. This does not imply that a journalist has a motive against company 'x' (like some people blindly believe). With finite, (time & money, information available & your own intelligence), resources you have to be biased in what you select. THen of course if you are in fact biased towards a certain company then that compunds things even more. And the printed media does this also. Easy explanation for this is: "A Network Administrator will not get fired for choosing Intel."
I have read a few UK magazines that have a much bigger following than the quivalent in the US and some are very dismissive of anything other than NVIDIA with regards to graphics cards. You wont get fired for recommending the status quo. But NVIDIA have built up this trust themselves and this image. It is easy to shatter also with one big mistake all your hard earned credibility goes down the tube.
Again Rev, that was a good article you wrote and I hope to read more in the future - but that is not the subject of this topic.
Qroach, Rev:
I don't necessarily agree with doomtrooper, but Rev asked for comments and he got them, so don't bitch if he's said things you don't like. Try to extract something useful out of it, and if you can't then ignore it. This constant bickering and insults is getting really old, and I think doomtrooper (while being onesided and emotional at times) offered some useful critisism and you pretty much told him to go to hell.
Having said that, I tend to agree with Kristof. It's really about adding value. Think of your review/preview/interview like you would an academic paper. If you are writing a preview or an interview, it's likely that you are getting most of your information from one source. On the other hand though, if you want your article to be interesting, you'll need to make comments and speculation on this information, and that implies that you'll be drawing on information from other sources. Research this information well, and even better provide a bibliography and cite (in the article) your sources for any background information you provide. If you are going to comment on conflicting information, be prepared to back it up. Why don't you think nvidia's information is correct? How does this affect their stance? Is their information reliable? Why or why not?
Citing provides you with an indication of how well you've researched your topic, and wether or not you have more work to do. If all of your information (and arguements) comes from nvidia (or anyone for that matter), it's going to be biased, and thus your article will likely be biased, even if you yourself are not. If there is no way you can get information from other sources, then be prepared to explain why this is the case, and why you feel that your article is still valid given this fact. Try to examine your arguements and find inconsistencies and lack of source. Have your article peer-reviewed before you publish it. Be prepared to defend yourself, because it's a lot easier to shoot down claims than it is to defend them.
If you can provide varying sources, strong arguments, and an understanding of the topic, it will be much harder for people to find valid flaws in your articles.
Nite_Hawk
I don't necessarily agree with doomtrooper, but Rev asked for comments and he got them, so don't bitch if he's said things you don't like. Try to extract something useful out of it, and if you can't then ignore it. This constant bickering and insults is getting really old, and I think doomtrooper (while being onesided and emotional at times) offered some useful critisism and you pretty much told him to go to hell.
Having said that, I tend to agree with Kristof. It's really about adding value. Think of your review/preview/interview like you would an academic paper. If you are writing a preview or an interview, it's likely that you are getting most of your information from one source. On the other hand though, if you want your article to be interesting, you'll need to make comments and speculation on this information, and that implies that you'll be drawing on information from other sources. Research this information well, and even better provide a bibliography and cite (in the article) your sources for any background information you provide. If you are going to comment on conflicting information, be prepared to back it up. Why don't you think nvidia's information is correct? How does this affect their stance? Is their information reliable? Why or why not?
Citing provides you with an indication of how well you've researched your topic, and wether or not you have more work to do. If all of your information (and arguements) comes from nvidia (or anyone for that matter), it's going to be biased, and thus your article will likely be biased, even if you yourself are not. If there is no way you can get information from other sources, then be prepared to explain why this is the case, and why you feel that your article is still valid given this fact. Try to examine your arguements and find inconsistencies and lack of source. Have your article peer-reviewed before you publish it. Be prepared to defend yourself, because it's a lot easier to shoot down claims than it is to defend them.
If you can provide varying sources, strong arguments, and an understanding of the topic, it will be much harder for people to find valid flaws in your articles.
Nite_Hawk
Joe DeFuria
Legend
Well said, Nite_Hawk!
The problem with posting large feature charts from NVIDIA is that not everyone is going to actually READ the article. Some will just skim through it, see the charts and be like, "Oh, that is how it must be. R300 sucks." If you don't believe something to be true at all, don't post the charts. Comment on it, but posting the charts is bad IMO.
It is interesting that you comment on Matrox's FAA. Nobody has really covered it at all. Why? A large reason is that matrox won't even spill the beans (even a little), but frankly I'm surprised on the stunning lack of info on even some of the basics behind the algorithm (not to pat my own back, but to show how sad the coverage is, an article I wrote over a year ago has more info on it than anything I've read to date). Could a review work at it and try to figure out what they are doing? Yup. it takes time and research, but it is possible to come up some good theories on it. <Plug myself time> If you subscribe to InovaPC, watch for such an article in the next few months.</Plug myself time>
It is interesting that you comment on Matrox's FAA. Nobody has really covered it at all. Why? A large reason is that matrox won't even spill the beans (even a little), but frankly I'm surprised on the stunning lack of info on even some of the basics behind the algorithm (not to pat my own back, but to show how sad the coverage is, an article I wrote over a year ago has more info on it than anything I've read to date). Could a review work at it and try to figure out what they are doing? Yup. it takes time and research, but it is possible to come up some good theories on it. <Plug myself time> If you subscribe to InovaPC, watch for such an article in the next few months.</Plug myself time>
Dave said:
I agree, but we did discuss possible implementations right here on the board, think I gave a description of a possible algorithm right here. And I mailed Wavey a little explaining article quite a while ago which might go into our Parhelia article, if we ever get to do one
K-
Doomtrooper said:
Thanks for the link DT. I agree they did an excellent job of being neutral. Though Tech-Report's article was started for different reasons than Reverend's. When Reverend did his article it was response to NVIDIA's presentation on CineFX. I agree with others that it could have been better if he went the route of Tech-Report instead, but considering the source of the material and the focus he took there's not much more he could have added to it without more info from ATI. It would have been prefferable to have this info BEFORE the CineFX presentation.
Why he don't have better communication with ATI is something I don't have much experience with. When I was doing articles for Jon Peddie Associates' The Peddie Report I never had a problem talking to any of the vendors. Though that was mainly due to who I was employed by. Maybe if Reverend is concerned enough about it he might try to distance himself from his site and/or employer?
Anyway, I loved the article by the way. It definitely caught me up on how the programmable shader stuff got started and what it means.
Tommy McClain
My $0.02 on all this..
Rev's article contains my requirements for a piece of "journalism" which are the following:
-----------------
1) Proper identification of the source and/or purpose of the discussion.
Rev does this right on the onset with-
Source identified- an NVIDIA PR Rep. and NVIDIA PR Rep. that knows a bit about hardware.
Purpose identified. It's a walk-through of an NVIDIA PR document, with some commentary from the folks that have brought this to you.
-----------
2) Some amount of questioning or analysis versus blindly rehashing information given.
This is done in a series of "NOTE:" style inclusions by Rev, such as:
and
---------------
So, in all, I dont fault the article nor see anything wrong with it. Given the source and the purpose as *clearly* outlined, there are obviously going to be some debatable or questionable material presented, and some of it has been debated and questioned in the article.
It makes no statements of "this is the way it is through testing and in depth research" but instead it reads as "the guys at NVIDIA have such and such understanding."
I can see an argument of allowing an ATI rep. be consulted or be allowed a rebuttal, but I dont think this article is the proper place for this. Instead, a possible future article that allows ATI a rebuttal and similar comparison should be allowed.. along with a small change to this article with something like ".. and see ATI's rebuttal to this at http://xxx""
I *disagree* with the intended audience for this article, that being of the gamer that just wants to know what is what for his/her gaming experience. PR-style information rarely yields this kind of insight and is always misleading as far as value is concerned. I think if an article intended for a "gamer" were to be written, it would require the evolution as such:
1) This article.
2) A rebuttal article by obtaining interest from an ATI rep with R300 PR.
3) A B3D or VE3D "rebuttal" to both that includes a feature box like:
Feature: DX8 | DX9 | OGL 2.0 | R300 | NV3X
(insert little grid of red checkmarks here from interpolated information from both sources)
Obviously, the point of IHV PR is to stretch a definition in a way to make one's own product appear to be on top of another product. By using different definitions, one can make one's own product have checkboxes and another not. By not including DX9 or OGL, an IHV can also lay claim to feauresets that will have no applicable value to the "gamer" as well. Albeit someone using a plug-in in Maya or coding "to the metal" assembly demos might be able to get some benefit, it is meaningless to the "gamer."
Just my opinion. I see nothing wrong with the article, OR the journalism in Rev's article. It doesnt give any intentions whatsoever that the information is "the way it is from Rev" but instead "let's go over this PR document with some additional commentary from the folks that have brought it to you." Since all this information is given, along with some questioning involved, there is nothing wrong with this article that I can see.. especially for the audience of B3D/VE3D readership that mostly understands all this.
Cheers,
-Shark
Rev's article contains my requirements for a piece of "journalism" which are the following:
-----------------
1) Proper identification of the source and/or purpose of the discussion.
Rev does this right on the onset with-
Source identified- an NVIDIA PR Rep. and NVIDIA PR Rep. that knows a bit about hardware.
Purpose identified. It's a walk-through of an NVIDIA PR document, with some commentary from the folks that have brought this to you.
-----------
2) Some amount of questioning or analysis versus blindly rehashing information given.
This is done in a series of "NOTE:" style inclusions by Rev, such as:
and
---------------
So, in all, I dont fault the article nor see anything wrong with it. Given the source and the purpose as *clearly* outlined, there are obviously going to be some debatable or questionable material presented, and some of it has been debated and questioned in the article.
It makes no statements of "this is the way it is through testing and in depth research" but instead it reads as "the guys at NVIDIA have such and such understanding."
I can see an argument of allowing an ATI rep. be consulted or be allowed a rebuttal, but I dont think this article is the proper place for this. Instead, a possible future article that allows ATI a rebuttal and similar comparison should be allowed.. along with a small change to this article with something like ".. and see ATI's rebuttal to this at http://xxx""
I *disagree* with the intended audience for this article, that being of the gamer that just wants to know what is what for his/her gaming experience. PR-style information rarely yields this kind of insight and is always misleading as far as value is concerned. I think if an article intended for a "gamer" were to be written, it would require the evolution as such:
1) This article.
2) A rebuttal article by obtaining interest from an ATI rep with R300 PR.
3) A B3D or VE3D "rebuttal" to both that includes a feature box like:
Feature: DX8 | DX9 | OGL 2.0 | R300 | NV3X
(insert little grid of red checkmarks here from interpolated information from both sources)
Obviously, the point of IHV PR is to stretch a definition in a way to make one's own product appear to be on top of another product. By using different definitions, one can make one's own product have checkboxes and another not. By not including DX9 or OGL, an IHV can also lay claim to feauresets that will have no applicable value to the "gamer" as well. Albeit someone using a plug-in in Maya or coding "to the metal" assembly demos might be able to get some benefit, it is meaningless to the "gamer."
Just my opinion. I see nothing wrong with the article, OR the journalism in Rev's article. It doesnt give any intentions whatsoever that the information is "the way it is from Rev" but instead "let's go over this PR document with some additional commentary from the folks that have brought it to you." Since all this information is given, along with some questioning involved, there is nothing wrong with this article that I can see.. especially for the audience of B3D/VE3D readership that mostly understands all this.
Cheers,
-Shark
Kristof said:I agree, but we did discuss possible implementations right here on the board, think I gave a description of a possible algorithm right here. And I mailed Wavey a little explaining article quite a while ago which might go into our Parhelia article, if we ever get to do one
K-
Got a link on that thread? I must have missed that one. Also, care to forward that email to my inovapc email. I'd like to see what you came up with vs. my own stuff.
- Status
