Quality of TH reports

[bogeybum]

The other week I posted a response to this article and the post got deleted and the comments locked. Now admittedly the article was an old one. But the point is that I got to the article via a web search and the point of my response was that the information in the article is misleading or at worst wrong. As far as I am aware the details of PCIe have never changed so it's not as if this article was correct when it was published.

Two paragraphs quoted from the article:

"PCIe slots come in different physical configurations: x1, x4, x8, x16, x32. The number after the x tells you how many lanes (how data travels to and from the PCIe card) that PCIe slot has. A PCIe x1 slot has one lane and can move data at one bit per cycle. A PCIe x2 slot has two lanes and can move data at two bits per cycle (and so on)."

"You can insert a PCIe x1 card into a PCIe x16 slot, but that card will receive less bandwidth. Similarly, you can insert a PCIe x8 card into a PCIe x4 slot, but it’ll only work with half the bandwidth compared to if it was in a PCIe x8 slot."

If you know anything about PCIe you'll see what's wrong in both of these quotes.

My question is why was my post deleted and why hasn't the article been updated or removed? It makes me wonder how reliable the information I read at TH is now. I also wonder what this sort of censorship is hoping to gain?

 

[Colif]

its policy to delete replies to posts made older than 6 months ago, its not censorship, its just forum management. Its nothing personal, often people reply to old threads asking for help but those in thread may not have been here in a long time, and other people won't look there, so we generally suggest people post a new thread (like you have). If its a comment on an article, the time frame was probably the reason.

I don't have any control over items on the front page.

 

[bogeybum]

Hi Colif thanks for your reply which makes sense. Do you know about the process for updating factual errors in articles?

 

[Rogue Leader]

I looked at both the article and your reply.

While in some cases some motherboards will utilize a physical x16 slot but only have the slot set up for x8 (or x4), this is described specifically in the manual of the device. This is a very minor alternate situation and I can see why it was not described that way in the article as it could cause confusion.

Your reply was deleted because it was both a response to an old article, and the issue you brought up was hardly a factual error but a minor semantic issue.

 

[bogeybum]

Thank you Rogue Leader for your response.

I read your reply as "TH can write anything that may be verging on inaccurate - but RTFM to be sure". I have have a higher opinion of TH for the last 20 years or so than that warrants. However yes, arguably it could be a semantic issue.

Do you think the second quote is a semantic issue? Do you think a x1 will receive less bandwidth in a x16 slot? Documentation in a technical environment will always lead to confusion and doubt unless its concise and exact so semanatics are still very relevant, minor or not.

 

[Rogue Leader]

Thank you Rogue Leader for your response.

I read your reply as "TH can write anything that may be verging on inaccurate - but RTFM to be sure". I have have a higher opinion of TH for the last 20 years or so than that warrants. However yes, arguably it could be a semantic issue.

Do you think the second quote is a semantic issue? Do you think a x1 will receive less bandwidth in a x16 slot? Documentation in a technical environment will always lead to confusion and doubt unless its concise and exact so semanatics are still very relevant, minor or not.

It IS a semantic issue. if you insert an x1 card into an x16 slot it will run at x1. Putting an x1 card into an x16 slot does not automatically magically increase its bandwidth to that of the slot its used in. Heck an x1 card doesn't even have the contact pins to reach the whole slot.

 

[hotaru.hino]

Do you think the second quote is a semantic issue? Do you think a x1 will receive less bandwidth in a x16 slot? Documentation in a technical environment will always lead to confusion and doubt unless its concise and exact so semanatics are still very relevant, minor or not.

The sentence can be interpreted in several ways. If taken at face value, it implies an x1 card receives less bandwidth in a x16 slot than in an x1 slot. If it were a PCIe 1.x card, that would mean it'd receive less than 250MB/s. However, it could also be interpreted that the slot doesn't provide the full bandwidth to an x1 card, so an x1 card receives less bandwidth in that sense.

 

[bogeybum]

As you explain hotaru.hino it can be interpreted differently. I think the real issue is the original author of the article didn't fully understand how PCIe works and in particular the distinction between the physical and electrical properties of the PCIe and their interpretation has lead to them writing loose facts which can and are being interpreted differently. This is what for me makes the article so questionable, especially for technical journalism. I wonder how much grief that article has caused PCIe novices?

 

[Rogue Leader]

As you explain hotaru.hino it can be interpreted differently. I think the real issue is the original author of the article didn't fully understand how PCIe works and in particular the distinction between the physical and electrical properties of the PCIe and their interpretation has lead to them writing loose facts which can and are being interpreted differently. This is what for me makes the article so questionable, especially for technical journalism. I wonder how much grief that article has caused PCIe novices?

I disagree with your assumption that the author does not understand PCIe. Your nitpicking of semantics does not conflate to misinformation or an error that must be corrected.

We understand your position, I will make the suggestion that some people had difficulty understanding the wording, but I believe and I'm sure the author will believe the issue is on the readers side.