Talk:CFexpress

Changes 2025-12

A user talk page is not the right place to discuss the content of an article.
1) {{Prose|date=April 2023}} is kept as the text still very much not flowing text and can/should be improved.
2) A product/standard history as well as key products (usually the first ones, great improvements or products that indicate industry wide adoption) are important as a standard is only worth it's existence if it is adopted by the industry (see also Theora vs. AV-1 or HD-DVD vs. BluRay - both both standards of the same type, one is used the otherone is not)
3) "going into far too much detail for Nikon cameras" - Nikon cameras were the first to introduce the standard. They are the first and by the pure fact of it notable. The retrofit to old cameras shows that it's not a hardware standard that can only be implemented with new hardware (if you have a XQD card slot). Show me one manufacturer who does a retrofit like that. If you want new software for you Apple device, you have to buy a new phone even the old one is capable of running it.
There was more information about other cameras as well, but someone "Removed the entire Compatible Devices section". Can't remember who it was though... I just kept the information that seemed to be notable enough to be kept. That is: The first implementation in a camera and the wide adoption across the industry (Nikon, Phase One, Canon, Sony). I did not include the details between the Canon EOS-1D X Mark III (2x CFe slots) and the Canon EOS R5 (1x CFe + 1x SD slot) as this seemed to be not important for me. If someone feels to add this information they are free to do so.
4) I said this in my comment and I'll say it again (here corrected and not a 1:1 copy of the edit comment): the last two reverts have shown you are not in for constructive work, but oh well... ¯\_(ツ)_/¯
If you want an example: (a) You just blindly removed information without transformation. You just trash other people's work instead of building upon it. (b) If other people try to build on past information you revert and revert again without any constructive additions, just for adding the same information you just removed before in later posts. The-FAQ?
I stand by my words and I am not afraid to repeat them: With destructive behaviour like yours a constructive, combined and positive work is not possible. I'm not against talking it out. I'm all in favour of discussion and finding the middle ground we all can live with. Blind deletions and reverts are not even considered to be discussed about for the obvious reason of it being wasted time.
--D-Kuru (talk) 19:01, 26 December 2025 (UTC)

In response:

I admittedly looked at the diff while half-awake while waking up, and assumed that your partial revert went much farther than it did. The compatible devices section used to be a complete mess littered with indiscriminate information. I apologize for the misunderstanding, and had added some information back before you reverted it back agian.

That being said, your presumption of bad faith on my side aside:

2.) I added the Nikon Z6 and Z7 back in recognition that these are the first cameras to use the standard. That being said, the current discussion into Nikon has way too much detail. Specific dates down to the day, specific firmware version numbers, and discussion of every model with firmware updates (D5, D500, D850) are not necessary for a reader to gain a good understanding of CFexpress. They clutter and lengthen the article unnecessarily and make it less readable, make the prose flow poorly, and arguably fall under WP:INDISCRIMINATE. A simple sentence, like the one I had in Form Factors ("Some older cameras with XQD slots have received firmware updates to allow use of a CFexpress Type B card in its XQD slot.") was sufficient. I rewrote the camera introductions to be more concise and smoother without the excess detail when I added that information back in, but you decided to scrap it entirely and replace it with the old, problematic text.

My various problems with your edits/reversions

1.) You have so far failed to give a reason for either changing the infobox image back to your image (which only shows a CFexpress Type B card rather than both Type A and B) or removing the infobox image entirely (most recent edit) and leaving only the CFexpress logo. The SD card article has an image showing each major form factor in its infobox, and my photo of the Type A card next to a Type B card had a similar role before you replaced/removed it without explanation.

2.) The Xbox claim is poorly sourced and does not belong in this article unless you find a better source. The first source is a website called ITIGIC that no one has heard of and is littered with ads and typos, and extremely unlikely to pass as a reliable source. Even then, it admits that CFexpress cards do not work on the Xbox Series One S/X even if the pinout is compatible. The second source is a forum post (meaning it is a self-published source), which is usually not acceptable as a reliable source. Even then, it veers on trivia; it is not particularly relevant to readers wanting to learn about CFexpress that a game console uses the derivative physical connector (but cannot actually use the cards themselves).

3.) Without the Xbox, CFexpress cards are exclusively used on cameras, and minor accessories like card readers to use those cameras with. At that point, there is no need for a separate "Compatible Devices" section when it would only list cameras and card readers, and become a huge unmanageable list like it was previously.

4.) You remove the cite I added for the Sony A7S III, leaving that uncited.

Defending this edit and removal of content

English Wikipedia is an encyclopedia meant for a general audience. Indiscriminate information, trivia, and excess detail are frowned upon, and "non-encyclopedic content should simply be removed" even if this means removal of people’s hard work.

We did not need to list every single CFexpress device - camera, card reader, card, and so on. We do not need a discussion of every single manufacturer of CFexpress equipment. It fails WP:LISTCRITERIA and WP:INDISCRIMINATE, and fails WP:UNDUE when some manufacturers receive a lot of detail and other manufacturers do not. There is not enough available editor labor to make a complete list and to keep it current. It is not encyclopedic and not useful to a reader to give a list of every single CFexpress device or to discuss every single CFexpress manufacturer, especially when it cannot be kept up to date. At that point, the best way to proceed is to remove all of the excessive information, which is what I did. 4300streetcar (talk) 01:05, 27 December 2025 (UTC)

I assume good faith unless there is a clear sign that it is not applicable. Your actions were such a clear sign to me and therefore good faith could no longer be assumed. I'm not unwilling to change my evaluation of the situation as IMO only assuming good faith makes Wikipedia possible. However, my evaluation only changes once there is a clear sign that such a re-evaluation is justifiable from my personal perspective eg. because responses overall turn into a productive co-work.

Section 1, 2) I don't mind removing the dates, I kept it because the information existed in the original post. I also don't mind removing the firmware version from the text, yet I would may keep them inside a <ref> tag. Having a note on a retrofit compatibility is IMO noteworthy as these cameras were never announced to have CFe support in the first place, but was later added resulting in the question why a camera announced in 2018 is the first one to support CFe if there is a camera from 2016 that also supports it. I disagree that these rather short notes (the entire paragraph has ~150 words) "clutter and lengthen the article unnecessarily and make it less readable, make the prose flow poorly". In direct comparison to today's article of the day (Trichogenes claviger) these are short notes. I still can't see how this "arguably fall[s] under WP:INDISCRIMINATE". I also can't find any convincing arguments in your text other than the claim of being so. Considering that CFe is based on the XQD card and this standard was announced by SanDisk, Sony and Nikon, Nikon might play a much more important and central role in this standard after all. Noting so would not be an over representation. "Some older cameras with XQD slots" does imply that there are more manufacturers who did so which is not true. Nikon did so and only for three of their cameras. Same with the text being "problematic" I don't see how and where exactly it should be problematic. For editing the many references makes editing much harder than the actual text itself.

Section 2, 1) I I moved the previous image (comparing Type A and Type B) to the section "Form factors" as it seemed to be better fitted in there than any other image to compare both form factors. I removed it from the top of the page as there is no reason to have it two times in the same article. I checked commons:Category:CFexpress for any other image and decided by visual quality that the best fit is either File:Delkin CFexpress.png or File:Transcend Information CFexpress card-top PNr°0910.jpg - I do not care which of these images is taken by whom, I care about the best quality. In the end I decided to go against File:Delkin CFexpress.png as the upload itself seems very fishy. The uploader claims that it's their own work, yet the overall style of the image, the amount of edits by the uploader and an old version of File:CFexpress Logo.jpg (where the same uploader claimed that the CFe logo is also their "own work") my impression that this was not the uploader's own work was too strong and I therefore decided against using it (even I think it's very fishy I assume good faith as long as I have some kind of evidence that the image is stolen from somewhere. If I had already found such evidence the file would already have been nominated for deletion). In the latest edit I decided to go against any image (other than the logo) as the article is about CFe as standard ("CFexpress is a standard for removable media cards") and not about any products build upon it (see also eg. Nikon (headquarters location), SATA (logo only), SD card (image about SD cards, "SD card is a [...] flash memory card format" -> card format and not the standard itself).

2) Both references lead to the same conclusion and after searching up images in the web, I evaluated the content as plausible enough to keep it. If I had found an image that would fit into the text (eg. side by side comparison) and was freely licenced I would have uploaded and included it. "no one has heard of" - you haven't heard of them, yet you apply your knowledge to everyone. "is littered with ads and typos" - advertising is how such websites finance themselves, that does not say anything about their quality. Neither does how many typos there are. Maybe not the best quality media on the planet, but if you are not a big media company you have to make some cuts. I read through the entire text and check for typos, yet I bet there are some left. However, I neither have the money nor do I want to spend it on a person checking my spelling. I wouldn't assume anything different from websites on a tight budget. For me only the information and whether it is true or not is important. Cards following the official standard do not work as the standard was adapted so it's not a big surprise that they can not be used. References are reference where the information is taken from and not primarily for the user "wanting to learn about CFexpress". The article is about the standard and an adaptation of the standard is relevant. Comparisons with other (or competing) standards (eg. VP8#Comparison_with_H.264) is not new or forbidden and neither is mentioning later versions, successors or derivatives. If so there would be the question why there is the comparison between CFe, SD, UFS, CFas and XQD cards. If there would be no need to mention later versions, successors or derivatives there would be a fast and easy solution for this discussion: We just speedy delete the article. As CFe has the same form-factor and interface as XQD and just uses the NVMe protocol there is no need to mention it on XQD card and there would be also no reason for it's own article. The NVMe protocol is actually probably just a minor adaptation. Yes, that adaption breaks compatibility but that's nothing that doesn't happen with every Windows 11 update, so we are fine here.
As you asked for other sources: (computerbase.de, german) de: "Der Anschluss entpuppte sich als CFexpress - Es hat sich schnell herausgestellt, dass der Anschluss der Expansion Card gar nicht so proprietär ist wie anfangs gedacht: Es handelt sich im Grunde um den neuen Speicherkartenstandard CFexpress, allerdings in angepasster Ausführung mit PCIe 4.0." | en: "The connection turned out to be CFexpress - It quickly became apparent that the expansion card connection is not as proprietary as initially thought: It is basically the new memory card standard CFexpress, but in an adapted version with PCIe 4.0." (Google Translate); (anandtech.com via archive.org, english) "The storage expansion units for Microsoft's newest Xbox Series X|S gaming systems are proprietary and only designed to be used with these consoles. Technically, these storage units resemble the CFexpress 1.0 Type-B design, yet they operate with a PCIe 4.0 x2 interface, whereas the CFexpress 1.0 Type-B specification mandates usage of a PCIe 3.0 x2 interface." Considering the section "Microsoft is thought to charge a licensing fee from makers of Xbox Series X|S storage expansion cards, and apparently Seagate was the only company to agreed to the software giant's terms, which is why it was the exclusive supplier of such storage units for over two years." it more than likely was to get more money from the same cards, but this would probably fall under original research.

3) There is no section called "Compatible Devices". Again: The article is about the standard, not the cards themselves. Having the standard adaptation by Microsoft just shows how a standard can be used and adapted to certain needs.

4) To be honest: I am a little bit amused by you complaining that I put in as little effort into my edit as you put in yours. Nevertheless, I checked and saw that the ref is already readded.

Section 3) Yes, the old list was a mess and I am not against trimming the information into something more useful.
Yet trimming is not Shift+Del + Enter. What is the definition of "non-encyclopedic"? Usually it's content people feel like that falls under it without having a clear vision of what it really applies to. In my opinion examples would be personal opinion or advertising and can also change depending on the article. Original research is included as long it's not real research (I prefer the german term "Theoriefindung" (word by word translation: 'theory finding') as to me this much more implies that the text is not based on "research"). Trivial details are trivial for me when they are either completely off topic (eg. "A CFe cards fits nicely into your hand") or repetitive and add nothing new (eg. provided explanation of EOS-1D X Mark III vs. EOS R5). Describing an edit as "non-encyclopedic" is probably the comment most opposite from an expressive, convincing or meaningful reason for a change. I could have just reverted your change with claiming it being non-encyclopedic leaving you the role to defend why it may is. "We did not need to list every single CFexpress device" - I agree. "We do not need a discussion of every single manufacturer of CFexpress equipment." Not every company, but first and big players. As I already said: The first in a field is a special position by the fact alone of being first. If a company has contributed more to a standard or a project there is no sense in reducing the description of their work down to the contribution of the least contributing entity (if so we would have to remove a lot about Google from the article about Android). If you think other manufacturers (Canon, Sony, etc.) receive too little detail, you are free to expand it. As you removed all information I doubt that you see any reason to expand these sections. My stance is: first and big players - later are IMO worth a longer description if they add something new (eg. a new product with the Type C standard or introducing a (theoretical and probably nonsensical or physically impossible) adapter from CFast to Type A). If a big player is only the next in line all information other than the time of adaption and first product is trivial information. Trivial information would be the complete lineup of supported cameras. The timeline of adoption is IMO useful up to the point of the adoption of the late majority (see ) as this makes the standard a standard (you could probably also call it market penetration). A counter example: On Linux there are many standards and some were created to collect competing standards into one and bundle developer effort - in reality all that happened was that there is yet another 'standard' that only few fully adopt. A complete list of devices never was nor is my target nor do I see it as something useful to have for the future. "It is not encyclopedic" - there it is again... "not useful to a reader to give a list of every single CFexpress device or to discuss every single CFexpress manufacturer" - again: I agree with this. "the best way to proceed is to remove all of the excessive information" - of course, the best way to have a nice hedge is cut the stem right where it comes out of the ground. How could I be so foolish and think that it's far better to give it a trim. Again: The difference for me is between trimming (reducing the text to important information) and Shift+Del + Enter (just deleting entire sections determining them as non-encyclopedic).

--D-Kuru (talk) 22:08, 27 December 2025 (UTC)

prose changes

2025-12-27

I tried to change the first two paragraphs into something that is less list-like. @4300streetcar: Feel free to comment, change or revert the edit. --D-Kuru (talk) 22:33, 27 December 2025 (UTC)