15
$\begingroup$

I recently made the suggested edit to:

Does Pi contain all possible number combinations?

The edit transcribed the text of the image to the page to gain the following benefits:

Visibility and ranking on Google

Granted, people can search by image on Google, and this page should come up if you search for the image.

The thing is, Google doesn't default to searching by image, it defaults to searching by text; people searching for searches around/related to this topic (e.g. "can pi be used to represent all words in the English language?") would benefit by having the text transcribed on the page.

Why is this important? Because from day one, Joel and Jeff designed Stack Overflow, and by extension, Stack Exchange, for pages to turn up to people that type into search engines.

Jeff has written extensively on the the importance of Google and search engine visibility.

This is not Stack Overflow, but all of Stack Exchange shares a set of basic tenets; this is absolutely one of them: to gain new traffic (always a good thing) which has the potential to get new users who will contribute to the site (another good thing as well).

Screen-readers/web accessibility

For those that use screen readers, accessibility issues, that image renders the question worthless to those people. From Wikipedia's entry on web accessiblity (emphasis mine):

For example, when a site is coded with semantically meaningful HTML, with textual equivalents provided for images and with links named meaningfully, this helps blind users using text-to-speech software and/or text-to-Braille hardware.

By not providing the transcript, you're reducing the overall accessibility of that page.

In summary, the edit would have provided reach for new views on the site (which could possibly translate into more members) as well as make the page more accessible to everyone who could possibly view it.

In light of those two benefits, why was the suggested edit deemed "pointless" and "unnecessary", especially when I took the time to point out one of the big benefits in the description of the edit?

$\endgroup$
16
  • 9
    $\begingroup$ I agree. The rejection of the edit as "pointless" and "unnecessary" is quite strange. There are examples where such edits have been made in the past, e.g. I recall Jeff Atwood OCR'ing on one of the images that I had posted. $\endgroup$
    – Bill Dubuque
    Oct 20, 2012 at 16:25
  • $\begingroup$ @BillDubuque Does that mean that the edit is a valid edit for M.SE? If that's the case, should I resubmit the edit and reference this post and your comment to that effect for reviewers? $\endgroup$
    – casperOne
    Oct 20, 2012 at 16:34
  • $\begingroup$ Let's wait a bit to allow the community to speak. Perhaps those who rejected the edit will share their thoughts on the matter (I have invited them to join this discussion) $\endgroup$
    – Bill Dubuque
    Oct 20, 2012 at 16:41
  • 5
    $\begingroup$ The point about screen-readers is more than legitimate; I simply didn’t think of it at the time. I’m not moved by appeals to visibility and ranking on Google: I simply don’t consider them important enough to outweigh the redundancy. Perhaps the best solution would be the one that I think I saw Arthur Fischer make: to replace the image with the text but leave a link to the image. $\endgroup$
    – Brian M. Scott
    Oct 20, 2012 at 17:03
  • 8
    $\begingroup$ @Brian Thanks for your input. The fact that the text will open up the post to searches is important. Besides making it more easily located by those seeking answers, it also helps to eliminate duplicate questions here (e.g. recall the many duplicate of the popular question about operator precedence) $\endgroup$
    – Bill Dubuque
    Oct 20, 2012 at 17:12
  • 1
    $\begingroup$ @Bill: Although I frequently vote to close duplicates if they come to my attention, I’m really not much bothered by them. And at least in the case of this particular question I rather think that between the title and the answers, it wouldn’t be hard for a seeker to find. But as I said, accessibility for the visually handicapped is a very legitimate concern, and I’d happily replace the image with the text, provided that a link to the image is maintained. (Especially since I’m still on dial-up, and the image is big enough to be a nuisance!) $\endgroup$
    – Brian M. Scott
    Oct 20, 2012 at 17:20
  • 4
    $\begingroup$ This was my rejection. When I was mading it I havn't thought about this benefits. Excuse me, this is my fault. My reasoning was - one don't need doubling information in textual and graphical form. My apologies. $\endgroup$
    – Norbert
    Oct 20, 2012 at 18:32
  • $\begingroup$ Thanks to everyone for responding. @casperOne Please post an answer indicating the outcome and accept it, so that the question will not be bumped in the future. $\endgroup$
    – Bill Dubuque
    Oct 20, 2012 at 18:39
  • $\begingroup$ I always try to do that myself, principally for the benefit of vision-impaired users of the site. $\endgroup$
    – MJD
    Oct 20, 2012 at 21:31
  • $\begingroup$ I've found [my suggested edit][1] from the list of activity on my page, but it doesn't show up in the history of the question. Is that a bug? [1]: math.stackexchange.com/review/suggested-edits/21195 $\endgroup$
    – TRiG
    Oct 20, 2012 at 21:42
  • $\begingroup$ I made exactly the same edit for exactly the same reason on the same question. And yet my edit now no longer appears in the history of the question. What happened there? Actually, my edit did go one better, because I also fixed the grammar while I was at it. > Does it make absolutely any sense ? should be > Does this make any sense at all? I made that change in my edit. How did my edit completely disappear? *** $\endgroup$
    – TRiG
    Oct 20, 2012 at 21:42
  • $\begingroup$ @TRiG: The edit was rejected by two users. It shouldn't be on the revisions list. $\endgroup$
    – Asaf Karagila Mod
    Oct 20, 2012 at 22:04
  • $\begingroup$ @TRiG: *Thinks*. That makes sense, actually. $\endgroup$
    – TRiG
    Oct 21, 2012 at 0:30
  • $\begingroup$ @Norbert No need to apologize, it wasn't a condemnation. Meta is the place to expand on reasoning especially when there are contrasting views on a matter regarding the governance of sites. We hashed it out, the system worked. =) $\endgroup$
    – casperOne
    Oct 21, 2012 at 14:07
  • $\begingroup$ @Asaf and TRiG: Are you not able to see that it was rejected, and the reasons listed for the rejection? Perhaps this requires a certain rep level? $\endgroup$
    – Bill Dubuque
    Oct 21, 2012 at 16:39

2 Answers 2

Reset to default
10
$\begingroup$

Purely textual information just plain shouldn't be posted as an image anyway. For accessibility reasons alone it's really bad; unlike text you can't change the color, and rescaling an image with zooming is much less effective than zooming text. Additionally blind users just won't get the text at all (granted I'm not sure how accessible MathJax is accessible there either...but we can't do much about that).

The image really doesn't add anything to the text. It's got a pretty red background around Pi and some numbers in the background for flourish, but it's all completely unnecessary. Due to that I don't think it's necessary to include the image at all (particularly considering possible copyright issues I don't care about enough to dig into).

As for SEO not being our business...not entirely true. Joel + Jeff's Google Juice only goes so far. If your question is incorrectly categorized (tags), poorly described (title) or excludes common keywords, it's much less likely to get found. We certainly don't have the "semantic web" yet, so Google only knows what your content is based on textual descriptions. That's yet another reason text shouldn't be hidden behind text.

$\endgroup$
4
  • $\begingroup$ Descriptive tags and words may increase Google hits, but why should anyone consider that when deciding on an edit? There are reasons to add those tags (or not) as content improvement judged by SE considerations alone independent of any search engines. I don't understand why the users should go beyond that and guess what is best for Joel or Jeff or Google or anything else outside of the content of the particular posting. $\endgroup$
    – zyx
    Oct 22, 2012 at 22:28
  • 1
    $\begingroup$ @zyx We're approaching this in reverse. Quality is the number one thing that all posters on SE should provide in their posts. If the post has quality content, then search engines will pick it up, and you don't have to think about it. That said, the image reduces the quality as well as the accessibility of the post. Put in those terms, I think we all agree the change was necessary. $\endgroup$
    – casperOne
    Oct 23, 2012 at 17:34
  • $\begingroup$ @casperOne, I'm at -3 (from 5 downvotes) on an answer with the statements "if content is good, search engine visibility will follow" and "converting [images to] text is a good thing". There really do seem to be people advocating that regular use of the site should include working for Joel or Jeff or Google, or performing search engine optimization, and that kind of thinking deserved a direct rebuttal. $\endgroup$
    – zyx
    Oct 23, 2012 at 18:17
  • $\begingroup$ @zyx It's not about working for those people, it's about taking pride and wanting to grow your site. If you provide quality content, those things will happen, searchability will improve, and all those good things (it just looks like you're working for them, but you're doing it because you take pride in your community). The image was simply not the best version of the post that it could be. $\endgroup$
    – casperOne
    Oct 23, 2012 at 18:20
-4
$\begingroup$

Converting text originally posted as images is a good thing, but search optimization is not a compelling reason to insist on it.

Why is this important? Because from day one, Joel and Jeff designed Stack Overflow, and by extension, Stack Exchange, for pages to turn up to people that type into search engines.

Search engine optimization is for Joel and Jeff to worry about. No user can be expected to prioritize that when writing or editing content. Search engines may OCR all the images or will soon do so and I believe SE can do commercial-grade OCR if it wants to.

Jeff has written extensively on the the importance of Google and search engine visibility.

If content is good, search engine visibility will follow. If visibility is "too low", which does not seem to be happening, I don't think users here need to consider it their job to fix Google or increase the quarterly earnings at Stackexchange. For anyone who does care about that it is more time-efficient to just write SE a cheque. Or send bug reports, contribute to questions & answers, and whatever else you normally do. Editing for search engine visibility per se is a waste of time.

$\endgroup$
3
  • 3
    $\begingroup$ I agree that we don't have to worry about optimization for search engines too much. But better placing in Google search is not better only for SE as a company (to increase the quarterly earnings at Stackexchange if I may use your words), but it is also service to SE community. For the users who contributed to the question it increases the chance that their work will be noticed by other people and it will be useful. $\endgroup$
    – Martin Sleziak
    Oct 21, 2012 at 7:44
  • $\begingroup$ @Martin: While I agree with you, I am also reluctant to agree. I can see why someone who would google my name might think it's a bad thing I spend so much time on a website instead of actually working. $\endgroup$
    – Asaf Karagila Mod
    Oct 21, 2012 at 17:51
  • $\begingroup$ @Martin, the rationale for increasing the search rank was that "traffic has the potential to get new users". This has good and bad effects and it is not clear at all that users who appear through searches are desirable compared to growth by word-of-mouth, articles in specialized forums, and other means of attracting experts and quality contributors. I also have trouble understanding how spending $n$ seconds converting an image to text helps search placement compared to $n$ seconds increasing/improving the site in ways not motivated by search placement. $\endgroup$
    – zyx
    Oct 21, 2012 at 19:05

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .