Re: Is semantic HTML equivalent alongside an image of text, OK, bad, or fail of SC 1.4.5 Images of Text

Thank you Patrick, very helpful indeed.

I am not too surprised to hear that I did not quite get the relationship to G140; I'll need to take another run at that :)

But my big takeaway is that I'm pretty much OK to not fail people on 1.4.5 if they accompany their image of text infographic with a text equivalent. I pretty much thought that would be the case, but it's very useful to get another view and find it coincides.

Regards,

Alan
. . . . -   . . - - -
Alan Bristow ( he / him / il )
Web Developer / Développeur Web
Elections Canada / Élections Canada
alan.bristow@elections.ca

________________________________________
From: Patrick H. Lauke <redux@splintered.co.uk>
Sent: Wednesday, March 8, 2023 2:02 PM
To: w3c-wai-ig@w3.org
Subject: Re: Is semantic HTML equivalent alongside an image of text, OK, bad,  or fail of SC 1.4.5 Images of Text

Ce message a été envoyé par un expéditeur externe. Veuillez faire preuve de prudence et ne pas cliquer sur les liens ou ouvrir les pièces jointes à moins de reconnaître l'expéditeur et de savoir que le contenu est sûr.

This message was sent from an external sender. Please exercise caution and do not click links or open attachments unless you recognize the sender and know the content is safe.

On 08/03/2023 16:56, Bristow, Alan wrote:
> Hi,
>
> For an infographic, how would people characterize:
>
> - using an image of text but passing SC 1.4.5 "Images of Text" by the
> use of an accompanying semantic HTML text equivalent, i.e. meeting ST
> G140 (one of SC 1.4.5's STs)
>
> versus
>
> - no image of text (or text equivalent) and the use CSS instead
> (assuming the CSS was carefully designed to be excellent, responsive, etc)?
>
> For example, is the former greatly frowned upon, entirely against the
> spirit of 1.4.5, but ultimately does not fail since it meets ST G140,
> and the latter the entire reason 1.4.5 was introduced and should always
> be used if at all possible?
>
> Or is the former not quite as good but perfectly OK and this is not much
> of an issue?

Purely my person opinion here, so take it with a pinch of salt...

1.4.5 was a cute little aspirational SC trying to nudge authors away
from doing things like "fancy headings done as images because they want
to use this special font", "navigation bars where every link is a little
graphic so it can have embossed characters, a drop shadow, desert chrome
effect or whatever".

Even for that, the wording in the understanding for 1.4.5 is actually
quite wishy washy "The intent of this Success Criterion is to encourage
authors..."

Encourage, eh?

It also has this escape hatch: "Images of text may also be used in order
to use a particular font that is either not widely deployed or which the
author doesn't have the right to redistribute, or to ensure that the
text would be anti-aliased on all user agents."

Nowadays, while yes it is *technically* possible to recreate most
laid-out/fancy infographics, or the classic "product show with extra
text laid over it with the product name, price, etc" you'd get on some
shopping sites, using SVG...the effort to do that is often
disproportionately higher, and just not realistic for certain sites
(e.g. a marketing site built with a big CMS, where the person updating
the actual content  has no graphic design and SVG coding skills - they
can only upload an image for a post/product, can't upload SVGs, and
wouldn't be able to create SVGs themselves anyway). In those common
scenarios, it feels very tight failing a site for not using something
like SVG...

Long story short: I'm often likely far more lenient that I should be
when it comes to 1.4.5. I may fail it as a very low priority issue, with
the understanding that my clients are unlikely to spend any time
actually trying to tackle it.

However, I *think* you've slightly misread/misinterpreted what technique
G140 is saying? It's not about "using an image with text, but passing
1.4.5 because you double-up and also have the same info as text", like
you seem to imply? It's noted as a technique for 1.4.5 for doing
something like "put the large background of your infographic in the
background using CSS, and then put the text of the infographic over it
using semantic markup, rather than having it baked into the image". So
there's no tension between the two solutions / techniques you mention,
because they both drive to the same conclusion?

Regardless though, there is an argument that if you have the exact same
information that's currently locked up in an image with text also
available right there on the same page as actual HTML text, then this
wouldn't necessarily fail 1.4.5, because then the image with text is
just an alternative representation of the information that's there in
HTML...so you wouldn't fail?


--
Patrick H. Lauke

https://www.splintered.co.uk/ | https://github.com/patrickhlauke
https://flickr.com/photos/redux/ | https://www.deviantart.com/redux
twitter: @patrick_h_lauke | skype: patrick_h_lauke

Received on Wednesday, 8 March 2023 20:28:17 UTC