1. 34
  1.  

  2. 18

    It would be nice if the author’s blog followed his own advice.

    1. 22

      The recent lobsters thread about unkind comments suggested replying if we thought a post was unkind. I’d much prefer to downvote with an “unkind” reason, but in place of that, I’ll register here that I don’t like this comment.

      I think it makes the tone of the forum slightly less kind. The same comment could be said in a kinder way with no loss of clarity.

      Additionally, dev.to is a community site and not the authors personal blog.

      1. 2

        How should I have mentioned it then?

        1. 2

          i liked your original comment. thanks for it.

          1. 1

            Not mentioning it at all is the most appropriate option. You could have added “, assuming the author controls the site where he blogged.”

            1. 3

              but the author can still decide where to host their blog

              1. 3

                It doesn’t matter where they decided to post about it, it’s just good manners. Maybe they decided to post there because it was convenient, had an audience - or maybe they were trying to reach an audience of users who were as clueless about the topic as the creator of the site? If I was advocated for a particular approach I’d try to hit the places where the developers don’t get the topic.

                1. 1

                  yeah good point

        2. 10

          Well what do you know, it actually does. dev.to is not the author’s blog.

          Edit: I downvoted with incorrect, since you didn’t do your research:)

          1. 2

            I cannot find a link to his site on the article. Nor could I find a link from his site to the article in question.

        3. 3

          Are we all supposed to be saying ‘a11y’ instead of ‘accessibility’ now? That seems especially ironic.

          1. 4

            ‘a11y’ is a common term in the accessibility space, so it’s fine.

            1. 4

              I frequently see it argued (on the fediverse) that a11y is not fine, and that it is less accessible than “accessibility”. I personally have no dog in the fight, but I default to believing the people who say it’s not.

              1. 1

                What is the argument? That it’s easily misread using some fonts?

                The term is an analog of i18n for “internationalization”, and l10n for “localization”. The fact that it seems to spell out a positive term is a bonus.

                As I am the husband of a blind spouse I’m especially interested in the space…

            2. 1

              It is some american meme. They like to make up these kind of nonsensical expressions. I guess it is not that strange for them as they are accustomed to English language, which is not a phonetic language.

              My favourite is PPPoE (Point to Point Protocol over Ethernet) pronounced as poppy, but take SQL pronounced as sequel.

              Also this leet-speek transcript of ally is meant to be somehow supportive I guess, but I personally find it non inclusive as it also carries the meaning that we are with you but we are not totally like you. So to sum up I think this expression defeats its purpose in multiple ways.

            3. 2

              Whenever I work with HTML, I always remember myself when I just began opening the source of somewebsites and thinking to myself, “when I get a hold of this, I’ll never use div tags, at all”. I’m not that “radical” anymore, but the educational value of code is certainly something I want consider.

              1. 2

                Also, template languages like haml or slim encourage to use divs for everything, because it’s default element in them (i.e. .foo creates <div class="foo">). And such template languages are very popular, at least in ruby community. At least Haml has <section> in example on front page.

                1. -5

                  “Semantic” and “HTML” don’t belong in the same sentence. HTML is a presentational markup — it describes things like headings and emphases and tables — and it was never really designed to carry meaning.

                  In a way it is disappointing that XSLT never took off, because then we could have served meaningful data through XML (which, for all its evils, is very easy to define, standardise and validate against schema definitions) and transform it into something pretty for humans using XSLT and then we wouldn’t have to worry so much about whether a11y devices or search engines can make sense of it.

                  1. 7

                    Headings and emphases and tables describe semantic relationships. I’m not sure there are any presentational tags left in HTML5. Even <b> and <i> were redefined in terms of semantic usage.

                    1. 5

                      In a way it is disappointing that XSLT never took off,

                      I actually worked on old IE app that was all in on xsl and xslt server and client side. Xslt is an abomination. It works great for simple stuff. Start adding namespaces and versions to the schema and it falls apart completely. Has to do with having to match input namespaces in your xslt for whatever xml input you’re given iirc. I recall we had to add a step to all our inputs to strip namespaces off tags.

                      1. 3

                        I have also worked with XSLT. I cannot blame anyone, as I chose that because It seemed the right tool for the task. It wasn’t. :)

                        I dodn’t even get to use namespaces, I already hit some hard walls and had to do terrible hacks to overcome its limitations.

                        I remember NetBeans had a somewhat adequate editor and maybe debugger for XSLT…

                        I’m happy XSLT didn’t take off.

                        1. 3

                          A colleague of mine at Lonely Planet wrote a Ruby DSL (called RSLT, if memory serves) specifically to avoid having to deal with XSLT :)

                          1. 3

                            If I’d understood XSLT better, I’d have made the DSL generate XSLT - the ruby-in-ruby DSL was a performance bottleneck we didn’t need.

                            The main problem I was trying to solve was ‘how do I encode several thousand similar rules, many of which are not yet known’. That’s a problem where the answer is basically always “create a new language”.