Some semblance of a blog*

Friday, July 5, 2013

I have recently had the chance to go over a number of transcripts that were professionally transcribed real-time. Given that they were transcribed real-time, I think they were pretty good.

That said, a good number of them still contained a lot of errors, some more obvious than the others. The most obvious ones include things from impossible spellings to complete sentences that don’t make any sense, and the less obvious ones, incorrect Twitter names and the like.

I don’t know why some of the transcripts are better than the others. Was it a different kind of transcriptionist? (Say, court reporters?) Because I know a computer would not have helped.

If people are going to put up transcripts like this, should they get a proofreader to go over them before they put them up for the public to read?

Sunday, June 23, 2013

I recently read a book called Inside Japanese Ceramics by Richard L. Wilson. Midway through the book, I was puzzled to see the word bisque misspelt as biscuit.

You know, this is an egregious mistake.

What bothered me was the fact that the author was an experienced potter, and bisque is a word that even absolute beginners—those who have been exposed to ceramics for even just a single day—would know. In short, this is obviously not an error the author could have committed. But if this is not the author’s fault, then it suggests the misspelling has been introduced at the editing stage, by a copyeditor who was neither familiar with ceramics nor was bothered enough to learn the correct terminology.

This shows how important it is to find a copyeditor (or proofreader, or reviewer) who either has the requisite subject knowledge or is willing to research into the field to gain it. You don’t want to write a book, have it published and printed, only to be ridiculed by absolute beginners when you, yourself, are the expert and you wrote nothing wrong.

The copyeditor’s mantra is “This is not my book.” But this can only go so far before this attitude degenerates into not being serious, before the copyeditor is doing the author a disservice instead of providing the author with a service.

Sunday, June 23, 2013

It is very hard to have a typographically good custom tab on ProZ. Why? Because a lot of perfectly good CSS is either illogically filtered out, or incompetently changed into syntactically invalid junk.

Take the length of the measure. It is a well-known fact that the measure of body text should be, say, around 70 characters if the text is in English. (See, for example, Book Typography by Mitchell & Wightman, p. 35). Yet max-width is filtered out. We can’t even have a custom width for our measure; this pretty much totally defeats the purpose of having a custom tab.

And we can’t choose our fonts. Yes, of course, we can choose our fonts, but we just can’t choose our own truly custom fonts. You know, Extensis is running a webinar series called Web Safe Fonts Are Dead and they are absolutely right in that web safe fonts are dead. You can’t even assure Georgia is on your potential customer’s browser—It could be an Android phone, for what it’s worth.

We can’t have rounded corners. We can’t have video. Ok, we can have video, but the visitor is going to click the No button and the video won’t get shown; that’s as good as not allowing us to have video.

So why this hostility towards webfonts, measure lengths, videos and rounded corners? Honestly I’d say it’s not so much hostility as abject ignorance: Ignorance of what valid HTML and CSS look like, ignorance of what really constitute dynamic content in today’s world, ignorance of what CSS does and what’s important and what’s not.

The quality of ProZ’s HTML is generally bad. (See, this page’s H1 is “ProZ Ideas.” What has that to do with this page, that it should deserve to be the page’s main heading?) Yet we can’t even use CSS to bring the state of things back to a semblance of a professionally acceptable quality.

We are promised we can name our tabs up to 15 characters—It’s an unreasonably short limit (just try a Finnish word), but let’s face it, it’s not our system. Yet it’s really not so much a 15 character limit but a 15 byte limit in reality. Titles in Chinese get cut off after a mere 5 characters. French, German, and Polish get cuts off a little later, but we are shortchanged. Non-English speakers are getting less than what we’ve been promised.