{ "version": "https:\/\/jsonfeed.org\/version\/1.1", "title": "Ilya Birman’s Blog: posts tagged text", "_rss_description": "Ilya Birman’s blog on design, cities, music, and life", "_rss_language": "en", "_itunes_email": "ilyabirman@ilyabirman.net", "_itunes_categories_xml": "<\/itunes:category>\r\n<\/itunes:category>\r\n\r\n", "_itunes_image": "https:\/\/ilyabirman.net\/meanwhile\/pictures\/userpic\/userpic-square@2x.jpg?1573933764", "_itunes_explicit": "no", "home_page_url": "https:\/\/ilyabirman.net\/meanwhile\/tags\/text\/", "feed_url": "https:\/\/ilyabirman.net\/meanwhile\/tags\/text\/json\/", "icon": "https:\/\/ilyabirman.net\/meanwhile\/pictures\/userpic\/userpic@2x.jpg?1573933764", "authors": [ { "name": "Ilya Birman", "url": "https:\/\/ilyabirman.net\/meanwhile\/", "avatar": "https:\/\/ilyabirman.net\/meanwhile\/pictures\/userpic\/userpic@2x.jpg?1573933764" } ], "items": [ { "id": "341", "url": "https:\/\/ilyabirman.net\/meanwhile\/all\/how-to-trim-text-the-right-way\/", "title": "How to trim text the right way", "content_html": "

Recently, I wrote about Facebook’s crappy handling of long posts<\/a>: it’s trims the text randomly and loses your reading position when expanded. So how should it be done? Let’s deal with the trimming first.<\/p>\n

Trimming text in a random place is disrespectful to the author and the text, and is a further proof that Facebook doesn’t care what you write. It’s also an example of deep technology dependence: it’s easier for a machine to trim to a certain number of characters, and so it does.<\/p>\n

There could be an aesthetic rationale behind trimming text, i.e. to make it fit into a certain design element. Of course this is also disrespectful to the author and the text, but at least the motives are humanistic. However trimming after a fixed number of characters is not even this: the physical size of lines of equal character length is usually variable: compare “iii” and “WWW”. Even in my Facebook example, a few more words would fit in the same box:<\/p>\n

\n\"\"\n<\/div>\n

Sometimes trimming randomly not only ruins the meaning of the text, but changes its to something indecent:<\/p>\n

\n\"\"\n<\/div>\n

So, if you can’t do without trimming, you have to trim carefully. How does Aegea do it?<\/p>\n

The character limit is considered a rough guideline, not an exact value. If you need to fit in 140 characters, and the text is 143 characters long, then Aegea just won’t touch it. If the text is noticeably over the limit, then Aegea will divide it into sentences and try to take as many full sentences as it can fit. Here, for example, is the snippet of this post:<\/p>\n

\n\"\"\n<\/div>\n

However, if the result is too short (or no whole sentence fits at all), Aegea will look for other signs of safe trim positions: semicolons, dashes, commas, brackets. Only if none of these are present will it trim at a word boundary. And only if there are no word boundaries, will it trim by the number of characters. The absence of word boundaries is a sure sign that the author themselves didn’t care about the meaning, so it’s fine if neither does Aegea.<\/p>\n

In general, Aegea will try to make it as close to the desired length as possible, but so that the meaning does not suffer.<\/p>\n

If I can code this, what would stop Facebook from doing it? I think it’s that it hadn’t even crossed anyone’s mind. Why would Facebook even care about doing anything well? What a nonsense. There is no such metric.<\/p>\n", "summary": "Recently, I wrote about Facebook’s crappy handling of long posts: it’s trims the text randomly and loses your reading position when expanded", "date_published": "2022-10-25T12:50:10+03:00", "date_modified": "2022-10-25T12:49:26+03:00", "tags": [ "Aegea", "being human", "interface", "text" ], "image": "https:\/\/ilyabirman.net\/meanwhile\/pictures\/fb-see-more-3@2x.jpg", "_date_published_rfc2822": "Tue, 25 Oct 2022 12:50:10 +0300", "_rss_guid_is_permalink": "false", "_rss_guid": "341", "_rss_enclosures": [], "_e2_data": { "is_favourite": true, "links_required": [], "og_images": [ "https:\/\/ilyabirman.net\/meanwhile\/pictures\/fb-see-more-3@2x.jpg", "https:\/\/ilyabirman.net\/meanwhile\/pictures\/liz-truss-anal@2x.jpg", "https:\/\/ilyabirman.net\/meanwhile\/pictures\/aegea-snippet-of-this-en@2x.jpg" ] } }, { "id": "161", "url": "https:\/\/ilyabirman.net\/meanwhile\/all\/compose-new-message\/", "title": "The stupid “Compose New Message” Mail.app menu item", "content_html": "

Let’s say you’ve right-clicked the Mail.app’s dock icon and want to write a new mail:<\/p>\n

\n\"\"\n<\/div>\n

Why, why is the menu item called “Compose New Message”? Why does it say “Compose”? No sane person would ever say: “Honey, I need to compose<\/i> a message”.<\/p>\n

In every other app on Earth<\/i> there is no verb before “New”. Just “New Window” in Safari. Just “New Event” in Calendar. Heck, in iMessage it is just “New Message”. Who does Apple make me spend several seconds trying to find the line I need in Mail.app? This is one of those things you cannot get used to.<\/p>\n", "summary": "Let’s say you’ve right-clicked the Mail.app’s dock icon and want to write a new mail", "date_published": "2014-05-26T23:25:48+03:00", "date_modified": "2014-05-26T23:26:38+03:00", "tags": [ "interface", "Mac", "rants", "text" ], "image": "https:\/\/ilyabirman.net\/meanwhile\/pictures\/compose-new-message.png", "_date_published_rfc2822": "Mon, 26 May 2014 23:25:48 +0300", "_rss_guid_is_permalink": "false", "_rss_guid": "161", "_rss_enclosures": [], "_e2_data": { "is_favourite": false, "links_required": [], "og_images": [ "https:\/\/ilyabirman.net\/meanwhile\/pictures\/compose-new-message.png" ] } }, { "id": "147", "url": "https:\/\/ilyabirman.net\/meanwhile\/all\/what-is-brick\/", "title": "What is Brick?", "content_html": "

So this thing, Brick<\/a>. What is it? It says:<\/p>\n

\n

Webfonts that actually look good.<\/p>\n<\/blockquote>\n

So it is a list of good fonts for the web? A collection of them? A book on them?<\/p>\n

There is no explanation .There is no About link. There are just two buttons: “Get stated” and “Browse fonts”. So presumably I get started<\/i> with the fonts by doing something else than browsing<\/i> them.<\/p>\n

Below, I see a heading: “Why Brick?” Wait, first of all what<\/i> is Brick? The question “Why Brick” makes no sense to me. It says: “Beautiful”. What is beautiful? The fonts? The website?<\/p>\n

\n

We don’t modify or subset the font in any way, so they are rendered the way they were meant to be seen.<\/p>\n<\/blockquote>\n

That is amazing now that I know what you don’t do<\/i>. But what do<\/i> you do? Why do you even say that you don’t modify or subset the font? I mean, me too, I don’t modify or subset fonts either. Heck, I should put it on my frontpage! (Is it what everybody else does?)<\/p>\n

It then says: “Fast”.<\/p>\n

\n

Brick is served to your users through Fastly’s industry-leading CDN network throughout.<\/p>\n<\/blockquote>\n

OK, so Brick is served to my users. Why?<\/p>\n

It then says: “Open source”. Ok, this probably explains why I can’t figure it out. Anyway, having no other option, clueless and confused I click “Get started”.<\/p>\n

And I end up on Github. And the project description starts with — wait for it —<\/p>\n

\n

Brick was built to be easy-to-use.<\/p>\n<\/blockquote>\n

I bet.<\/i> But I have no fucking clue<\/i> what Brick is.<\/p>\n

Web design 101: Use nouns to describe things.<\/p>\n", "summary": "So this thing, Brick. What is it? It says", "date_published": "2014-03-22T01:31:19+03:00", "date_modified": "2014-03-22T01:31:43+03:00", "tags": [ "design", "rants", "text" ], "_date_published_rfc2822": "Sat, 22 Mar 2014 01:31:19 +0300", "_rss_guid_is_permalink": "false", "_rss_guid": "147", "_rss_enclosures": [], "_e2_data": { "is_favourite": false, "links_required": [], "og_images": [] } }, { "id": "131", "url": "https:\/\/ilyabirman.net\/meanwhile\/all\/great-title\/", "title": "This is how you write a great title for a blog post", "content_html": "

Several years ago, if you were to write an article on article titles and you wanted its own title to work best, you would name it something like “Coming up with a great title for an article” or even “10 qualities of a great title”. That was how you attracted clicks.<\/p>\n

But recently something has changed. The articles now have titles like “Here is what you should know before updating iOS” or “Look at this amazing picture of Mars made by Curiosity”. As if the first sentence of the text was used in place of a title. The more and more we see these titles that try to shift your attention from themselves to what goes after them. It is the content of an article, right?<\/p>\n

Wrong. What really<\/i> goes after them is a link. That is how twitter sharing works. Now that the title itself is no longer clickable, there no need to attract clicks to it. Instead, a title should look like a direct speech of someone who shares the link, adding to the credibility of the link that follows.<\/p>\n

I find it fascinating how technology changes language.<\/p>\n", "summary": "Several years ago, if you were to write an article on article titles and you wanted its own title to work best, you would name it something like “Coming up with a great title for an article” or even “10...", "date_published": "2013-11-03T00:54:02+03:00", "date_modified": "2014-07-06T22:57:26+03:00", "tags": [ "English", "text" ], "_date_published_rfc2822": "Sun, 03 Nov 2013 00:54:02 +0300", "_rss_guid_is_permalink": "false", "_rss_guid": "131", "_rss_enclosures": [], "_e2_data": { "is_favourite": true, "links_required": [], "og_images": [] } } ], "_e2_version": 4269, "_e2_ua_string": "Aegea 12.0a (v4269e)" }