Easyling Release Notes May, 2021

With over half of the residents in Hungary having received their first dose of the vaccine against Covid-19, life is slowly returning to normal. As such, we are hard at work to improve Easyling for every user. This month, we added many useful features to the backend. The Ignore by classes feature on the Advanced settings now supports wildcards and the Crawler can now store and submit cookies sent by the original server. We also added an option to ban robots from externalised pages (or at least discourage them). Read the full article for the details.

Ignore by classes wildcard

The Ignore by classes feature has always been a great tool to mark parts of the site as non-translatable. Sometimes, however, you just don’t have a full list of the classes that you wish to ignore. Consider a site that has lots of blog pages, similar to ours. Let’s assume that every article has a small introduction that the client doesn’t want translated. These all have a class header-${id_of_article}. Previously, in order to exclude them, you’d need to add the class with every single ID on the site. Now, you can use a wildcard like header-* to ignore them all!

Ignore by classes wildcard

Note that this feature does not allow you to use regular expressions and that the * can only be used at the end of the expression.

Improved crawler

The crawler is the core of the translation workflow. It is being relied on to monitor content changes on thousands of websites. However, as the sites evolve, new features are added and technologies change. To keep the crawler compatible, we added a few things to improve it.

It used to have a hardwired user agent, a slightly outdated version of Google Chrome. Now, instead of that, it is the browser that you start the crawl from. According to statistics, chances are that it’s still a slightly outdated version of Google Chrome but this way, your automatic updates will update the crawler’s agent as well. Additionally, if there are browser-specific sections or features of the site you wish to translate, you can pick that content up as well by starting the crawler from that browser. This can also be a great way to separate the crawler in your analytics. Simply enter something unique, like “Easyling crawler” on the Miscellaneous tab of step 4 of the Crawl wizard. Just please don’t try Internet Explorer.

One way to block simple bot traffic is to provide a session cookie and expect it on further requests. The crawler is now prepared to circumvent such blocks. It can now gather cookies and return them to the original server. The feature is still in beta. If you wish to try it, drop us a line at support@easyling.com. We expect to make the toggle public soon™.

We also added a feature that detects if the site being crawled is badly damaged (e.g. the pages are redirecting to themselves) and stops the crawl. This can save you lots of credits that would’ve been unnecessarily spent. The crawler will only retry a page at most 10 times and if there are more than 10 thousand unsuccessful requests, it will stop.

Ban robots from externalised pages

There isn’t much reason why robots should visit pages that you don’t translate but if they are linked to, they will. This can cause SEO penalties because of duplicate content. As a countermeasure, you can now choose to let robots know that these pages should be avoided. You can access this feature on Publish → Global settings → Externalized page redirection handling.

Ban robots from externalised pages

Other improvements

Naturally, this month we had a few small improvements that are worth mentioning:

  • We added a new cache-busting system to avoid caching of wrong resource files when loading the Dashboard immediately after a new version of the Dashboard is deployed.

  • We fixed an inconsistency in the Workbench that caused translated entries to appear as deleted.

  • We added a button to insert soft hyphens in the Workbench. It can be found alongside other non-printable characters on the right of the target text entry field. Soft hyphens can be used to tell web browsers (and other programs that support them) where to hyphenate words if they don’t fit. This can be especially useful for target languages with long words such as German or Russian.

  • Web browsers’ cookie policy changes blocked HTTP pages in the Workbench. As a workaround, we now force the Workbench to load via HTTPS. As a side effect, this makes the Highlight view for HTTP pages inaccessible. The List and Pop-out views aren’t affected. No action should be needed on your end but if you experience any weirdness, please refresh the page.