Selaras Bahasa

I began to work as a regional language editor for Kompas daily (PT Kompas Media Nusantara), the leading and most influential newspaper in Indonesia, in December 2003. My daily tasks included correct typographical errors and check for grammar and consistency issues in pre-published articles to ensure they are well written, logically structured, and accurate. I got four helpers for these tasks: Buku Panduan Kompas (a guidebook), Kamus Besar Bahasa Indonesia (KBBI/a dictionary served as the primary reference for standardised words in Indonesian), Pusat Informasi Kompas' digital scrapbook as a part of Kompasnet, and internet search engines.

The primary reference was the guidebook due to its collections of standardised words according to Kompas, figures names and occupations, foreign words equivalents in Indonesian. KBBI was secondary in terms of standardised words. PIK had a massive (perhaps gigantic) collection of articles from 1965 to present day. The last hope was internet search engines if all else failed.

They did not come without weaknesses. The guidebook was clearly impossible to update in real time. KBBI was quite thick (and heavy!) and sometimes outdated. Each article clipping in PIK's data base was stored as is so mistakes were also preserved. Please do not ask me about accuracies of internet contents.

At the same time, my job demanded accuracy and ability to work quickly from me. I could be reading and checking data in 5 minutes without sacrificing accuracies due to approaching deadlines. These resulted in me being tired and stressed out.

I decided to create my own database of words, names, etc which could be updated anytime as my second primary reference. Tit for tat, my colleagues and superior in Jakarta shared the same enthusiasm. I started building a PHP & MySQL application with these criteria:

• A database which stored entries such as standardised Indonesian words, figures names and occupations, acronyms and abbreviations, foreign words equivalents in Indonesian, and nations names.

• Authenticated users could search through the data using full-text search.

The test results were encouraging. Users in average could find what they searched in less than a second, thus they could made informed decision faster. This improved the speed of previous methods wich was 2-3 minutes on average. Other advantages of this app were that the data was always easily updated and was validated by the direct superior, thus single source of truth.

As time passed by, the app got more features added such as Facebook-like wall, work schedule, and users management. Its users ranged from editorial language editors, advertising language editors, journalists, book editor of Penerbit Buku Kompas (acbook publisher), and language editors of Kompas.com and Kompas TV.

The URL changed several times. First was selaras.yudhawijaya.com hosted on my server. In march 2011 we bought the domain selarasbahasa.net until 2015 when the app was moved to company's Linux servers and its address was changed to selarasbahasa.print.kompas.com. Finally from 2017 until now, the app settled at selaras.kompas.id.

Before moving to the lates address, I rewrote the app using Laravel 5.4. Since my superior advised me not to use full-text search anymore, I began toying around with Elasticsearch which was at that time in also used in the company's applications. This was my first time also using Vuex to manage states.