2 ms·

Do you expect anyone to believe the Tiobe index? They say Apache Groovy has risen from #60 a year ago to #16 now [1]. This is obviously wrong since software sit

by vorg 8y ago

Do you expect anyone to believe the Tiobe index? They say Apache Groovy has risen from #60 a year ago to #16 now [1]. This is obviously wrong since software sites have been dropping Groovy for other JVM tech. [1] https://www.tiobe.com/tiobe-index/ https://www.tiobe.com/tiobe-index/

zmmmmm 8y ago
I agree with you mostly about Tiobe, it's entirely driven by search trends which can fluctuate for reasons having nothing to do with whether a programming languages is actually being used. Usually even bad benchmarks are comparable relative to themselves but Tiobe's methodology means you can't even really correlate a language's performance relative to itself in the previous survey. But Groovy seems to be going strong. There was definitely a difficult period after it had to transition to Apache, but momentum seems to be picking up again, and if anything it is getting more used than previously, though in different ways to how it used to.
- vorg 8y ago
  > more used than previously, though in different ways to how it used to Are these new uses of Apache Groovy really what we would call "using" a programming language? Although many "use" Groovy in Gradle, it's mainly 20-liners without any procedural code, so really only the parsing functionality is being used. Jenkins goes further by transforming the AST after parsing to such an extent that none of the collections-based functions work, and the runtime behavior isn't really Groovy at all. The adoption of Groovy in these uses is so dependent on the nitty gritty behavior of its Antlr 2.x -based parser that the Apache Groovy PMC seem to be stalling on moving along the upgrade to Antlr 4 in Groovy 3, keeping it in alpha mode for as long as possible, and even removing the optional Antlr 4 preview feature from the Groovy 2.6 beta. Most other uses of Groovy out there use it like "Bash for the JVM". For building actual systems, sites use Java (with lambdas), Kotlin or Scala rather than the static typing that was bolted on for Groovy 2. Even dynamically typed uses of Groovy are losing momentum, evidenced by how few Grails 2 websites and plugins have been upgraded to Grails 3, and how few new websites have been built in Grails 3. The official download numbers for Groovy are grossly manipulated, and Tiobe's ranking for Groovy has been gamed in the past by faulty search numbers, e.g. 2 yrs ago it was Baidu giving bad results. I suggest your view of Groovy's momentum might be skewed by what you see immediately around you, and not by overall worldwide trends.

by vorg 8y ago

zmmmmm 8y ago

I agree with you mostly about Tiobe, it's entirely driven by search trends which can fluctuate for reasons having nothing to do with whether a programming languages is actually being used. Usually even bad benchmarks are comparable relative to themselves but Tiobe's methodology means you can't even really correlate a language's performance relative to itself in the previous survey. But Groovy seems to be going strong. There was definitely a difficult period after it had to transition to Apache, but momentum seems to be picking up again, and if anything it is getting more used than previously, though in different ways to how it used to.

vorg 8y ago
> more used than previously, though in different ways to how it used to Are these new uses of Apache Groovy really what we would call "using" a programming language? Although many "use" Groovy in Gradle, it's mainly 20-liners without any procedural code, so really only the parsing functionality is being used. Jenkins goes further by transforming the AST after parsing to such an extent that none of the collections-based functions work, and the runtime behavior isn't really Groovy at all. The adoption of Groovy in these uses is so dependent on the nitty gritty behavior of its Antlr 2.x -based parser that the Apache Groovy PMC seem to be stalling on moving along the upgrade to Antlr 4 in Groovy 3, keeping it in alpha mode for as long as possible, and even removing the optional Antlr 4 preview feature from the Groovy 2.6 beta. Most other uses of Groovy out there use it like "Bash for the JVM". For building actual systems, sites use Java (with lambdas), Kotlin or Scala rather than the static typing that was bolted on for Groovy 2. Even dynamically typed uses of Groovy are losing momentum, evidenced by how few Grails 2 websites and plugins have been upgraded to Grails 3, and how few new websites have been built in Grails 3. The official download numbers for Groovy are grossly manipulated, and Tiobe's ranking for Groovy has been gamed in the past by faulty search numbers, e.g. 2 yrs ago it was Baidu giving bad results. I suggest your view of Groovy's momentum might be skewed by what you see immediately around you, and not by overall worldwide trends.