Stories
Slash Boxes
Comments

SoylentNews is people

SoylentNews is powered by your submissions, so send in your scoop. Only 17 submissions in the queue.
posted by janrinok on Wednesday November 04 2015, @12:47PM   Printer-friendly
from the written-for-pirates? dept.

The R Consortium and the Linux Foundation are investing in a new code-hosting platform that will help streamline the development and distribution of software packages for R, the popular statistical programming language.

Titled R-Hub, the platform will offer development, building, testing and validation services for R packages. R developers proposed the creation of R-Hub in July 2015 to serve as "the everything-builder the R community needs."

A description of the R programming language can be found on Wikipedia.


Original Submission

This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 3, Informative) by VLM on Wednesday November 04 2015, @02:45PM

    by VLM (445) on Wednesday November 04 2015, @02:45PM (#258370)

    I'm a little confused, I've used CRAN and R-Forge looks interesting. So this is a third competitor because of NIH or its intended to wipe out CRAN and R-Forge or they're actually re-launching R-Forge and pretending the old one doesn't exist or is it a political thing where people aren't getting along so form their own little colony or ...

    The press release is full of the usual promises to do everything for everyone instantly for free everywhere, full of community-speak as if there's only one all encompassing one, LOL, but they carefully avoid discussing the existing systems or existing competition, which is weird.

    I'm in the periphery and I wonder next time I need something for a work project and go to CRAN, if it'll be gone or unaffected or ... ? I'm an interested user, but not interested enough to be a "R" guy and get all into the details. Just curious whats going on, at least as it affects me, if all the content-free corporate-speak is cut away.

    Its a strange amount of money to spend. I've seen OS'es and RTOS written at a cost of a couple cans of Mt Dew (well, OK a couple cases...) and I've seen major corporate initiatives blow $20M and accomplish absolutely nothing, but this is a peculiar in between value. Enough to hire like four dudes for three month contracts.

    • (Score: 0) by Anonymous Coward on Wednesday November 04 2015, @06:32PM

      by Anonymous Coward on Wednesday November 04 2015, @06:32PM (#258456)

      "This is the first of a series of $200,000 total of grants that the R Consortium's Infrastructure Steering Committee intends to award over the coming months."

      Is there an evil goal here?
      https://www.r-consortium.org/ [r-consortium.org]

    • (Score: 2) by gringer on Wednesday November 04 2015, @08:46PM

      by gringer (962) on Wednesday November 04 2015, @08:46PM (#258503)

      I've used CRAN and R-Forge looks interesting. So this is a third competitor because of NIH

      My guess is that this is a competitor to Bioconductor, which is a more integrated version of R package management, including validation and dependency tracking.

      Regardless of what it's trying to replace, it's the usual aggregation dilemma and will probably just result in adding another place to the mix of repositories people need to search on:

      https://xkcd.com/927/ [xkcd.com]

      --
      Ask me about Sequencing DNA in front of Linus Torvalds [youtube.com]
      • (Score: 1) by shrewdsheep on Thursday November 05 2015, @10:06AM

        by shrewdsheep (5215) on Thursday November 05 2015, @10:06AM (#258764)

        I concur that there is a risk of fragmentation of the R package world. Be reminded, however, that the number of R packages starts to rival the number of perl packages and exceeds 10.000 all sources taken together. So some specialization is possible and bioconductor is one example, focusing on the analysis of biological data.

        From a developer's point of view CRAN is certainly insufficient. The submission procedure is a PITA (very unfriendly handling), the infrastructure does not allow for bug-reporting, testing, code-inspection etc. I have started to host on github from where you can easily install into R. However, you are not indexed. IMO there is much room for improvement. The critical thing would be to maintain a global index including the major other hosting platforms. I hope (but fear it's not) that this is addressed by the current project.