back to article Drinks-for-stars promo on open-source GitHub project scrapped after disgusted devs shame it into oblivion

An open-source project on GitHub has been thrown into disarray after contributors complained it was being promoted via a free drink in return for a star. The promotion, now withdrawn, appeared on a site operated by SK Telecom, the largest wireless communications provider in South Korea, which sponsors the project. There were …

  1. Anonymous Coward
    Anonymous Coward

    Why it's best to self-host

    When you screw up in a big way (this is just an embarrassing example, but say someone wrongly published secrets in the bug tracker or so), being self-hosted allows you to properly scrub the problem bits without having to rely on the patience and willingness of whoever is hosting your repo's instance.

  2. KittenHuffer

    As my other half always says .....

    ..... Wherever there's a scheme there's a schemer!

    My version is that as soon as you introduce a 'Metric' people will stop working the job and instead start working the 'Metric'.

    1. Anonymous Coward
      Anonymous Coward

      Re: and instead start working the 'Metric'.

      I think this is pretty well known. Of course, and depending on the type of job tasks, the problem is that without a metric they may or may not be working on what you want, and even if they are you can't always tell. A metric is a double-edged sword, to be sure, but sometimes it might be better than a sword with no edges.

      However, if you happen to be working in pre-metric electromagnetism (https://arxiv.org/abs/physics/0610221), the use of a metric is necessarily deprecated :-)

      1. KittenHuffer
        Facepalm

        Re: and instead start working the 'Metric'.

        I've seen a number of cringe-worthy examples of metrics introduced to improve a process that actually ended up making the situation worse.

        One of my favourites was a First Time Yield (FTY) metric added to a checking process that was intended to shorten the average time the process took end-to-end. The intention being that if the checks were successful the first time then it elminated the time necessary to go back and correct the problems that failed the checks.

        What actually happened was that the supplier to which the metric was applied decided to replicate and re-implement the checking process, effectively to pre-check the checking process. This was done specifically to raise the FTY returned on the customers checking process.

        The customer was quite happy that the FTY was increased within their checking process, and I'm sure that some Mangler got a pat on the back and a bonus for the idea.

        But what was actually happening was that the bulk of the time in the end-to-end checking process had been pushed upstream to the supplier, out of sight of the customer. And the true end-to-end time for the checking process actually became greater than it had been before the metric was introduced.

        I do remember that the supplier then introduced a FTY metric to their system, with which they could 'encourage' their subcontractors to 'improve' their checking, to reduce the end-to-end times for their pre-checking process.

        I always wondered if the subcontractor then decided to introduce their own pre-pre-checking system to improve their FTY on the suppliers pre-checking system!

        And this was how it ended up as turtles all the day down!!!!

  3. GeorgeMendes

    When you screw up in a big way (this is just an embarrassing example, but say someone wrongly published secrets in the bug tracker or so), being self-hosted allows you to properly scrub the problem bits without having to rely on the patience and willingness of whoever is hosting your repo's instance.

    xnxx xxx

    my issue got solved!!

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Biting the hand that feeds IT © 1998–2019