Beefy Boxes and Bandwidth Generously Provided by pair Networks
Pathologically Eclectic Rubbish Lister
 
PerlMonks  

Re: Telling Users What They Want is Stupid Never Works

by raiph (Hermit)
on Jan 20, 2012 at 08:47 UTC ( #948910=note: print w/ replies, xml ) Need Help??


in reply to Telling Users What They Want is Stupid Never Works
in thread Hockey Sticks

Rakudo went off the rails when it's continually failed to improve its bus number
Rakudo hasn't gone off the rails. pmichaud was lead dev, and he is of course intermittently dropping out of the picture due to his wife's illness. That sucks. In the last 12 months there has been loads of work by jnthn, moritz, tadzik, japhb, and colomon, and many others.
reimplemented large portions of Parrot badly
Whiteknight, current Parrot lead dev: "6model is far superior to what Parrot provides now".
forked into a dead branch from which Star releases slowly petered out (hey, even compiler releases slipped and skipped)
They missed 1 monthly compiler release last year and chose to defer an update to Rakudo Star till the new branch was stable. A new Star is due out in a few weeks.
and went into yet another rewrite mode which is still suffering from scope creep?
Rewriting isn't an end in itself. There's an overall objective. Ship a sufficiently complete, bug-free, fast, documented, product as fast as possible. If it helps towards that, it isn't scope creep.
"Hey, this proposed rewrite has some huge risks..."
To have NOT done the rewrite involved the ultimate risk: relying on Parrot to fix technical and project problems that the Rakudo team considered life-and-death for Rakudo. The Rakudo team is glad it did 6model, qregex, etc. even if merging it all has taken a long time, because it was ultimately necessary to give Rakudo a shot at reasonable performance.
Every major Perl 6 implementation has had severe project management problems
It's taken 12 years so far. It clearly didn't go smoothly. But ignoring Parrot, and other than the usual problems that beset even reasonably well run OSS projects, like a low bus number, what are the problems you are talking about?
blaming skeptics for not jumping up and down at the message "Yay, a single number measurement with no meaning behind it has increased, Christmas is coming!"
You may think there's no meaning behind numbers except that which is mathematical, and maybe you had to be there during the weeks that ran up to it, but I wasn't the only one who felt that 100.02% was magical, and I was sharing a celebration, my first Meditation in 3 years. If you don't feel like sharing in the good vibe, that's fine, but you could still choose not to pour cold water over it.
If you want to fix the Perl 6 marketing message, release usable software
Imo we're nowhere near the point that we can fix the Perl 6 marketing message as it relates to non-contributors. It's almost universally ridiculed, just like Mozilla was the year before Firefox was released. Imo it's worth trying to fix the message as it relates to potential contributors, so we can get the bus number up. Perl 6.0.0 will ship before Christmas but it could do with help if that's to be this Christmas (or, perhaps more realistically, Christmas 2013; maybe even later -- really, it's very hard to predict).


Comment on Re: Telling Users What They Want is Stupid Never Works
Scope Creep
by chromatic (Archbishop) on Jan 22, 2012 at 07:14 UTC
    Rakudo hasn't gone off the rails.

    Read the front page of Rakudo.org. That August 2011 nom distribution release still hasn't happened. The Star release you mention has been a few weeks away since July.

    A project which cannot (or will not) release working software has indeed gone off the rails.

      I'm trying to combat my own scope creep, namely responding to your scope creep, but I'm going to do this one more time...

      Read the front page of Rakudo.org. That August 2011 nom distribution release still hasn't happened. The Star release you mention has been a few weeks away since July.
      Quoting from that front page: "I donít want to specify an exact date for the next release; certainly it will occur within the next three weeks." The next monthly Rakudo compiler release (they don't refer to these as "distributions") happened exactly 3 weeks later. It was based on the new object model, as they had hoped.

      The original plan was for Star releases (they refer to these as "distributions") to be on a quarterly release schedule in 2011. They did Jan, a couple of April variants, and July. The next was scheduled for October. Then...

      Patrick, the lead dev (and rakudo.org webmaster), more or less dropped out of the project late September last year. Patrick had nom work underway. It was possible he would return, so it made no sense for others to dig in to his work. Instead they initially chose to defer the Star release until he returned, finished and merged his work, and did the release, as he had been doing them up till then.

      jnthn and Moritz took over as lead devs. From late October, news from Patrick dried up. By December, jnthn concluded he needed to go ahead with a Star release without Patrick's code and help. He took a proper Christmas break and now has a Star release nearly ready. (Ironically enough Patrick recently appeared on #perl6 and said he hopes to be back in action this coming week.)

      I ask that you (and anyone else) consider the need to read things carefully, be wary of jumping to conclusions, consider asking perl6ers (I recommend #perl6) to get the latest news, be considerate of Paula's condition, and avoid being assertive about problems with Perl 6 unless you know you're right, because unfair negative comments may make a difficult situation (personal pain) worse as well as discouraging potential contributors.

        Fine, you win. I couldn't deploy it because I didn't want it hard enough—there were no technical barriers in the way, just my bad attitude.

        Now let's all put on our happy faces and pretend Rakudo Star has been going swimmingly, just according to plan.

        I don't want to discredit Patrick, far from that. He's done some amazing work. And the same counts for Johnathan. On top of that, it's always nice to catch up with either of them if I run into them at a conference.

        But it's worrysome that a project almost comes to a halt if a single person, for whatever reason, is suddenly unavailable. That's not a sign a project is anywhere near the state of "mature". If I were an IT manager of a company that may consider switching to Perl6 at some time in the future, I'd take notice.

Log In?
Username:
Password:

What's my password?
Create A New User
Node Status?
node history
Node Type: note [id://948910]
help
Chatterbox?
and the web crawler heard nothing...

How do I use this? | Other CB clients
Other Users?
Others examining the Monastery: (4)
As of 2014-12-22 06:25 GMT
Sections?
Information?
Find Nodes?
Leftovers?
    Voting Booth?

    Is guessing a good strategy for surviving in the IT business?





    Results (111 votes), past polls