Talk:Rosetta Code/List authors of task descriptions/Full list

Anomalies in Task Lists

Firstly, the Go entry has a bug - my regular expression is picking up 'Category:Requesting_Content' as a draft task (thereby increasing Short Circuit's count by 1) which is easily fixed.

However, whilst trying to reconcile the overall Task Lists with the list of tasks completed for Go, I noticed the following anomalies:

1. For some unknown reason, the [Reflection/List methods] task (authored by Outis) is showing up in both the 'full' and 'draft' task lists (its page shows it as a 'draft'). The Perl 6 entry has somehow detected the duplication (and shows it as 'full') but the Go entry hasn't (it's easy to fix). I don't know whether the Phix entry has detected it or not.

2. There are two 'putative' tasks, [FiveNum] and [Mutex], which aren't showing up in either list because their pages don't actually flag them as tasks. It appears that they definitely should be (there are plenty of implementations) but I thought I'd wait to see what Thundergnat has to say about it before changing them to tasks myself.

3. There are some entries in the tasks completed for individual languages which aren't actually tasks but adjuncts to them. Examples include [N-queens problem/dlx go], [Robots/Phix] and [Perl/LWP]. I don't know why this happens or whether there is anything we can do about it but it's worth bearing in mind if you're trying to reconcile the overall lists with those for your language.

--PureFox (talk) 15:15, 11 January 2020 (UTC)

I fixed the Fivenum Rosetta Code task,   it looks like an oversight when it was created.     -- Gerard Schildberger (talk) 15:54, 11 January 2020 (UTC)

(As for item # 1, above, Reflection/List methods):   it appears that it might have been   copied   and then possible modified,   and then the older version wasn't deleted.   I ain't touching it,   that's a job for some administrator.     -- Gerard Schildberger (talk) 15:54, 11 January 2020 (UTC)

Yeah, I don't think there's any doubt that Fivenum should be a (draft) task - I should have spotted the oversight when I added the Kotlin entry. Not sure about Mutex though. From the discussion page, it looks like Short Circuit originally intended it to be an 'encyclopedia' entry but, despite this, implementations were added for 28 languages and are showing up as 'tasks completed' for those languages. My vote would be to regularize matters by converting it to a task.

As you say Reflection/List methods is definitely one for Thundergnat to sort out if indeed it can be sorted out at this rather late stage. --PureFox (talk) 16:11, 11 January 2020 (UTC)

For what it is worth, [FiveNum] was marked as a draft task until September 12 2019 when the tasks original author removed the draft (and task) markup. My suspicion is (s)he was confused about what the markup should be. Ah well, fixed now, Thanks Gerard. It appears that [Mutex] never had correct draft/task markup even though Short Circuit was an early contributor to that page. Task markup added. Perl 6 has also detected [Reflection/List methods] as both a draft and a task. At this point I just selected one and went with that. (I believe, though I am not positive, that the dual nature is due to some of the Reflection/ entries being full tasks and that one being a draft so it shows up in both categories.) It's the only one like that, it never bothered me enough to invest much time figuring it out and, until very recently, Perl 6 was the only language that had made any attempt at solving this task. I'm going to go ahead and promote it out of draft (It's got many entries and no controversy). We'll see if that clears up the issue.

As far as the sub pages: [N-queens problem/dlx go], [Robots/Phix], etc. As long as there is a link on the main task page to the sub page, they should be counted correctly; at least that has been my experience. --Thundergnat (talk) 00:17, 12 January 2020 (UTC)

EDIT: ADDED LATER: As it turns out, I was wrong. Many of the sub-pages WERE being counted twice (or more!). I have gone through and checked all of the top 60 languages and fixed the markup to not be counted twice. There may be other sub pages that need fixing but I am out of time and energy for the moment. --Thundergnat (talk) 16:43, 18 January 2020 (UTC)

Thundergnat:   you've changed (at least) one REXX entry to that new thingy format,   (it was an "entry" under a page for solving a sudoku puzzle)   of the "entry" form of:

=== REXX ... ===     and was   not   of the form:

== REXX ... ==     which I believe is the form that you are interested in changing.

By the way, there was one more entry on that page that you missed.   Also, the spelling of REXX is REXX,   not Rexx.   Some people in past have used the other spelling, and at that time, it was miscounted   (at least, in the sense that it was   not   counted under the REXX computer programming language).   I corrected those (two) misspellings, and they were not easy to find.   Those misspellings did show up under the Rexx programming language (or maybe a category)   somewhere,   and they proved (at that time) difficult to locate and change as there wasn't a place for them be to listed.   I had since changed my REXX program   (used to count computer programming language entries as per the Rosetta Code task:   Rank languages by popularity)   to allow a lowercase or mixed-case spelling to be considered the same programming language.   It should be interesting to see how the changes you're making will change the counts for ranking languages by popularity.   I'm going to re-run my program and update the results.     -- Gerard Schildberger (talk) 17:39, 18 January 2020 (UTC)

Ah. My apologies for the wrong spellings. I should have noticed that I was screwing it up. In my (very feeble) defence, I had been called in to work for an emergency last night and had only gotten about 2 hours of sleep so was not at my finest (assuming I have a finest...). Thanks for fixing my misspellings. Hopefully these changes don't cause other issues. --Thundergnat (talk) 22:10, 18 January 2020 (UTC)

Not a problem.   What you need is a clone to help you out.   (As in the film   Multiplicity.)

It looks like promoting [Reflection/List methods] to a full task has solved the problem as it's no longer showing up on the draft task list. With both [FiveNum] and [Mutex] fixed too, all should now be well :) --PureFox (talk) 00:54, 12 January 2020 (UTC)

I was going to look into it when I got some spare time.   I had thought there might be   two   versions (separate entries), one was a draft task,   the other (slightly spelled different perhaps?) a (full) task,   and I was going to compare the complete contents of one versus the other.   I don't know what I would've found, if anything.     -- Gerard Schildberger (talk) 06:37, 12 January 2020 (UTC)

Good work on sorting out the sub-pages that were causing some tasks to be double-counted which seems to have worked fine. There's one sub-page which still needs some attention which is the aptly-named Psyco which is showing as a task completed by Python even though it's a now defunct JIT compiler for that language. --PureFox (talk) 15:05, 20 January 2020 (UTC)

Just checked and it's not in the overall Task Lists and is not therefore showing as a task authored by Hajo. --PureFox (talk) 15:19, 20 January 2020 (UTC)

Fixed Psyco. There's still a few double counted or misrepresented task entries around. As I find them, I'll fix them. If you notice one, either fix it or point it out to me. Thanks. --Thundergnat (talk) 23:12, 20 January 2020 (UTC)

As far as misrepresented task entries (for counting), there are several, and I do not know where they all are.   Such entries are where all the   BASICs   were lumped into one section,   and each flavor of   BASIC   was then entered under a sub-section   (most often on a different page),   and those sub-sections of BASIC aren't counted   (as per all the types/flavors of BASIC).   One such Rosetta Code task is   99 Bottles of Beer, (the BASIC section)   where they were all clumped together under BASIC.   This, if I remember, was part of a campaign to reduce the size of some (large) Rosetta Code pages by grouping the   BASICs   into one (sub-section),   and then have a   BASIC   page for that Rosetta Code task.   I have no idea how deep that project went.   One by-product of that is,   some Rosetta Code tasks have   (for instance)   Wee Basic   under   BASIC,   others have it at the end of the programming language list, under the   Ws,   making it a tad harder to locate a particular BASIC when browsing Rosetta Code tasks solutions.   But the bottom line is that most flavors of BASIC are under-counted.   And some   BASICs   like Locomotive Basic   (note the difference in the capitalization of BASIC in these last two examples),   according to some,   shouldn't even belong under the BASIC umbrella).     -- Gerard Schildberger (talk) 00:01, 21 January 2020 (UTC)

<quote>If you notice one, either fix it or point it out to me.</quote> Ok, let me put some caveats on there. The whole Basic debacle is a can of worms I'm not willing to open right now. IMO it would have been better off to just give every dialect its own task header and not try to "group" Basic dialects together, but that was decision made and partially implemented by others and would be nightmarish to try to untangle at this point. Especially trying to do it manually through the web interface. So yeah, you're sure right about that, but I'm not taking it on right now. --Thundergnat (talk) 01:18, 21 January 2020 (UTC)