Anonymous user
Rosetta Code:Village Pump/Uses Algorithm Template: Difference between revisions
Rosetta Code:Village Pump/Uses Algorithm Template (view source)
Revision as of 11:21, 10 November 2010
, 13 years agoAdd template
No edit summary |
m (Add template) |
||
| (2 intermediate revisions by 2 users not shown) | |||
Line 1:
{{Vptopic
|topic=Uses Algorithm Template
|summary=Whether it is justifiable to have a template for characterizing the tasks that are focused on particular algorithms, and what are the consequences of this.
}}
I have often struggled with tasks which specify that a specific algorithm be used.
Line 21 ⟶ 25:
:::: TBH, TLDNR, and there may have been elements of an SEP field; I'm not so focused on implementation details as end result, so I try to let RC users figure things out. What you're describing with the hierarchy sounds like a simple transformation of what I was thinking, but I'd stop short of forcing the automatic creation of parent tasks when the itch at hand is a particular algorithm or solve a particular scenario. It forces too much empty space too quickly for the few contributors who work at implementing ''everything'', and the site already has a bit of a problem with the number of tasks with unnecessarily-small example sets (A task that doesn't contain a language the viewer is familiar with isn't helpful for comparative purposes.). Does that make sense? --[[User:Short Circuit|Michael Mol]] 18:49, 22 June 2010 (UTC)
::::: Ok, so... as a suggestion or guidline this might work but it should not be rigidly enforced nor implemented as a structural requirement? --[[User:Rdm|Rdm]] 19:55, 22 June 2010 (UTC)
:::::: As is convenient to create them, I can see it manifesting as a set of "Related tasks" links. "More general forms of this task" "More specific forms of this task" Some of the existing tasks with distinct optional components might easily break out into that latter set. But this isn't a decision I want to make by fiat. I would '''specifically''' like to see some more input from other regular contributors. --[[User:Short Circuit|Michael Mol]] 02:10, 23 June 2010 (UTC)
: There's value in both having “implement this algorithm” and “achieve this goal” type tasks; they serve different requirements. Implementing a specific algorithm is about doing detailed comparison between broadly-similar languages (e.g., I'd use them to study the differences between [[Ruby]] and [[Python]], but probably not between either of them and [[Prolog]] as they just have different approaches to considering the task). Achieving a goal is more about showing how there are different ways to achieve something and is suited to comparing languages that are quite different (whereas largely similar languages will probably not show much difference at that point). What there ''should'' be is semantic links between the tasks; e.g., one task for doing language-idiomatic sorting, many other tasks for specific sorting algorithms, and relations so that everyone knows that in production code it is probably best to use the idiomatic form. I'm not sure that we can maintain all these things automatically yet, but manual maintenance is better than nothing. –[[User:Dkf|Donal Fellows]] 11:18, 10 November 2010 (UTC)
| |||