Summary

New tag for "limited solutions"

Discussion

In some instances, it may not be possible for a language to provide an exact match solution to the task, due to limitations of the language or its utilized subcomponents. It might be a good idea to have a "limited solution" marker would be a good idea. (Similar to the "incorrect solution" tag), that reads something like 'The solution provided does not meet the exact specifications of the task, due to limitations of the language or its subcomponents. The author has tried to provide an "approximate fit" or "best fit" solution based on those limitations'. I know there are tasks on the wiki, where the solutions is a close approximation, rather than an exact fit. This might be quite useful. This would give better scope for making comparisons of languages, because it would be possible to see the limitations of the language and it tools for a particular task, (rather than the implementation being simply omitted, giving us nothing to use for evaluation of the language at all). Markhobley 09:55, 24 July 2011 (UTC)

Unfortunately, the example for which you would like to add the tag: Pi#GUISS is reason to reject the notion as one gain so little when comparing that language example with the others. That page would be better if the GUISS example were dropped altogether. In light of this, I would rather we continue our current "ad-hoc" methods as we don't currently have a need to go beyond that. --Paddy3118 12:36, 24 July 2011 (UTC)

There are some other examples: some of the timer related tasks, file handling tasks, I think some of the UNIX shell solutions, Unix/Windows incompatibilities, levels of support with different concepts, etc. I'll paste any that I find. Markhobley 14:23, 24 July 2011 (UTC)

The REXX solution for "Flatten a List", works with a string representing a list rather than a list itself. That is a close approximation (and acceptable IMHO). Markhobley 16:43, 24 July 2011 (UTC)

I see this as a consequence of your Pi#GUISS "solution" being rejected and you are trying yet another way to weasel out of it. I don't think there is any place for such a template. For a task, there are correct solutions, and there are not-so-correct soutions written in good faith. What constitutes "in good faith" is subjective, so people must be allowed to scrutinize them instead of letting them hide under some umbrella tag. Case in point: by most people's standard, the GUISS pi digits "solution" would not be a good faithed attempt (want a bet?), yet you still want a tag there so you can point at it and say "don't remove it, it's tagged such, thus acceptible". --Ledrug 19:27, 24 July 2011 (UTC)

It's nothing to do with that. I am trying to make the documents as good as they can be, and I think being relaxed and inclusive is the best way forward. I don't think we need to be quite so strict with an "this is not an exact match, so it must be deleted" type of approach. After all, the solution was valid, whether it gets published here or not. I can always publish it on another website. That is not a problem. It's just convenient to work with Rosetta Code, because it has similar goals to a project that I have been developing, and the infrasture is already in place here, and I quite like how the project is evolving. FWIW, a policy would alleviate the need for the tag, or it is easy to add "If your language cannot achieve this task exactly, then make an approximate implementation, and state divergencies and limitations in the provided solution" to every task. But I think this is extreme. A policy would be sufficient (if only really for the purpose of clarification of conflict between different developers ideas). Markhobley 20:47, 24 July 2011 (UTC)

Policy

We could just have a policy that allows approximations, if an exact fit is not available, providing that the solution states the limitations and any differences in behaviour from the task requirements. Markhobley 16:49, 24 July 2011 (UTC)