What is the "right" way to search for a package/class to solve a random problem? Whenever I need to find a new package or class, I never feel like I am doing it efficiently. I tend to use a mixture of CTAN, Google, and TeX.sx. The issue I have with CTAN is it doesn't provide any guidance on quality. Google tends to be filled with how-tos of questionable quality from various eras. TeX.sx seems to give the most up to date answer with a voting based judgment of quality (e.g., this question on CVs), but only when the question has been asked.
|
I happen to have reflected on this question. :-) The packages that have repeated functionality is not a large percentage of the archive as a whole, but for jobs that people often want to do you do see redundancy. We can imagine a number of attacks on that issue. For instance, CTAN could track downloads and rank related packages that way. The issue here is that distributions such as TeX Live include the most widely-used packages so for instance the often-recommended geometry would be rarely downloaded from CTAN. Or a web crawler could look at online docs to see what packages are actually used. That has the advantage of being automated, but an immediate difficulty is that many documents on the web are ancient and their package choices limited by their creation date, even assuming that majority vote is a good recomendation. A person could think to weigh more recent docs more highly .. but I don't know of any effort on this. Maybe there could be some peer review process. That might help improve packages as well as give a person some recommendations. One issue here is that it could create another layer of documents to wade through and the great number of TeX users are already struggling with complexity. But the main issue is: who will bell this cat? Finally (its the last idea that I have but no doubt someone else has better ones) we could have people write articles for TUGboat, etc., that give advice. I have this one (PDF) but it is getting old and anyway such articles by others would be a good thing. (Alternatively, such an article could be a wiki-type thing). One disadvantage here is that over time the "best" choice can change. In addition it has the disadvantage that some recommendations are just choices. For instance, I like Asymptote while others like TikZ and to some extent a recommendation is arbitrary. |
|||||||||||||||||
|
|
With "vanilla" TeXLive, you can use Example:
Answer:
|
||||
|
|