Are only software-specific questions allowed?

Question

I am on this site already for several years, and before (when I had more free time to give answers) I saw a lot of good questions asking how to solve a problem using whatever software, and people were suggesting solutions with QGIS, ArcGIS and other tools, and then the topic-starter was choosing the best way for him. Such questions have collected a lot of useful tips over years, like these:

• Calculating average width of polygon?
• Removing small spaces (slivers) between polygons?
• Creating large amount of random points in binary raster?

Recently this has changed? Such conceptual questions are not allowed anymore? Old questions started to get closed like these:

• Calculating all distances from single points to multiple polygons; and
• Programmatically adding field to shapefiles and inserting shapefile name in every row of field?

New questions are closed immediately like Shortest path creation between two lines (where an answer could be to use a centerline, and then he could research himself how exactly to do this). And under this question there is a comment

you need to cut and paste your question into four separate questions, then edit so each one says precisely one of "using ArcGIS", "using Matlab", "using QGIS" or (if you want) "using R". Them's the rules.

Maybe it was a joke. For me doing this feels a bit ridiculous.

I noticed this situation when I asked my own question recently (where I need a mathematical algorithm in the first place but if necessary I can just mark it ArcGIS-only)

• Detecting branched polygon shape?

So, I just would like to clarify if there was such change in rules (or maybe I was misunderstanding rules before), in order to ask here allowed (one-software but not theoretical) questions in future.

Answer 1

As a moderator, who uses several software programs on a daily basis, I admit I struggle with this myself sometimes. However, when the site was in its infancy, broad questions were allowed. This is not the case so much anymore, as we are a focused Q/A site. By tagging multiple software packages, it is essentially asking multiple questions in one, which is not allowed.

Sometimes I wish that I could ask one question with multiple programs, because in the end, I don't really care which one I use, I just need to get the problem solved. But, it is generally agreed (and there has been much discussion about it, which I don't have time to search for now) that limiting a question to a single program is best.

The question you linked to not only tags multiple programs, but also doesn't show much in the way of effort undertaken. If the OP included what they had tried (with screenshots and code for each program they use, I may be inclined to bend the rules on that one. However, as it stands, they just gloss over the details.

Answer 2

I think I agree most with @Aaron comments. However, I think over the years GIS and GIS related programming have evolved rapidly - or perhaps my own understanding of it all has grown with experience.

Now there are many open-sourced solutions to do GIS work with an expansion of more capabilities to perform GIS with not only a graphical interface, but line-by-line. When I first got started in GIS there was ArcGIS and GRASS. The questions could be answered with one software type because there were limited options available. Now we have QGIS, GDAL, Python, R, etc. Which all excel at different parts of spatial analysis, some are quicker to perform certain tasks, some easier to learn, and most of all they are free. Many of us use multiple programs and languages in our workflow and choose which we use based on a certain problem or limitation with software/language we encounter. More experienced answerers may have a reasoning for choosing these methods.

I myself have provided answers 1 2 where I disregard the OP's software tags because it's easier to do in R or there's a better tool in QGIS.

In cases where a question does not provide a specific software - but is comfortable with most answers we shouldn't close the question. Perhaps as more programs are freely available and easier to set up we can open up our community restrictions to be less so - and allow questions with no set software requirements as it limits the community - who are more than capable than using more than one proprietary software. Closing questions for not choosing a specific tag for software is limiting the exposure and potential answers that may help OP. Additionally, asking OP to create separate questions for each software or package would toe the line of being closed as a duplicate -- and would be excessive.

A different scenario is one where a user has not specified a specific software because they are new and forgot a tag or are inexperienced or limited to one software - where the quality of the question could be improved with being more specific.