Yes, it's true. The reason may be nostalgia, but a gopher: URL is sexy. However, I see no reason for the connection between protocols and data types that many posters as well as the manifesto writers make. HTTP can serve plain text data as well as HTML. If you put plain text data on an apache server, the performance and network load will be only marginally worse than with gopher. Every good http server will also be able to generate directory pages automatically if there is is no index.html or equivalent in a directory. Thus, while the HT in HTTP stands for hypertext, gopher-typical contents can be served over HTTP as well.
However, Gopher seems to have one severe problem, namely the data type descriptors. While HTTP users enjoy all the benefits of MIME data types, Gopher allows only single letter descriptors. For all data types that the Gopher makers have not forseen, the data type can at best be guessed from file name extensions or by magic numbers. Gopher+, of course, addresses this problem, but does so in a somewhat artificial way.
To come to a (not all that conclusive) conclusion: On the side of protocols, http has a clear advantage by the incorporation of MIME types. It makes a good deal of sense, however, to serve up information by more that one protocol if the types of information served lend themselves easily to being served in different ways. Why, after all, should the same text file not be available via HTTP, Gopher+, and FTP? On the side of content types, it is certainly true that plain text is underused today althogh it is very suitable for shorter documents that don't make a whole lot of references either within the document or to external documents. HTML is certainly more powerful, and hopefully we will see the standardization of other SGML languages in the future that will provide for better possibilities to mark up things like academic papers semantically. Looking at images, Gopher's native GIF format can probably be considered dead thanks to the strange efforts of Unisys. But it certainly is worthwile to discuss the question whether Gopher should remain in use as a protocol separately from the question of abuses of HTML.
Yes, it's true. The reason may be nostalgia, but a gopher: URL is sexy. However, I see no reason for the connection between protocols and data types that many posters as well as the manifesto writers make. HTTP can serve plain text data as well as HTML. If you put plain text data on an apache server, the performance and network load will be only marginally worse than with gopher. Every good http server will also be able to generate directory pages automatically if there is is no index.html or equivalent in a directory. Thus, while the HT in HTTP stands for hypertext, gopher-typical contents can be served over HTTP as well.
However, Gopher seems to have one severe problem, namely the data type descriptors. While HTTP users enjoy all the benefits of MIME data types, Gopher allows only single letter descriptors. For all data types that the Gopher makers have not forseen, the data type can at best be guessed from file name extensions or by magic numbers. Gopher+, of course, addresses this problem, but does so in a somewhat artificial way.
To come to a (not all that conclusive) conclusion: On the side of protocols, http has a clear advantage by the incorporation of MIME types. It makes a good deal of sense, however, to serve up information by more that one protocol if the types of information served lend themselves easily to being served in different ways. Why, after all, should the same text file not be available via HTTP, Gopher+, and FTP? On the side of content types, it is certainly true that plain text is underused today althogh it is very suitable for shorter documents that don't make a whole lot of references either within the document or to external documents. HTML is certainly more powerful, and hopefully we will see the standardization of other SGML languages in the future that will provide for better possibilities to mark up things like academic papers semantically. Looking at images, Gopher's native GIF format can probably be considered dead thanks to the strange efforts of Unisys. But it certainly is worthwile to discuss the question whether Gopher should remain in use as a protocol separately from the question of abuses of HTML.