The question of the nomenclatural availability of preliminary electronic versions of taxonomic papers, distributed online ‘ahead of print’ by some journals and publishing companies, is addressed again. We disagree with Krell's suggestion to ‘distinguish between content and bibliographical metadata’. The concept of ‘bibliographical metadata’ does not exist in the Code and should not be incorporated into it. The citation of publication date, issue and page numbers are part of the relevant information that is useful in bibliographic references, in text citations and in synonymic lists that appear in taxonomic publications (which have a much longer life than most other publications), and should be considered part of the ‘content’ of a taxonomic paper. In particular, the page of first appearance of a name or of a nomenclatural act is traditionally cited in taxonomic revisions, monographs, faunas and catalogues, whether printed or stored in online databases, and this information is very useful for serious taxonomists. Having two different sets of information in this respect, in a first version of the work first published online but then discarded from the website of the journal, and in a subsequent one included into a journal issue, would be an indisputable source of confusion, which would not be solved by calling artificially both these versions the ‘version of record’ as if they were identical. The fact that ‘pagination is not regulated’ by the Code is fully irrelevant here: many aspects of taxonomic works are not regulated by the Code but are of crucial importance for the discipline of taxonomy. Krell's proposal is motivated, according to his own words, by some publishers’ desire to make their publications more ‘attractive as outlets of taxonomic research’. But the purpose of the Code is not to make some journals more attractive than others; it is to help working taxonomists in their daily work, to make it easier, more straightforward, efficient, reliable and useful, and less prone to ambiguity. We recommend rejecting Krell's suggestion. On the other hand, we make the new proposal of the creation of a ‘label’ to which some online journals and publishing companies might adhere, taking the engagement to publish online only one single version of each paper, with its final date, issue number and pagination. Authors will then have the possibility to choose their publication outlet among those having this label or those following the practice of ‘early view’. Currently, such preliminary versions of papers are unavailable under the 2012 Amendment of the Code.