Il giorno 21/dic/2011, alle ore 19:53, Adrià Cereto Massagué ha scritto: > 2011/12/21 Sebastian Hilbert <[email protected]> > On Wednesday, December 21, 2011 07:12:03 PM Hartmut Goebel wrote: > > Am 21.12.2011 00:19, schrieb Adrià Cereto Massagué: > > > While browsing the source codeI noticed that there's a variable called > > > 'datas' and the word 'datas' is used in the coments as a plural of > > > 'data'. > > > This is a grammatical error, since 'data' is already the plural of > > > 'datum', so 'datas' is like 'feetes', 'mices' or 'phenomenas'. > > > > You are absolutely right. But I'm afraid, we can not change it that > > easy, since `datas` is used in the spec-file. Thus correcting this error > > will break al existing spec-files. I would like to change this too, but > > the price seams to be to high for the benefit. > > Just a thought. Since the next major release potentially breaks spect files > anyway you could target the nexe major release. > > > I have some .spec files from 1.5.1 that do not work with current trunk > (mainly because of the need to specify there the path to _mountzlib.py). So I > don't think it would be such a high price for the next major release.
The fact that it broke doesn't mean that we agreed on breaking it, and thus does not justify further breakage. In my opinion, it makes sense to break .spec files only when we introduce the new-generation spec files that we have already designed; up to that point, I would avoid breaking them. For the data vs datas specific point, it is sufficient to accept both data and datas in Analysys and treat them as synonymous; then we fix Makespec to generate "data" instead of "datas". This way we fix it for new users and we save backward compatibility. We'd love a patch :) -- Giovanni Bajo :: [email protected] Develer S.r.l. :: http://www.develer.com My Blog: http://giovanni.bajo.it
smime.p7s
Description: S/MIME cryptographic signature
