From kwrite-devel Wed Dec 04 11:28:12 2013 From: =?ISO-8859-1?Q?J=2E_Pablo_Mart=EDn_Cobos?= Date: Wed, 04 Dec 2013 11:28:12 +0000 To: kwrite-devel Subject: Re: Non-free file in pate js_utils plugin Message-Id: X-MARC-Message: https://marc.info/?l=kwrite-devel&m=138615650523503 MIME-Version: 1 Content-Type: multipart/mixed; boundary="--===============4488851013563061650==" --===============4488851013563061650== Content-Type: multipart/alternative; boundary=001a11c353005ff64d04ecb3b5ac --001a11c353005ff64d04ecb3b5ac Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: quoted-printable 2013/12/4 Philipp A. > 2013/12/4 Kevin Kofler > >> Unfortunately, that is not true. (Wikipedia is right.) The issue is that >> JSHint is derived from JSLint, so the code derived from JSLint is of >> course >> still covered by the JSLint license, considering that the author refused >> any >> relicensing requests (except, ridiculously, IBM's). > > > not exactly. i think i=92ve figured it out: > > JSHint stable (2.x) has a main file that=92s derived from JSLint. and tha= t > is under doug=92s license. (so we can=92t ship JSHint stable) > > HOWEVER, JSHint is currently being rebuilt, and as a result, jshint.js wa= s > totally rewritten =96 so no derived code remains in JSHint=92s master bra= nch. > > so we can ship the unstable JSHint version if we want. also, as you said, > there are alternatives =96 while i don=92t think OCaml and Java code is > practival to be included, =93JavaScript Lint=94 is python! > > 2013/12/4 J. Pablo Mart=EDn Cobos > > If jslint is an external dependence like the before solution is this a >> problem? >> >> When Alejandro Blanco and I coded this feature we used a wrapper coded b= y >> him for this reason: >> >> https://pypi.python.org/pypi/pyjslint >> > > no, because the code still calls into code using doug=92s license: > > pyjslint *says* it=92s BSD licensed, but it still downloads JSLint, so wh= en > using it, the user ends up with nonfree code on his PC. so pyjslint > basically lies about its license. > > pyjslint is python wrapper of JSLint, and its code is BSD-licensed. It is true that it downloads jslint and invokes it using node.js, but how do changes the fact that the python wrapper is BSD-licensed? jslint is released under a modified MIT license. It is arguably if that license is free software or not. But even in the case of it not being free (arguably), why can't a BSD-licensed software interact with it? Nowhere in the BSD license is that restriction. And since it is pyjslint who downloads the jslint file, it is not present in the KDE repository, so KDE is not distributing an arguably non free software, even if the jslint license grants freedom of distribution. > _______________________________________________ > KWrite-Devel mailing list > KWrite-Devel@kde.org > https://mail.kde.org/mailman/listinfo/kwrite-devel > > --001a11c353005ff64d04ecb3b5ac Content-Type: text/html; charset=windows-1252 Content-Transfer-Encoding: quoted-printable
2013/12/4 Philipp A. <flying-sheep@web.de>
2013/12= /4 Kevin Kofler <kevin.kofler@chello.at>
Unfortunately, that is not true. (Wikipedia is right.) The= issue is that
JSHint is derived from JSLint, so the code derived from JSLint is of course=
still covered by the JSLint license, considering that the author refused an= y
relicensing requests (except, ridiculously, IBM's).
<= br>
not exactly. i think i=92ve figur= ed it out:

JSHint stable (2.x) has = a main file that=92s derived from JSLint. and that is under doug=92s licens= e. (so we can=92t ship JSHint stable)

HOWEVER, JSHint is currently being reb= uilt, and as a result, jshint.js was totally rewritten =96 so no derived co= de remains in JSHint=92s master branch.

so we can ship the unstable JSHint version if we want. also, as you said, t= here are alternatives =96 while i don=92t think OCaml and Java code is prac= tival to be included, =93JavaScript Lint=94 is python!

2013/12= /4 J. Pablo Mart=EDn Cobos <goinnn@gmail.com>

If jslint is an external dependence like the before solution is this a= problem?

When Alejandro Blanco and I coded this feature we used a wra= pper coded by him for this reason:


no, because the code still = calls into code using doug=92s license:

pyjslint *says* it=92s BSD licensed, but it still down= loads JSLint, so when using it, the user ends up with nonfree code on his P= C. so pyjslint basically lies about its license.


pyjslint is python wrapper of JSLint, = and its code is BSD-licensed. It is true that it downloads jslint and invok= es it using node.js, but how do changes the fact that the python wrapper is= BSD-licensed?

jslint is released under a modified MIT license. It is = arguably if that license is free software or not. But even in the case of i= t not being free (arguably), why can't a BSD-licensed software interact= with it? Nowhere in the BSD license is that restriction. And since it is p= yjslint who downloads the jslint file, it is not present in the KDE reposit= ory, so KDE is not distributing an arguably non free software, even if the = jslint license grants freedom of distribution.
=A0
____________________________________________= ___
KWrite-Devel mailing list
KWrite-Devel@kde.org
https://mail.kde.org/mailman/listinfo/kwrite-devel


--001a11c353005ff64d04ecb3b5ac-- --===============4488851013563061650== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ KWrite-Devel mailing list KWrite-Devel@kde.org https://mail.kde.org/mailman/listinfo/kwrite-devel --===============4488851013563061650==--