X-Authentication-Warning:	delorie.com: mail set sender to geda-user-bounces using -f
X-Recipient:	geda-user AT delorie DOT com
X-TCPREMOTEIP:	207.224.51.38
X-Authenticated-UID:	jpd AT noqsi DOT com
Mime-Version:	1.0 (Mac OS X Mail 7.3 \(1878.6\))
Subject:	Re: [geda-user] About reinventing the wheel, and how to avoid it
From:	John Doty <jpd AT noqsi DOT com>
In-Reply-To:	<alpine.DEB.2.00.1509121245420.3577@lichen>
Date:	Sat, 12 Sep 2015 09:51:12 -0600
Message-Id:	<9D45499A-BC67-4B2F-A285-34C49ECC20B0@noqsi.com>
References:	<alpine DOT DEB DOT 2 DOT 00 DOT 1509111150410 DOT 2482 AT lichen> <B94E2C24-8B63-46FC-A551-9E4B14B817C9 AT noqsi DOT com> <201509111624 DOT t8BGOPYV000685 AT envy DOT delorie DOT com> <F24E14C9-8427-4688-A27A-1EBE0F669995 AT noqsi DOT com> <201509111732 DOT t8BHWnF7005271 AT envy DOT delorie DOT com> <960028A6-F16B-4276-B838-E9F0D2959A6F AT noqsi DOT com> <55F32375 DOT 2020505 AT ecosensory DOT com> <FC9C2933-89CE-44FD-A9EE-CA2721405ACE AT noqsi DOT com> <201509111941 DOT t8BJfqWl010427 AT envy DOT delorie DOT com> <D3D5A923-FAF9-4665-A0DB-A8D6A8BEA5A3 AT noqsi DOT com> <CA+uY=MS5ZOTHXDe9taav+0tCiZOqhtAJU2KO-xyRA3SrQVvLNw AT mail DOT gmail DOT com> <201509112020 DOT t8BKKBgI012564 AT envy DOT delorie DOT com> <E872F1BC-8FA7-449E-BA6D-5841AC79A465 AT noqsi DOT com> <201509112058 DOT t8BKwF3b013774 AT envy DOT delorie DOT com> <6341DF3E-543D-4E36-9B01-6B5B950208C4 AT noqsi DOT com> <alpine DOT DEB DOT 2 DOT 00 DOT 1509121245420 DOT 3577 AT lichen>
To:	geda-user AT delorie DOT com
X-Mailer:	Apple Mail (2.1878.6)
X-MIME-Autoconverted:	from quoted-printable to 8bit by delorie.com id t8CFpK61014399
Reply-To:	geda-user AT delorie DOT com
Errors-To:	nobody AT delorie DOT com
X-Mailing-List:	geda-user AT delorie DOT com
X-Unsubscribes-To:	listserv AT delorie DOT com

On Sep 12, 2015, at 9:18 AM, Roland Lutz <rlutz AT hedmen DOT org> wrote:

> On Fri, 11 Sep 2015, John Doty wrote:
>> things like gnet-spice-noqsi have *zero* impact on what youĒre doing. ThatĒs a consequence of good factoring.
> 
> In the context of Free Software, a model where people contribute their code back upstream usually turns out to work better.  Where's the difference?  If your code is in the main codebase, other users can find it, and even more importantly: the developers are aware of it.

IĒm fonder of the model that Python and LaTeX have, where the core code is separate from the application modules. We kind of have that with gedasymbols.org (thank you DJ!), and I have been remiss in failing to publish gnet-spice-noqsi there. The Python and LaTeX communities have done a better job of documenting and indexing their add-ons than we have, though.

> 
> If there is for example an API change, a simple grep will reveal the problematic code line; the developer can fix it right in the same commit.

Test cases are important, but I donĒt think itĒs necessary to have add-on modules maintained with the core sources in order to test them together.

> 
> You may have noticed that I included your spice-noqsi backend, along with Igor2's dump backends, in the Xorn repository.

Yes. IĒm very impressed. That was a fair amount of work.

>  (Thanks for putting it under the GPL which made this possbile!)  I had a twofold motivation for that: by including these backends in the regular test suite, I could be sure I didn't break anything by accident; and after updating fourty other backends, it was much easier (and less error-prone) for me to update another one than if I had ignored and left that task to you.
> 
> In the end, everyone benefits from such a model: the contributors aren't left alone with their incompatible code (see Firefox for a negative example on that), the users have non-bitrotten code even long after the original contributor has left the project, and the developers need to worry much less about what code outside the repository may break with a change.

John Doty              Noqsi Aerospace, Ltd.
http://www.noqsi.com/
jpd AT noqsi DOT com

- Raw text -