Mail Archives: geda-user/2016/02/29/14:15:49
--Apple-Mail=_5D43DDBC-6365-4BAC-9E4F-C856790BD1D2
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
charset=windows-1252
On Feb 29, 2016, at 11:05 AM, DJ Delorie <dj AT delorie DOT com> wrote:
>=20
>>> Hey, it's software, at least you *can* rewrite it to do anything. =
You
>>> missed the message completely: It's *difficult* to do. Perhaps you
>>> could put some effort into making it easier?
>>=20
>> There's a distinction between rewriting and scripting.
>=20
> You missed my point completely. Any software can be, at the worst
> case, rewritten to do anything else. Gschem can be rewritten to be a
> video game. It's possible. It's *difficult*. Saying something is
> "possible" is a weak excuse to not make it easier.
You missed my point completely: nobody needs to rewrite gschem to =
represent any circuit. This is in contrast with pcb, which must be =
rewritten to handle any printed circuit structure that it doesn=92t have =
a specific feature for.
>=20
>> Who are the target users?
>=20
> At this point, I'd like to say we're targetting the people who are
> already using gEDA and trying to make it work for them. People are
> complaining and you're not listening.
You mean =93pcb consumers are complaining=94. And there=92s a serious =
disconnect between the design of gschem and what pcb consumers expect. =
But pcb consumers expect an Accord, where gschem is a Wrangler. The =
solution isn=92t to change the Wrangler into an Accord: Wranglers are =
better when the going gets tough.
>=20
> It really sounds like you want the target users to be "John Doty" in
> which case, fork the tools and go away.
>=20
>> It used to be "anybody computer-literate who's using any tool that
>> can read a netlist",
>=20
> But the people who are trying to consume that netlist want the whole
> process to be easier, and you aren't listening to them. Where is the
> spice model database?
Copyrighted, unfortunately. You can=92t fix this by changing gschem.
> Where is the osmond back-annotation tool?
I could put that on gedasymbols after some cleanup. So many things to =
do. Changing gschem won=92t help here, either.
> You're using pcb as a whipping boy to avoid making the tools better
> for everyone.
You have a different metric of better.
>=20
>> Because I think there should be a toolkit for those of us who need
>> the open-ended capability that the Unix philosophy leads to.
>=20
> I agree, but why make that toolkit as difficult as possible for people
> to use? Why can't you listen to the people who *want* to use the
> toolkit, who are complaining that it's difficult to use?
Windows consumers complain that Linux is difficult to use. Does that =
mean that we should make Linux more like Windows? Word consumers =
complain that LaTeX is difficult to use. Does that mean that LaTex =
should be more like Word?
It is very common to overestimate the value of complexity and =
underestimate its costs.
>=20
>> Imitating KiCAD is a dead end.
>=20
> Bringing up the KiCAD strawman is yet another excuse.
But one common complaint is that KiCAD is more popular. Do we want to be =
popular and limited like Windows, or unpopular and powerful like Linux?
>=20
>> No, it's just like using an editor to create program source code, a
>> page at a time. The editor is not responsible for the things that
>> make, the compiler, and the linker do. That's good modularity.
>=20
> Then your argument is that gschem is like the old "edlin" that ONLY
> edits files, and leaves everything else - like search and replace - to
> scripts, where people want something like emacs or vi that have syntax
> highlighting and a robust basis for expansion. See? I can abuse a
> metaphor too.
Emacs does this with scripting in Lisp. This stuff isn=92t wired in. =
Gschem supports the emacs approach.
>=20
>> But it's the cost of a
>=20
> No, that's an exuse again. If you want a hard-to-use toolkit that
> only the Supreme Chosen People can use, you won't find a willing flock
> here. A toolkit can be flexible and extensible without being sparse
> and obtuse.
Gschem isn=92t obtuse. Sparseness is a virtue, making the tool easier to =
use. One great thing about gschem that I appreciated from the beginning =
is its simplicity. For Viewlogic, I could spend all day discovering that =
none of the many features did what I needed. Features have cognitive =
costs and trip over each other (why is adding buried vias to pcb so =
hard?).
>=20
>> You want Word, I want LaTeX.
>=20
> I want a schematic editor. You want a paint program.
>=20
I want something as powerful and productive as gschem actually is. You =
want a toy.
John Doty Noqsi Aerospace, Ltd.
http://www.noqsi.com/
jpd AT noqsi DOT com
--Apple-Mail=_5D43DDBC-6365-4BAC-9E4F-C856790BD1D2
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
filename=signature.asc
Content-Type: application/pgp-signature;
name=signature.asc
Content-Description: Message signed with OpenPGP using GPGMail
-----BEGIN PGP SIGNATURE-----
Comment: GPGTools - https://gpgtools.org
iQIcBAEBCgAGBQJW1JjMAAoJEF1Aj/0UKykRQ4oP/1Ke93z+FX9hC0innnS8sehJ
HXxYbzDRNatecEmLgzD6y9+4G2mWsKZxVNXh9HMCh2Owj3MS6C+YL/5mwYoUAwfR
eNfppREGyWQ4qYi+Hpt9e9H4nwALKnA/ivX4zxlrO1gKbuLqQ/AtxnNehbwIQnh1
Nlg2g2zvYL2MYC6CPEtBBaQfKUJsuEEthdy4occl6mPPS3yePug7hSwCxqZ/okko
J8kqmYzSWxLVSQcFIf2IMJOzCl+dcrpIfWhvyNvAsl/1qsC1D1uLFmoiV309+IBI
aGMU4/D07JImB0gsr3D+TbTnFB6aMGZNcEg/c3WmI2kTRpHbRR5ypkIfF7h7pKOi
BYNzXYkTETy0Md5qLeXbSka3NF4adWmL3g/xnLdHYMYsT5BjAmGKRVw5Nnqo8kXm
imhodSLncgXgxuHoEqWiZ9PR22b74rTGy9nPuFjbFzAn2+CXlufchN+bQcvESkMu
QBMmeB3QysNQNXkQ/XLcQW4U1lAAsARi/woAwcLoiLkEtvzAFROAJrozdZEBobNB
GgF7eaRg0+zl2iJPQcMR3mqiiy49ygDcpSIRs6es65lfDMtGD1G8G7m4YU4Gblfk
5n8uMDPvL6x5haem0l3h4xpdsdcCOl7+ZNecUevlO+qXq5g5URnQR5PVUkIK6rNL
KBjuUhVjGWFnIIIW7IFX
=D67Z
-----END PGP SIGNATURE-----
--Apple-Mail=_5D43DDBC-6365-4BAC-9E4F-C856790BD1D2--
- Raw text -