← Back to team overview

kicad-developers team mailing list archive

Re: boost_root driven by https


On 31 August 2013 22:06, Dick Hollenbeck <dick@xxxxxxxxxxx> wrote:

> > I just upgraded to openssl 1.0.1e, and am now building the DLL fine in
> a-mingw-us,
> > although I think this 'e' version had nothing to do with this.  Seems
> the Configure script
> > is picky about sequence of arguments.
> >
> >
> > We now need to package the openssl DLL goodies so they can flow into a
> situation where
> > windows builders of GITHUB_PLUGIN can use those.
> >
> Come to think of it, I don't know how I am going to find the openssl
> headers on Windows
> for the GITHUB plugin.  Certainly we cannot expect them to flow with a
> KiCad installation.
> Maybe we will simply need to reproduce the ExternalProject_Add() in the
> GITHUB plugin
> CMakeFile, conditioned on Windows.   Gosh yes, what was I thinking?
> How you guys stay on Windows is beyond me.   The package management
> concept has long been
> an established concept.  But "add/remove programs" is a joke by comparison.
> Then, you have competing C++ compilers so the story would have be told in
> duplicate
> anyways.  I don't expect MSVC to be around much longer, given that it does
> not generate
> revenue (not like the Android patent blackmailing is).    And Android is
> showing up
> everywhere now, including wrist watches.  I give MSVC about another 6
> years max.
> How good can that be for your character, knowing that most of your revenue
> is coming from
> patent blackmail?  Maybe Mr. B is retiring because of that.  (Note I am
> assuming he has
> that much quality of character.)
> You hear that New Zealand just outlawed software patents?  Call your
> representatives
> please.  It will take a lot of voices to drown out the voice from
> Microsoft, which as I
> mentioned, now gets a lot of patent revenue from Android OEMs, who are not
> strong enough
> to fight back.

Yes, I thought about the header/dependency situation for SSL after your
previous email. I understand your frustration, believe me!

Perhaps Apple will stop suing people for making black boxes without their
permission, or charging people for manufacturing a bluetooth device that is
to connect to one of their devices too. But it'll never happen. The cogs
will keep turning and we'll have to sort the crap out.

With regards to the compiler situation; As much as I don't really
appreciate MSVC, the execution speed differential of GCC (with -O2
optimisations) and MSVC compiled wxWidgets samples is incredible. MSVC wins
by an order of magnitude, especially for start-up times. I wanted to
investigate the ELF header information. Apparently GCC can be a bit overly
verbose with regards to ELF information which can slow load times

Anyway, I'm kind of glad to hear the scripting window is not uniquely
broken on Windows builds. In that case, I can forget about fixing it before
releasing scripting through KiCad-Winbuilder. So I can get on and release
that ASAP then. :)

Best Regards,


Follow ups