On Tuesday, September 8, 2015 at 7:02:32 PM UTC-4, Mouse wrote:
>
> Crypto++ is a library that is not useful unless linked to some 
> application. IMHO this means that only those who build applications 
> themselves could/should have interest in it. And if a person is able to 
> compile his application - it shouldn’t be too big a big deal for him to 
> compile a library provided in a source form. 
>
> As an alternative, I’d like to point out several software repositories 
> that serve *and maintain* Crypto++ in binary form (as a compiled library 
> and a set of header files):
>
>    - Macports (for Mac OS X)
>    - Homebrew (for Mac OS X)
>    - Ubuntu Software One (or whatever’s the name of the main repo)
>    - CentOS (RPM)
>    - …probably others… 
>
>
> If there’s no such repository for Windows, what can I say? Crypto++ 
> developers/maintainers have neither time nor desire (and probably not the 
> ability) to maintain binary releases for all the platforms Crypto++ is 
> designed to run on. Windows platform definitely won’t be an exception.
>
 
IMNSHO, I think the prebuilt binaries are a good idea.

I think the pain point is we have not adopted a strategic approach and no 
one has stepped up to offer them.

There's a secondary issue. If someone does step up, then users could suffer 
the "web effect", where potential garbage shows up all over the web like a 
feral dog sprays its urine.

I've observed the "web effect" with OpenSSL for years. Bad advice shows up 
everywhere, and bad binaries built on the bad advice show up everywhere. It 
was one of the reasons we (the OpenSSL devs and a handful of interested 
folks) started the OpenSSL wiki (we also had good results based on the 
Crypto++ wiki, and I was able to relay them when arguing for their wiki).

The third underlying issue is stale/outdated repos. I've tried to address 
that with friendly relationships with maintainers, like the Fedora, Ubuntu 
and Debian maintainers. If we make it easy on them and keep them informed, 
then they will be more likely to pickup our changes.

A fourth and even more subtle reason is distros usually carry the latest 
stable, and not the cutting edge stuff. So unless we release it and the 
distro deems it useful, then it won't be picked up. We are positioned to 
begin moving things forward, so this will be more important in the future.

*****

When I get some spare cycles, I want to address the strategic gaps and get 
process in place so we can do the prebuilt binaries.

I want to create them for Windows, and add a Debian or Ubuntu PPA 
(https://askubuntu.com/questions/4983/what-are-ppas-and-how-do-i-use-them). 
I also want something for Red Hat/CentOS/Fedora. Finally, I want to supply 
the cross-compile stuff.

Right now, there are two items that need to be addressed. First, we need a 
PGP-like long term signing key for releases, but we have not hashed out the 
details on its storage, access and use.

Second, Windows is holding me up because Visual Studio Community is an 
expiring trial, so I can't do things like convert a project and build the 
artifacts from the command line using msbuild.exe. (Related: 
http://stackoverflow.com/questions/32438478/how-to-upgrade-a-solution-and-project-files-from-the-command-line).

Jeff

-- 
-- 
You received this message because you are subscribed to the "Crypto++ Users" 
Google Group.
To unsubscribe, send an email to [email protected].
More information about Crypto++ and this group is available at 
http://www.cryptopp.com.
--- 
You received this message because you are subscribed to the Google Groups 
"Crypto++ Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to