[concept_check] Reference page concept links out-of-date
Not sure if this has been reported before, but any links on the Reference page of the concept_check library that point to SGI.com now take you to a page that says 'SGI.com Tech Archive Resources now retired'. Regards, Spencer Collyer
On Sun, Apr 8, 2018 at 6:07 AM, Spencer Collyer via Boost < boost@lists.boost.org> wrote:
Not sure if this has been reported before, but any links on the Reference page of the concept_check library that point to SGI.com now take you to a page that says 'SGI.com Tech Archive Resources now retired'.
I've noticed this with links from other Boost libraries (e.g. Boost.uBlas)
to the SGI's former webpage. (Way to not go, Hewlett Packard.) I see a
few options to work around this:
1. Point the links to the last versions archived by the Wayback Machine <
web.archive.org>. It's the most immediate fix, but Boost might lose the
perception of being a cutting-edge set of libraries that doesn't point to
outdated web pages.
2. Point the links to corresponding documentation on the latest C++
standard
On Apr 8, 2018, at 12:59 PM, Cromwell Enage via Boost
On Sun, Apr 8, 2018 at 6:07 AM, Spencer Collyer via Boost < boost@lists.boost.org> wrote:
Not sure if this has been reported before, but any links on the Reference page of the concept_check library that point to SGI.com now take you to a page that says 'SGI.com Tech Archive Resources now retired'.
Unless you have HTTPS Everywhere installed and have previously visited http://www.sgi.com, in which case it continues to automatically redirect to https://www.sgi.com, which no longer exists. Arguably, a connection error is a better result than a 200 page which only adds insult to injury.
I've noticed this with links from other Boost libraries (e.g. Boost.uBlas) to the SGI's former webpage. (Way to not go, Hewlett Packard.) I see a few options to work around this:
1. Point the links to the last versions archived by the Wayback Machine < web.archive.org>. It's the most immediate fix, but Boost might lose the perception of being a cutting-edge set of libraries that doesn't point to outdated web pages.
2. Point the links to corresponding documentation on the latest C++ standard
. This would also require reworking the concept checking class templates and the concept archetypes, possibly breaking existing user code or other Boost libraries (e.g. BGL). 3. Host our own versions of those pages. It may take the longest, but we can avoid the drawback of option 1, write new concept checking classes instead for option 2, and have complete control over our documentation.
They're not mutually exclusive. A link to the Wayback Machine is much better than a dead end -- how about starting with that to repair the ruined links, so the information is accessible to visitors while we devise a more permanent solution? Josh
Cromwell Enage via Boost wrote:
Not sure if this has been reported before, but any links on the Reference page of the concept_check library that point to SGI.com now take you to a page that says 'SGI.com Tech Archive Resources now retired'.
I've noticed this with links from other Boost libraries (e.g. Boost.uBlas) to the SGI's former webpage. (Way to not go, Hewlett Packard.) I see a few options to work around this:
1. Point the links to the last versions archived by the Wayback Machine [...] 2. Point the links to corresponding documentation on the latest C++ standard
. [...] 3. Host our own versions of those pages. [...]
This problem affects many C++ libraries and so hosting at isocpp.org might be another option. Regards, Michel
On 4/8/18 3:07 AM, Spencer Collyer via Boost wrote:
Not sure if this has been reported before, but any links on the Reference page of the concept_check library that point to SGI.com now take you to a page that says 'SGI.com Tech Archive Resources now retired'.
Regards, Spencer Collyer
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
This is an incredible shock to me. I left a nasty comment on the website voicing my displeasure. On a more constructive note, I'll find a copy of those pages and host them on my web site at rrsd.com. It's the only way that I feel confident that these unique historical, referential, and didactic resource can be preserved. It's incredible that the C++/Generic Programming world has come to this. Note that my motivation is not just preservation for it's own sake. These are eminently practical documents for understanding STL and the form of designing Concepts and APIs. I refer to them regularly. As far as I know, they are the only description/example of "formal" documentation. If anyone has pointers to a little additional information, such as who the author of these pages is (I'm guessing it steponov and lee) I would like to hear about. Maybe I can preserve some of this also. Robert Ramey
participants (5)
-
Cromwell Enage -
Josh Juran -
Michel Morin -
Robert Ramey -
Spencer Collyer