I am opening a post here for more people to see who will be affected by the changes …
I had a conversation with @jon over the phone the other week and we talked about some technical problems with the wiki and how we can make it better. I had some time to implement these already. So here they are:
This will help us to find pages that are no longer linked, but still might show up in some search index. It also helps us to have a more logical structure of the wiki - which is mostly helpful for the authors.
Removal of the sidebar
I came to the conclusion that this does not really have a purpose. It wastes a lot of screen space and I do not see what it adds that wasn’t there before.
Pages are now wider (especially on mobile devices) and there is more space for the content
Comments when pages are restored
When someone restores an older version of a wiki page, there is now a comment field to make clear why something was reverted.
I guess you all know why this is necessary
Remove older versions from the Google Index
Deleted or older versions of some pages show up on Google. This is annoying, because improvements to that page won’t be seen by users. They will simply view an outdated page.
It might take a couple of days until they are all removed. If you see anything in a couple of days, please let me know.
I suggest retaining this article. It might be shortened.
Some of the passive cards could still be in use and causing otherwise unexplained slow transfer speeds, as people move to higher bandwidth Internet plans.
I’m uncertain how to ascertain whether a chip not listed on the page is active or passive. A relatively high level test that I have used is to dd a file of ~100 MB from a server over
NFS (the latter avoids encryption overhead) to a workstation having the chip under test.
That was repeated at least once, to get the file cached at server end, with “systemctl vm.drop_caches=3” in between, at client end.
It reported 116 MB/s, to two different workstations, via a DLink Gb switch, which is a practical topology for most situations. From which I concluded that:
the tg3 chip in the dual-core 1.3 GHz server is probably active
the RTL8168 chips in the quad-core 2.7 GHz & dual-core 1.8 GHz workstations are either active or the difference is not significant with that amount of CPU power.
thanks for cleaning up the wiki. This page should be preserved, as it is important for people who are willingly to provide another mirror for the project.
However, it is not directly related to development but I guess it does not make sense to open up a “infrastructure” category as we do not have anything else fitting into it (perhaps the postmaster information?).
Is it worthwhile to create a “package” area in (e.g., https://wiki.ipfire.org/package with https://wiki.ipfire.org/package/gnutls)? Or just delete the gnutls wiki page?
I think for the time being we should concentrate on the add-ons that actually require configuration.
I do not think that it is wrong to have some documentation about standard packages and commands of the OS, but probably very fewer people will benefit from this.