The wiki went down a few hours ago. Ive redirected requests to the fallback server; the redirection exposes its IP address.
The wiki was frozen two years ago and will eventually be removed. If you still rely on it please take a moment to write to me about whats needed for you to switch to the up-to-date documentation at code.kx.com.
Thanks
Stephen
Stephen Taylor | Librarian | Kx | +44 7713 400852 | stephen@kx.com
Personally I look forward to the old wiki permanently being removed. The inconsistency in page design and the redirect banner at the top of the wiki page after clicking kx links from Google results is a very mild but constantly occurring pet peeve.
It would be nice to have an ability to go directly to the keyword reference by some URL/keyword for the sake of qPad’s “F1 @ keyword”, but new documentation’s search can be used instead (differs by one additional click on a first result), so no that big a deal, I suppose.
That url is useless. It does not provide direct link to ssr. Read below.
Why we may argue about whether it’s important, integration with google search is different then searching on site.
I’ll try to explain my workflow.
working in jupyter.
ctrl+t open new tab ; type kdb ssr ; scan results from top ; click first result ; read doc ; ctrl+w close tab = 2 command shortcuts + 1 list scan + 1 click
keeping typing in jupyter with refreshed knowledge
This is about 2 seconds to get to ssr definition. My flow is not interrupted.
with what you suggest i will need to:
open code.kx.com (or scan through tabs and find it open) = 1 scan tabs + 1 click OR 1 shortcut + type window title search string (code.kx)
focus search fild = 1 click
scan results (no result for SSR) = 1 scan
click on any link because they all are wrong = 1 click
scan for ssr (won’t do it), instead will type ctrl+f ssr = 1 shortcut + 1 scan for yellow highlight
While I adore kdb, it’s dynamic language, so no any help from IDE, so I use google a lot.
With removal of wiki, i’m out of usable manual.
to make things worse, try searching for “ss” on the site. Compare it to google “kdb ss”.
WHAT TO DO:
process your documentation. It seems to be structured enough. Generate partial pages with proper urls without id/anchor, put it in “/funref” directory and let google index it.
Add enough keywords, add “kdb” (because no kdb in url, only kx).
Reformat funref to contain no extra space for novices.
Then compare search hit rate on both funref and current pages, and you’ll get what I mean.
Now I see I replied to everyone, Ive removed the password challenge. V2 is not yet ready to publish, but your comments are welcome.
A word about context. Alex is right to mention novices. My primary mission is to lower the learning gradient for q. For those of you who found your way through the wiki, any improvements to code.kx.com are likely to be marginal. The least I can do is stay out of your way, and Im grateful for information about what does and doesnt work for you.
There is some prospect of condensing the Reference articles by writing more tutorials and moving examples and discussion to them.
And something I just learned about the Jupyter notebooks:
Just put the cursor on ssr within the Jupyter notebook and hit Shift Tab a couple of times the built-in context-sensitive help takes you straight to the correct page on code.kx.com without having to lift fingers.
Stephen Taylor | Librarian | Kx | +44 7713 400852 | stephen@kx.com