External links: Whether & how to distinguishing them from internal links, and to open them

How do other IAs/UXD’s treat external links & what are the perceived pros/cons of different options. Specifically looking for recommendations on:

  • whether & how to visually distinguish external links from internal links
  • whether to open them in the same window/tab or a different window or tab

I’ve found some great feedback re these questions on http://www.ixda.org/search.php?tag=external+links & http://www.useit.com/alertbox/open_new_windows.html

Looking for additional opinions, thoughts & especially any usability test findings contributors of this site might have re these questions.

It is decidable whether a pushdown automaton will accept a word?

I’m asking myself if the problem of decide whether a push down automaton will accept a word is decidable.

I would say that you can simulate a push down automaton with a Turing Machine and, if it doesn’t loop forever, accept the word when the automaton accepts, or reject when it does. In this case, the problem of decide if a push down automaton accept a word would be semi-decidable. But I’m not sure if that is correct and also, I can’t find a way to prove if it is decidable or not.

Any ideas here?

Is there a paper or model about detecting whether musical instrument playing right note or not?

I want to create a model that can detecting whether musical instrument play right note or not.

Example: provide mp3 file and musician are playing in piano. How to check the musician play right note according to mp3 file.

I had search GG but not found anything related.

Uncertainty whether $\{a^i b^j c^k \mid i+j \le k\}$ is context-free or not

I’m having trouble with this particular language: $ $ \{a^i b^j c^k \mid i+j \le k\}$ $

If it’s not context-free, I don’t know how to correctly apply the Pumping Lemma for CFLs; if it is context-free, I don’t know how to create a context-free grammar that generates this language.

Which one applies? Can you help me out?

is it possible to determine using a single depth-first search, in O(V+E) time, whether a directed graph is singly connected?

I’m working on exercise 22.3-13 of CLRS (Intro to Algorithms 3rd edition):

A directed graph $ G = (V, E)$ is singly connected if the existence of a path from $ u$ to $ v$ implies that $ G$ contains at most one simple path from $ u$ to $ v$ for all vertices $ u, v\in G$ . Give an efficient algorithm to determine whether or not a directed graph is singly connected.

Previous threads on this question are here and here. In that second thread, one of the answers links to a preprint by someone who corresponded with one of the authors of the book, and stated (in the abstract of the preprint) that the solution the authors intended was (if I’m understanding correctly) to call DFS-VISIT(G, u) (from p. 604 of CLRS) on each vertex $ u$ of the graph, resetting all vertex colors back to white (to indicate unvisited) in between calls, and return false (for “not singly connected”) whenever a forward edge or cross edge is found, and returning true otherwise. This would take O(VE) time.

There seems to be some confusion in the answers and comments section of those threads: the above scheme is referred to as “calling DFS on every vertex”. In fact, this is different from the DFS algorithm presented on p. 604 of CLRS, which calls DFS-VISIT on all unvisited nodes in the graph, without resetting the colors in between calls, which takes $ O(V + E)$ time.

My question is, instead of doing what is stated in the preprint, which takes O(VE) time, is it sufficient to simply perform the ordinary $ O(V + E)$ DFS on the graph, and return false (for “not singly connected”) whenever a forward edge or a cross edge that is located entirely within a single DF tree is found? (As correctly pointed out in one of the comments in the second thread linked above, a cross edge within a single DF tree means the graph is certainly not singly connected, but a cross edge between two vertices in different trees does not necessarily mean so).

In other words, is it possible to have a non-singly connected graph, such that for some ordering of the adjacency lists, performing the DFS algorithm from p. 604 of CLRS will not find any forward edges or cross edges within individual trees, but possibly only cross edges between trees, if any?

How can I determine whether I’m vulnerable to SPECTRE/Meltdown/L1TF on Ubuntu

I’m running Ubuntu 18.04, and I have the following linux-image and intel-microcode packages:

ii  linux-image-4.18.0-17-generic         4.18.0-17.18~18.04.1                   amd64        Signed kernel image generic ii  intel-microcode                       3.20180807a.0ubuntu0.18.04.1           amd64        Processor microcode firmware for Intel CPUs 

Assume the image is immutable, so I can’t install and run the speed47 script or install additional packages—I will make a new image if this one is vulnerable.

What canonical (but not necessarily Canonical) sites can I visit to determine whether the packages contain the necessary mitigations? I’m having trouble finding this information. For example, https://wiki.ubuntu.com/SecurityTeam/KnowledgeBase/L1TF only points to information for up to 4.15, not the 4.18 kernel I have.