Install Kolab 3.4 on Debian Jessie

John Candlish

The current version of Kolab 3.4 for Debian 8 does not configure after installation.

Mostly because their package cyrus-imapd is compiled against perl-5.18 and Jessie ships with 5.20.

So the first thing that you have to do is build an -nmu package from their sources.  Kolab sources require a single change to the file debian/cyrus-imapd.install.

Another little gotcha is that setup-kolab ldap fails to configure because:

This patch gets it going:

To check that the cyrus-imapd configuration is correct:

Once self-signed certificates are working with cyradm it is safe to configure roundcube. The following stanzas must be present in /etc/roundcubemail/config.inc.php :

The following text in /var/log/roundcube/errors is an indication that the cafiles are not readable:

To make the cafile is readable it should be set chgrp ssl-cert and the user www-data should be a member of the ssl-cert group.

Then it should be possible to log-in to the roundcube web client.

REF:
https://docs.kolab.org/upgrade-guide/kolab-3.4.html
https://docs.kolab.org/installation-guide/debian-community.html
https://docs.kolab.org/administrator-guide/using-the-kolab-command-line.htm
lhttps://docs.kolab.org/administrator-guide/setup-kolab-cli-reference.html
https://git.kolab.org/T492

WordPress does LaTeX

John Candlish

My February '95 Statistics 462 midterm notes.  In LaTeX.

$ {\bf Independence} $

$ P(A \cap B) = P(A) \cdot P(B) $

$ f_{X_1,X_2,\ldots,X_n}(x_1,x_2,\ldots,x_n)=f_{X_1}(x_1)f_{X_2}(x_2)\cdots f_{X_n}(x_n) $

$ P((X,Y)\in R)={\displaystyle {\int\!\!\!\int}_{R}} f_X(x)f_Y(y)dy\;dx $

$ {\bf Moment \: Generating \: Function} $

$ M_X(t)=E(e^{tX}) $

$ E(X)=M_{X}^{'}(0); \quad E(X^2)=M_{X}^{''}(0) $

$ Var(X)=M_{X}^{''}(0)-[M_{X}^{'}(0)]^2 $

$ {\displaystyle \bar{Y}=\frac{1}{n}\sum_{i=1}^{n}Y_i\quad Var(\bar{Y})=\frac{Y_1}{n}} $

$ {\bf Chebyshev's \: Inequality }$

$ {\displaystyle P(|X-\mu| \geq \epsilon ) \leq \frac{\sigma^2}{\epsilon^2} = \frac{Var(X)}{\epsilon^2} }$

$ {\bf Special \: Distributions } $
$ {\rm Binomial \: Distribution} $

$ X \sim b(n,p) \Rightarrow P(X=K)={\displaystyle {n \choose k }}p^k(1-p)^{n-k} $

$ E(X)=np \quad Var(X)=npq $

$ {\rm Poisson \:Limit}$

$ {\displaystyle \lim_{n \rightarrow \infty } { n \choose x }p^x(1-p)^{n-x} = \frac{e^{-\lambda}\lambda^x}{x!}} $

$ {\rm Poisson \: Distribution} $

$ X \sim P(\lambda) \Rightarrow P(X=k)=\frac{e^{-\lambda}\lambda^k}{k!} $

$ E(X)=\lambda \quad Var(X)=\lambda $

$ \mbox{MLE for }\lambda = \bar{X} $

$ {\rm Poisson \: Approximation} $

$ \mbox{If }X \sim b(n,p) \mbox{ then } {\displaystyle P(X\geq K) \doteq \sum_{i=k}^{\infty} \frac{e^{-np}(np)^k}{k!} }$

$ {\rm Normal \: Distribution} $

$ X \sim N( \mu, \sigma^2 ) \Rightarrow P( a < X < b ) = {\displaystyle \frac{1}{\sqrt{2\pi}\sigma} \int_{a}^{b}e^{\frac{-1}{2}[(x-\mu)/\sigma]^2} dx} $

$ X \sim N(\mu,\sigma^2)\quad \Rightarrow \quad{\displaystyle \frac{X-\mu}{\sigma}}\sim N(0,1) $

$ \hat{\mu}={\displaystyle \frac{1}{n}\sum_{i=1}^{n}}y_i=\bar{y} \quad\quad\quad \hat{\sigma}^2={\displaystyle \frac{1}{n}\sum_{i=1}^{n}}(y_i-\bar{y})^2 $

$ Y_1+Y_2\sim N(\mu_1+\mu_2,\sigma_1^2+\sigma_2^2) $

$ \mbox{If }{\displaystyle \bar{Y}=\frac{1}{n}\sum_{i=1}^{n}Y_i} \quad\mbox{then } {\displaystyle \bar{Y} \sim N(\mu,\frac{\sigma^2}{n}}) $

$ {\rm DeMoivre-Laplace} $

$ \mbox{If }X \sim b(n,p) \mbox{ then } \lim_{n \rightarrow \infty} P \left( c < {\displaystyle \frac{X-np}{\sqrt{npq}}}< d \right) = P \left( c<Z<d \right) $

$ {\bf Consistancy} $

$ P(|W_n-\theta| < \epsilon ) > 1-\delta \quad n>n(\epsilon,\delta) $

$ \mbox{let }\delta = {\displaystyle \frac{Var(W)}{\epsilon^2}} \mbox{ and use Chebyshev to find } n $

$ {\bf pdfs \: of \: Common \: Estimators} $

$ \mbox{uniform distribution} $

$ Y_{max} \:\:\:\:\:\:f_W(w;\theta)={\displaystyle \frac{nw^{n-1}}{\theta^n}} $

$ Y_{min} \:\:\:\:\:\:f_W(w;\theta)={\displaystyle \frac{n}{\theta}\left(1-\frac{y}{\theta}\right)^{n-1}} $

$ {\bf MLE } $

$ \mbox{likelyhood function} $

$ L(\theta)={\displaystyle \prod_{i=1}^{n}}f_Y(y_i;\theta) \quad \Rightarrow \quad \ln L(\theta)={\displaystyle \sum_{i=1}^{n}}\ln f_Y(y_i;\theta) $

$ \mbox{solve } \quad 0 =\frac{\partial}{\partial\theta} {\displaystyle \sum_{i=1}^{n}}\ln f_Y(y_i;\theta) $

$ {\bf CI} \: 100(1-\alpha)\% \mbox{ for the binomial parameter } p $

$ {\displaystyle \left( \frac{y}{n}-z_{\alpha/2}\sqrt{\frac{\frac{y}{n} \left(1-\frac{y}{n}\right)}{n}},\frac{y}{n}+z_{\alpha/2}\sqrt{\frac{\frac{y}{n} \left(1-\frac{y}{n}\right)}{n}} \right) }$

$ \quad\mbox{where}\quad {\displaystyle y=\sum_{i=1}^{n}X_i } $

So yes, WordPress does LaTeX, but it is finicky and missing a lot of functionality. If you try to use LaTeX that doesn’t work inside the math environment (such as begin{align} ... end{align}), it wont parse.  Also, you need to use the displaystlye keyword judiciously.

Edit: I've uninstalled JetPack and so lost its Beautiful Math feature.   Therefore I've installed the WP LaTeX plugin. It is more configurable, and can offload formatting to a real LaTeX running alongside WordPress.  With a little hacking I've got {align} going. Why someone thought is was sensible to limit WordPress' LaTeX support to the math environment is beyond me.

One requirement of the WP LaTeX plugin is that the WP unformatted plugin is also installed, so that the wptexturize() function can be disabled on LaTeX posts.

Established 1999

John Candlish

Why does the website say “since 1999” you might ask?  Because that is when I registered the candlish.net domain and established an online presence.

How ’bout that wordwrap?

Fairly narrow layout. :/