Two Server Setups, Two Spreads

A web team times how long a page takes to load, in seconds, for randomly chosen visits served by two independent server setups:

The two groups are different visits, so this is an independent-samples comparison, not a paired one. The groups also differ in size and look very differently spread. So instead of assuming one shared standard deviation, the team uses Welch's t statistic, which lets each group keep its own sample variance:

$$t = \frac{\bar{x}_A - \bar{x}_B}{\sqrt{\dfrac{s_A^2}{n_A} + \dfrac{s_B^2}{n_B}}}$$

Here $\bar{x}$ is a group's sample mean, $n$ is its size, and $s^2$ is its sample variance, which divides by $n - 1$:

$$s^2 = \frac{1}{n-1}\sum_{i=1}^{n}(x_i - \bar{x})^2$$

What is the value of $t$? Keep the order A minus B, so the sign is part of the answer.

Round your answer to 2 decimal places.