When installing Gitlab on FreeBSD, charlock_holmes doesn’t compile when executing the following step:
sudo -u git -H bundle install --deployment --without development test mysql aws
Even if you have devel/icu installed, charlock_holmes doesn’t look for header files in /usr/local/include where devel/icu installs them.
The fix for this is to perform an intermediary step:
sudo -u git -H bundle config build.charlock_holmes --with-opt-include=/usr/local/include/ --with-opt-lib=/usr/local/lib/
then do the ‘bundle install’ step.
(Hat tips to https://github.com/brianmario/charlock_holmes/issues/9#issuecomment-10370071 and http://blog.herrbeesch.fuettertdasnetz.de/?p=24 for this solution.)
Julia is the new open source, high-level scientific, technical and numerical/data analysis computing language. With design goals of high-performance, numerical accuracy, syntax familiarity and distributed parallel execution, Juila is already being touted as both complimentary and competition to data analysis tools such as MatLab, R, and Python/NumPy/SciPy.
So come join this high-level introduction to Julia. We’ll be looking at it’s goals, features and interesting language constructs. We’ll also be looking at how well it plays with other data analysis tools. And of course, there will be examples.
The generally accepted method of determining the size of PostgreSQL objects (databases, tables, indexes) is to use one of the pg_[database relation]_size functions. These work fine for scripts and whatnot. But what if you just want to get a quick look, and not have to go through all the typing hassle?
A much quicker method (if you’re in the psql client) is to simply use
"+" after a
For instance, while doing some db housecleaning tonight, I wanted to see how large my indexes were. Being already in psql, I could have typed:
"select pg_size_pretty(pg_relation_size('schema.index-name'));". Not fun. Not only is it a lot of typing, but there are two problems. First, one needs to know–or lookup–the object name of interest. Second, each object has to be looked up individually, so immediate side-by-side comparison isn’t possible.
Instead, I used the much simpler
"\di+ schema1." and received a nice table of index metadata–including size–resembling the following:
database> \di+ schema1. Schema | Name | Type | Owner | Table | Size | Description --------+----------------+-------+---------+---------+---------+------------- schema1 | daily_pkey | index | dbowner | daily | 1070 MB | schema1 | date_idx | index | dbowner | daily | 469 MB | schema1 | name1_idx | index | dbowner | name1 | 537 MB | schema1 | name2_index_1 | index | dbowner | name1 | 768 kB | schema1 | name3_pkey | index | dbowner | name1 | 912 kB |
No object name lookup, less typing, and all in one spot. And if shell or script use is desired, it’s a simple
"psql -A -F , -t -c "\di+ schema1. " .
This is my Minnebar 2011 presentation, Brain-Computer Interfaces: Hacking Wetware.
Thanks to all who attended, and I hope you enjoyed the presentation.
Other links, for those who might be interested in hacking:
Short and sweet, this is my Minnebar 2010 presentation entitled “I’m a guru, not a god!: A Tao of System Architecture” for those of you who requested a copy. Sorry for the delay. Note that I removed the “Insertion Model” diagram, as I’ve not received permission to distribute it.