![]() ![]() Verify that MySQLdb is correctly installed:īy running exit() get out of the python shell, and verify that django-admin.py is in your path: Open Terminal and run the following commands in the python shell (start this with the python command). You should be all set now, but let’s verify this right away. Sudo ln -s /usr/local/mysql/lib/ /usr/local/mysql/lib/mysqlĪll these adjustments are required because we are building and installing the driver on Mac and not on Linux. The last step required is to create a symbolic link with the following command: If instead, you don’t receive an error but you see warnings about files not required on this architecture, don’t be concerned about them. If you still get an error (and only in that case) you’ll need to edit the site.cfg file within the same folder and set threadsafe = False, before running the two commands above once again. Now, from the MySQL-python-1.2.2 folder run: usr/include/sys/types.h:92: error: two or more data typesĮrror: Setup script exited with error: command 'gcc' failedĪt this point, edit the _mysql.c file and comment out lines 37, 38 and 39 as follows: usr/include/sys/types.h:92: error: duplicate 'unsigned' Please follow these exact instructions because the source code won’t compile out of the box and will give you the following error when trying to build it: Get MySQL-python-1.2.2.tar.gz from SourceForge. Choose the one for x86 and Mac OS X 10.4. You can install MySQL by grabbing and running one of the packages that are available on the official site. I would normally recommend PostgreSQL, at least until we have DB2 on Mac, but I realize that many of you use and prefer MySQL, which also seems to be the only one that requires special instructions due to a few installation issues when trying to get MySQL and Python to work together. Alternatively, you could just create a symlink to the utility django-admin.py in /usr/bin, but I recommend the former approach. ![]() ![]() In case you are using the bash shell (as I do) then you should have a. How you go about doing this, depends on the shell you are using, and I’m assuming you are able to export a shell variable on your own. Users/Antonio/Code/django_trunk/django/bin ![]() We need therefore to add the following to the PATH (again, change it to your own location): The bin directory within the django folder (which is inside django_trunk itself) contains several management utilities. Of course, change it to the full path location of the directory on your filesystem. Within this file, place only one line containing the path to your django_trunk folder. Library/Python/2.5/site-packages/django.pth What you need to do is inform Python about the location of your django_trunk directory. You can verify this by running python in the Terminal (use exit() to get out of the python shell). Mac OS X 10.5 already ships with Python 2.5.1, thus you won’t have to install it. The development code is also rather stable and used by most people in production mode, even for sites like the Washington Post. The reason for this is that the trunk version implements a few new features. MacPorts is not required for this how-to.įirst and foremost, we are going to install Django from its svn repository, as opposed to obtaining the 0.96 release archive. I’m writing about this for the benefit of those of you who may attempt and struggle with this feat. (MySQL upgrades on the Mac are also horrendous.)īottom line, there are far faster, easier, cheaper options.Installing Django on Mac OS X Leopard is supposed to be very straightforward, but if you are new to it, you may encounter a few puzzling questions and, in the case of MySQL, even a couple of headaches. Whatever the case, IMHO, breaking your back for hours trying to get fscking MySQL to install on your Mac just isn't worth the time. If you don't want to install VirtualBox (or aren't comfortable with it), there are tons of options like Vultr where for a staggering $6/mo, you can get a pre-baked OS with MySQL/Maria to play with. So, standard fare development is unlikely to get it to break a sweat. Point being, it's working with a dataset over 10mm records and returning results in a few seconds. When I run the same query on an indexed column, you get: 1 row in set (4.703 sec) Just for fun, I just spun up my own copy of the OpenBSD/Maria instance just now on my Mac and ran this on a smaller dataset I'm doing some ad hoc queries on. Unless you want a reason to spend the money on a new lappy, if your Mac was made in the past decade, you'll rarely (never?) find it wanting and likely don't need to replace it. Our guys are doing serious development with some tables containing a lot lot LOT of data. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |