How to install MariaDB rpms in CentOS 6.4 using rpm (not yum cmd) + handling mysql-libs conflicts

Question

I am trying to perform a wrap around of the ASCII alphabet characters in order to perform a shift from a key. For example, if the key is 2 then shift the letter A to become C. But, how do I wrap around Z in order to get to B using modulus arithmetic?

I am trying to implement the following formula:

ci = (pi + key) % 26;

Where ci is the ith ciphered letter and pi is the ith letter to be ciphered.

Solution 1

Instead of " rpm -ivh ..." you should have done "rpm -Uvh ...".

Except for perhaps kernel, -Uvh should ALWAYS be used.

The Obsoletes: tags are not performed when using --install,-i.

Solution 2

You don't. You use yum so that it handles this situation cleanly. It operates perfectly well offline using RPM packages directly, provided that all the necessary dependencies are available or already installed.

1. Download the appropriate MariaDB RPMs to your USB stick or DVD-ROM.

   yumdownloader MariaDB-client MariaDB-common MariaDB-compat MariaDB-devel MariaDB-server MariaDB-shared
   

2. Take your USB stick or DVD-ROM to the target machine and copy them in.

3. Install them:

   yum --disablerepo=* -y install MariaDB-*.rpm
   

Comments

[Michael Hampton]

I don't understand how this helps with doing an offline installation?

[GioMac]

added to the end :) Guy knows how to download and put files to the server.

[Danila Ladner]

localinstall goes into the internets to check for dependencies.

[GioMac]

hahah, yum localinstall -y MariaDB-*.rpm

[Michael Hampton]

@GioMac On EL6 they can be used interchangeably. There was only a difference on EL5.

[Danila Ladner]

unless you configure local repo on the box and enable only that repo.

[GioMac]

Yes, I see, actually never used it (I'm fine with rpm), still not satisfied with how yum works, RHEL5 yum is a headache. RHEL7 will be fine.

[Michael Hampton]

@GioMac RHEL7 will also include MariaDB by default :)

[GioMac]

That's what happens with the product when maintainer fails to communicate with OSS community :) developers.slashdot.org/story/13/09/09/2259206/…

[chux - Reinstate Monica]

Note: Unexpected results with test_shift < 0. Also off-by-1 with letter <= ASCII_CAP_LETTER_OFFS + NUM_OF_LETTERS which should be letter < ASCII_CAP_LETTER_OFFS + NUM_OF_LETTERS. <= --> < in 2 places.

[Toto]

Thanks for that hint @chux. I've made some corrections. A negative shift is now possible ( by adding NUM_OF_LETTERS which makes it a shift in positive direction)

[chux - Reinstate Monica]

Good improvement. Also for your design review, why short in short shift vs int? Usually short makes sense for arrays to reduced size. int is the usual workhorse type.