svn checkin to #mysqlfs to avoid ./configure incorrectly-reported failures…

Uncategorized Comments Off on svn checkin to #mysqlfs to avoid ./configure incorrectly-reported failures…
svn checkin to mysqlfs to avoid ./configure incorrectly-reported failures and a checksum error.

Uncomplicated Version Numbers

Uncategorized 1 Comment »

I don’t understand the reason for the various Vanity Versioning. Seriously, it only makes sense to the one guy who assigns it. For example if MC Hammer released an upgrade to his album, which do you think it would be?

  1. MC Hammer: Don’t Hurt ‘Em: 1.3.7-rc2-hit-em-harder
  2. MC Hammer: Don’t Hurt ‘Em: 2

OK, that’s a strawman, but can you tell me what the additional detail of “Example-1.3.7-rc2-conan-the-barbarian” means over “Example-2”, or even over “Example-1.3.8” ?

Also, computers and scripting systems have a wretched time parsing thing stuff. They can understand how “1.3.7” and “1.3.8” rank, and “1.4” vs “1.3.9”, but strcmp() doesn’t work comparing “1.3.1”, “1.3.2-rc2”, and “1.3.2”. Invariably, it screws up, and with the habit of OpenSource to create-rather-than-reuse, inventing your own vanity versioning format causes a lot of spare-time engineers to waste their time figuring out your system.

The “release candidate” for a version is really the last version with a point release. Truly. If you have a version 1.4, then obviously something in the 1.3.X line is going to roll-up to 1.4.0. Suppose it’s 1.3.8, or 1.3.12, or 1.3.999; do any of these cause you confusion? No, obviously. Is there any doubt that one of them will be a 1.4.0? Obviously not. Everything is a Release-Candidate, so why bother marking that? It’s like saying “this is a version”, which it obviously is.

I’ve yet to hear a clear reasoning for RC markers, -preX markers, or vanity “the version has ‘snoop’ in it cuz my nickname is Dawg” markers. Seriously. Can you think of one? The comments are open…

WP Theme & Icons by N.Design Studio
Entries RSS Comments RSS Log in