Re: Re: PEAR_ErrorStack Quality

Bertrand Mansion wrote:

Greg Beaver wrote:

Aidan Lister wrote:

Are we going to underscore all good packages with PEAR_ then? Sorry, but I think this is highly stupid.

What is your problem? This has nothing to do with other packages, and the name has been set for months. The package is an error stack that resides in PEAR and is a core component recommended for use by PEAR packages, so it is called PEAR_ErrorStack. You want to change the name because you think "PEAR_" will make people not want to use it. That is indeed highly stupid. Offense intended when I say: if you were thinking rather than saying ridiculous comments like the above, you'd see that the name is descriptive, and the quality of the package is what matters most. If you have any similar "constructive" criticisms like this message, please take them off-list until you have thought them through thorougly. I'm always happy to address any criticisms of the code's content, or design choices.

Greg, don't be so mean with Aidan. I posted my other mails before reading yours. IMO, you should take these comments into consideration even if they look stupid for you at first glance. I

???? The next time someone tells me that what I've said with careful consideration is "highly stupid," with no response to anything I have written, I'll be sure to listen carefully. That makes a load of sense. Did you read Aidan's messages? When he first posted a carefully considered argument for changing the name, I responded with my reasoning behind the name. Rather than debate intelligently, he resorted to insults. That is not how you get your way with me. Maybe it works with dumberer people. If you want to change my mind, you need to use a solid argument based on reasoning. Getting back to the original argument, let me put this another way. if "PEAR" is a stigma, then why do you use or develop for it at all? Isn't it to provide high quality code through an easy distribution system? Are we all really so victimized and helpless to effect change to the PEAR system? Most people download from the site because it has been less than a year of a truly stable PEAR installer, in spite of the misleading stable label on 1.0 and even 1.2.1. In addition, the windows PEAR is still extremely difficult to use, and is not installed by default, so users who wish to run PEAR on the most common operating system have a load of trouble. This is not to denigrade the work done on the windows side of things, but this is the way things work. If someone were to write an installer for windows that is an .exe file and worked out of the box, PEAR would instantly gain a following double of its current size - using the installer. Any new concept takes a while to sink in, no matter how brilliant, and word of mouth ("this installer thing is really much easier to use than unzipping") will be the thing that causes this if the quality is high enough. If the quality isn't high enough, that is another story. I've been breaking my back to make the quality of code high enough in the installer, through PEAR_PackageFileManager, through PEAR_ErrorStack, and through the PEAR installer. The same thing is true in any field. One of the greatest quartets around today was virtually unknown 10 years ago, but they've been around since 1976. They had a personnel change 10 years ago, and it took 6 or 7 years before the world noticed there was a quality change. Certainly, it takes time for change, but high quality code is not going to be ignored. Greg