On Tuesday 09 February 2010 15:51:03 Brice Leroy wrote: > 2010/2/9 Mike Ramirez <gufym...@gmail.com>: > > On Tuesday 09 February 2010 15:12:21 Brice Leroy wrote: > >> Would you eat some food not cook by a professional ? > > > > I would be very wary of a person who hasn't cooked before... The > > difference is in experience. > > it's just that I don't like those ridiculous images that shows the > world as binary. You can effectively be sure that nobody want's to > flight in an home made plane and nobody want his butcher to be his > surgeon. But in the past I painted the wall by myself, and made myself > some pretty good pasta :p >
But in both cases this was only affecting you... I don't care if you shoot heroin, as long as that habit doesn't negatively affect others around you. This is the key to my argument. > >> > Because you can, doesn't mean that you should... > >> > > >> > It's this type of mentality that leads to security holes and bad web > >> > apps that hurt server performance and many different things... > >> > >> Not everybody needs to run facebook. > >> Your approach seams to be "if it's not perfect it should not be used". > >> Don't you think it's a little to extreme ? > > > > No, I may lose a bit of perspective because I don't think in terms of the > > layman, but at the same time I'm not about to get into something unless I > > know enough about it first and if that means taking classes at community > > college... That's what I do. > > > > I'm more against something like my mom putting together a website because > > she won't do the neccessary homework first. I think that without > > knowledge of what you're doing, you're doomed to fail. If webservers were > > either all VPS's with limits monitored by the host os or dedicated > > servers for each site, then I wouldn't say much on this, but virtual > > hosting is another story, where you have the potential to hurt others. > > This is the main problem here is that we think in terms of how it > > affects ourselves only. what about others? > > Actually I had that in mind. For example, the code generated for list > is using pagination to avoid mass data reading from the db. > The purpose of this tool is to go further than createproject and in > another way like being able to fork an existing project or quickly > include existing recognized pluggable. > I can see this being helpful, but I can't see it as a general hosting service installer system... That is just a very bad idea. In the end, I see this as three things, an Educational tool -- first (thanks for that pespective Richard), secondly a build tool for experienced programmers and lastly, a controlled SaaS. For me, I saw this first as a replacement for startproject which in my mind isn't perfect, but the best we have. Because I don't see how something like this can save that much time, nor am I a believer in there is only one way to code a good app, I came across negative. I don't usually have the same info or code in each project, outside of what startproject gives and refactoring, well I went over that. (Also goproject shaped my perspective as it's what I've been messing iwth mainly over the past few weeks in an effort ot learn/undertand Go). Unless you're going to go with something like wordpress.com and have it launch sites for others and you still maintain control over most of the code and that edits by users are limited... Really where they only need to deal with the frontend that's great, but even then... Experience matters (XSS and bad js can clog up browsers). I'm really wary about unknowledgeable users editting things that can mess up the site and the service. Making it easier for the laymen is like tossing free money into a crowd, sure it's a nice thing, but it has the potential for catastrophe. If that potential is removed or severly limited, I can jump on whole heartedly as a supporter. I think your plan is better as a SaaS, than to target the laymen to use django. The above is also something I think is fundamentally different from say dreamhost or webfaction adding this in to their software installers, which is where the potential for catastrophe starts to grow exponentially as users make edits that can blow things up. > >> > A bad django app made this way can hurt django's reputation. > >> > >> That's pretty severe. > > > > Maybe, but public opinion is easily swayed and django is still fighting > > against other more popular systems. > > Just to add to this, PHP is the example, it's rep is really bad, mainly because of bad programmers building bad applications. Now wp, joomla, drupal (before phpnuke and there was one more, forgot it), help this, but they all have issues that are driving users away. Mike -- PUNK ROCK!! DISCO DUCK!! BIRTH CONTROL!!
signature.asc
Description: This is a digitally signed message part.
