I held off on replying, since I was hoping to allow others a decent chance to voice an opinion before it being tainted by me. Sorry guys, I now have to speak.
On Fri, Apr 22, 2011 at 10:39 AM, Alessandro Molina < [email protected]> wrote: > I was taking look at the various 2.1.1 tickets and that raised me a > question when I saw http://sourceforge.net/p/turbogears2/tickets/56 > For any who do not wish to look, the ticket is about adding Elixir as an option to the quickstart. > Currently SQLA Declarative has quite made Elixir > superfluous, and I have to say that supporting multiple ways to manage > models might actually confuse the users, so I was asking myself if this > might do more bad than good to the project. > Also having both makes harder to have a bug free environment. > Personally, I want to see Elixir properly supported. There's no reason not to do it in my mind. Some people like Elixir. Personally, I like SQLAlchemy, but not everybody does. Now, as to why? SA, the templating engines, and Elixir are ways in which we can differentiate ourselves from the competition. If we have it properly supported and documented, then we can say we make things possible that the competition simply does not deal with at all. I'm not going to claim it's easy. I'm not about to claim it's the best thing ever. I am going to stick by the desire to have it properly supported, though. We have few reasons not to do so, and we can get a benefit from doing so that no one else provides. > We already have 5 supported template engines and this is causing > problems with external libraries like ToscaWidgets. > Also chameleon.genshi is partilly not working due to requiring > chameleon.core <= 1.0.2 which has currently been blocked on pypi. > This is simply one of the areas where we have to limit ourselves. We can't force ToscaWidgets (and all the other widgets that have been written) to support our templating engines. For someone to write so many different widget templates is just not going to happen. To my mind, what we should do is provide Genshi and Mako templates as our defaults, and then allow people to choose others. If they wish to use Jinja2, and the widgets in question don't work with Jinja2, then they must either write Jinja2 compatible templates or do without the widget. Honestly, I'm okay with either option for them. As for chameleon.genshi and blocking on pypi, that won't be a problem. One of the changes coming for 2.0.4 and 2.1.1 will be pointing people at our file index by default. It's possible for someone to get a later version of chameleon.genshi, but it's going to be somewhat harder. Again, not easy, and does leave us open to certain problems that should probably start getting a FAQ done, but I don't want to close off choices for people without a strong technical reason. A significant chunk of our user base comes from toolkits which limit them in some fashion, and they're looking to remove those limits. If we don't accommodate that desire, they're just going to move on again, and we will have missed a great chance to do something with/for them. > Just my two cents, I was mainly reflecting if we aren't causing more > damage than good with all those choices :D > The ages old conundrum of "how much choice is too much?". I'll point to Ubuntu for an answer: They install certain things by default, and make it possible to use anything else that you might choose. For instance, they don't provide emacs by default, but using it is a single command away. There's no reason we can't do similar. -- Michael J. Pedersen My IM IDs: Jabber/[email protected], ICQ/103345809, AIM/pedermj022171 Yahoo/pedermj2002, MSN/[email protected] My LinkedIn Profile: http://www.linkedin.com/in/michaeljpedersen -- You received this message because you are subscribed to the Google Groups "TurboGears Trunk" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/turbogears-trunk?hl=en.
