Hey Guys, I stumbled across this link http://jesusphreak.infogami.com/blog/vrp1 which is a blog post titled "Python web development and frameworks in 2007".
In it is a section about TurboGears(quoted below for people who don't want to read the whole article) discussing the future and disadvantages of it. Perhaps some developer can address those issues, i found some of them to be valid. For me, there were a few issues that were not discussed in this post, and other issues that were completely invalid such as the size of the community. I personally see that most questions go answered on this mailing list and take part whenever i can.(i don't know about the irc channel, i don't use that) Plus he probably means 1.1 and not 2.0 The issues that are relevant to me are the following: -Development seems to be slow. I know this is a community and volunteer effort. I am not saying this to insult anybody. But there seems to be just a couple of developers, and the outstanding tickets seem to be too involved for anybody to help with.(i personally would like to help if someone is willing to mentor me) -The use of existing components is excellent to me. However, i ran into the issue that TG often uses outdated version of the components and so when i need to look up something in the component documentation, i often have to consult the archives or the archived source code for help. Perhaps snapshotting the documentation is a good idea. -I agree to what he said about the TurboGears book. But this is not specific to TG, this will be true for any book about any framework. Which is why an electronic version is probably much better for constant updating. I would like to know two things, perhaps a developer can update on the status of remaining issues to 1.1 and the whole documentation effort. And the other thing is some responses and opinions from the TG community. Regards, Alaa Salman <quote> The main problems with TurboGears is it has been realized that there are issues with some of the functionality in 1.0 which is why 2.0 is being created and several key components are being swapped out or upgraded; SQLObject to SQLAlchemy, Kid to Genshi, CherryPy 2 to CherryPy 3, and a few other changes. Essentially, 1.0 is a dead framework or the 'old and busted' components that nobody really wants to use, but it is the only framework available right now. Most of the documentation is based upon 1.0, quite a few of the admin tools such as Catwalk simply won't work with 2.0 without a rewrite, etc. This has caused some real problems in the TG community. People don't really know whether to stick around using and working on improving 1.0and the documentation, or if they should wait for 2.0. Added to these two options is the third, less complete one to manually swap out components from that will be used 2.0 with those in 1.0, and this is what many have done. This level of confusion has caused the community to slowly decline. TurboGears was and is in a really difficult spot. It was almost ahead of its time, and since it was based around using the 'best of breed' libraries in the Python community it had to make these major changes. Only shortly after TurboGears really started to get popular, SQLAlchemy became a real alternative to SQLObject, and there were many cries within the community to use it instead. In addition, WSGI which had been somewhat of a secondary concept in Python soon became extremely important with the emergence of Pylons. The TG developers were faced with two options: either rework the framework from the ground up to include these newer components or to stick with what already existed and soon become outdated and an afterthought. The developers did the correct thing as difficult as it was and decided to rework the framework to integrate these newer libraries. Unfortunately it has led to the aforementioned confusion. Heck, even the TurboGears book released in November 2006 is largely outdated. The focus of 2.0 more than anything else is modularity. CherryPy 3 is much more WSGI-based than its predecessor and just about every component available is being broken down into decoupled packages (such as the Toscawidgets library). TG will also be using Paste Deploy and other Paste-related tools further in its stack. Infact, TurboGears 2.0 has many of the same goals and overall design as Pylons, and therefore there has been some talk of the two projects merging. The notable difference between the two frameworks, however, is TurboGear's reliance on CherryPy as its core, and this is not something easily reconciled with Pylon's design. Ian Bicking has offered a good comparison of the similarities and differences between the two frameworks at this link. This might not be an accurate view of the project, but I think TurboGears is in a lot of trouble. As I've mentioned, the community has declined a lot over the past year. Users have shifted over to both Pylons and Django. The maling lists tend to be pretty quiet, and the last time I viewed one of them it had quite a bit of spam. Go into the IRC room and ask a question and you'll be lucky to get an answer within the hour if you get one at all. In the end, even if 2.0 ends up being an amazing project (and I can't imagine it being implemented and well-documented in any time shorter than a year), first impressions are key, and many people still have an image of TurboGears as the frankenstein framework which couldn't quite figure out what it wanted to be. How will TG possibly remarket itself? This is all very unfortunate. TurboGears has gone from being perhaps the most discussed framework at the end of 2005 to one where the community is much smaller than Django's and which even Pylons has eclipsed. It is this combination of bad circumstances which leads me to believe that a merger between Pylons and TG would be the ideal situation if at all possible. If the main thing that separates the two projects is the reliance on Paste by Pylons and the reliance on CherryPy by TG, whatever steps possible should be taken to see if the frameworks can work together. However, it should be noted and has been stated by others that becuase the focus in both Pylons and TurboGears is modularity, the existence of one is not negative to the existence of the other. An admin application written for Pylons should theoretically be able to plug right into TurboGears 2.0 and vice versa. Despite this, it is with some sadness that I must say I have a very difficult time recommending TurboGears to a new web developer, though this could change in the future. There simply isn't enough stability right now. Nobody should have to spend a few months learning the set of tools available in 1.0 and then have to drop them and use the new ones for 2.0. </quote> --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "TurboGears" 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?hl=en -~----------~----~----~----~------~----~------~--~---
