On Fri, 2006-03-31 at 14:21 +0200, Daniel Bimschas wrote: > Thank you Malcolm, > > again it sounds like work is mainly being done on the magical-removal > branch.
This is true, but it is not a bad thing for people using the trunk. Bugs / problems / interesting architectural decisions on the trunk are known and people can work around them if they wish. It's not getting worse and it is not changing drastically. Not a bad target to work against if you are starting out. You should appreciate that you hit a real edge case in the example you are asking about, for example. It seems that hardly anybody has a requirement to use character fields as primary keys. I hit the same bug you did almost immediately after starting to use them. But don't take that to mean trunk is in some way the unloved son of Django. The main roads are well tested and well used. You will find problems in corner cases on magic-removal as well. > I wonder if it would be the best to switch to MR now? Is it > "stable"? Or when do you expect it to become stable? Let's differentiate between "stable" (in the sense of "does not crash; is usable") and "unchanging". What follows is entirely my personal opinion and not in any way sanctioned by the hive mind... It is definitely possible and practical to write fully functional applications using the magic-removal branch without much trouble. It does not crash randomly. All the features from trunk are there (if not, you should raise it; it may have been an oversight). There *is* a certain amount of relearning required because of design changes (the whole point of magic-removal). But having a copy of http://code.djangoproject.com/wiki/RemovingTheMagic close at hand makes that fairly straightforward. The documentation is reasonably up to date and various people are working on porting the remainder across (it mostly involves changing a few terms here and there). The tutorials have been mostly ported across, for example. The magic-removal branch is not unchanging, however. Things are being added. Some decisions are still being thought out and implemented (see, for example, Adrian's validator changes that he and now Nebojša are implementing). So you need to stay alert and keep up with the changes (watching http://code.djangoproject.com/timeline is one way). This might be more effort than it's worth for you at the moment. That being said, most of the work on magic-removal is tweaking now. The big changes have mostly landed, so it's not as likely that you will wake up one morning and somebody has changed the framework underneath you. Personally, I am only using magic-removal at the moment, except when answering questions about trunk. But I am probably not typical, in that I feel very comfortable with Python in general and quite a few databases and I think I understand the Django internals fairly well, so debugging problems is not something I mind. Realise that you hit a real edge case in the example you are querying about. It seems that hardly anybody has a requirement to use character fields as primary keys. I hit the same bug you did almost immediately after starting to use them. But don't take that to mean trunk is in some way the unloved son of Django. The main roads are well tested and well used. You will find problems in corner cases on magic-removal as well. > I started my project using 0.91 last week as a totally django-noob and > I am close to completing this first project now. Finish it first. Then do a one-time port to magic-removal (using the above web page as a guide). Let us know what we have forgotten to document about the port. --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Django developers" group. To post to this group, send email to django-developers@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/django-developers -~----------~----~----~----~------~----~------~--~---
