@DBJ: I'm aware of your version, but I thought it would make more sense to do it at build time instead of run time. This way, developers who prefer a safer version can build their own safe copy from the original source, and developers who prefer not to have a safe version can just remove them at build time (in my rather informal testing I only noticed a ±5ms difference between the safe and unsafe version--- though admittedly I only tested Safari and Firefox.)
I'm not a big fan of JScript conditional comments as they seem to be an IE only feature and still add extra overhead during run time. By the way, last time I looked at your safe jQuery version I noticed your hasOwnProperty checks in the Sizzle part of jQuery are incorrect. As for building specialized versions of jQuery; I would build a standards compliant version (i.e. a version that pretends there are no bugs in browsers) and a conditionally included section of code that does feature detection and works around bugs for the various browsers. The only question is how useful the "standards compliant" version would be, perhaps only as a sort of ACID test at first, aiming to increase compatibility and standards among browsers. Hopefully at some point in the near future we would then end up with all the alternative browsers (Safari, Firefox, Chrome, etc.) being able to run the standards version and IE using the "bug support" version. Testing of these versions could be done using John's Test Swarm (nice work btw!) Bram On Aug 27, 2:27 pm, DBJDBJ <dbj...@gmail.com> wrote: > @Bram, controversial but honourable attempt > ( same intent as :http://dbj.org/jquery.1.3.2.safe.slow.js ;-) > > As I said previously: using JScript conditional compilation one can do > a lot (on this subject) straight away. No building tools needed and no > new files. > > --DBJ > > On Aug 27, 9:39 am, Bram Stein <b.l.st...@gmail.com> wrote: > > > I've been thinking about building special versions of jQuery for a > > while as well. I'm not sure if building a browser or feature specific > > version is feasible, but it might be an interesting experiment. > > Recently I have been working on a few projects that target a specific > > browser, and it would be interesting to see if there is any advantage > > to using a version specifically built for that browser (or a group of > > browsers.) My main interest at the moment is however in creating a > > version that is safe to use with code that extends the native Object > > prototype. To that end I've forked the Sizzle code and added > > preprocessor statements for conditional "compilation". My next step > > would be to do the same for jQuery. Perhaps such an approach would > > also work here (i.e. building special version from the original > > source.) > > > My Sizzle fork can be found > > here:http://github.com/bramstein/sizzle/tree/masterhttp://groups.google.co...... > > > Bram > > > On Aug 25, 1:00 am, DBJDBJ <dbj...@gmail.com> wrote: > > > > I am trying deliberately to stay "abstract" in order not to appear to > > > be giving advices to the team how to do their job. I think they are > > > qualified enough. > > > > I would rather stay in the realm of "why". Ie "why" something would be > > > a "good thing" for jQuery. > > > And this idea of not having one jQuery file where everything is mixed- > > > in for all browsers, I think might be a good thing indeed, for jQuery. > > > > @Andrea : Sizzle is the good candidate. (its makeArray() mehod, for > > > example, is particularly good example) > > > > -- DBJ > > > > On Aug 24, 5:04 pm, Andrea Giammarchi <andrea.giammar...@gmail.com> > > > wrote: > > > > > Have you seen sizzle? The main core component that could make the > > > > difference > > > > since jQuery is mainly based on selectors and arrays manipulation? Try > > > > to > > > > imagine sizzle was called jQuery.core.selector, I think makes sense to > > > > start > > > > from the main dependency that could bring benefits for everyone, even > > > > non > > > > jQuery users. But this is just my opinion. > > > > > On Aug 24, 2009 4:33 PM, "DBJDBJ" <dbj...@gmail.com> wrote: > > > > > I thought my original aim was simple ? > > > > I just thought it might be proven feasible to have one jQuery for IE, > > > > and one jQuery for others ... > > > > > The team is building jQuery already, using tools , so (I thought) the > > > > same tools might be used to build these non-ie and ie versions, also. > > > > And then tests can be done , using the same testing > > > > infrastructure ... > > > > > This is of-course not an "zero effort" excersize, but I think we all > > > > agree that the non-ie version will be measurably faster on on non-ie > > > > browsers than the original "mix-in-everything" version. And the same > > > > will happen with ie-only version on ie browsers. > > > > > --DBJ > > > > > --~--~---------~--~----~------------~-------~--~----~ You received this > > > > message because you are subs... --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "jQuery Development" group. To post to this group, send email to jquery-dev@googlegroups.com To unsubscribe from this group, send email to jquery-dev+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/jquery-dev?hl=en -~----------~----~----~----~------~----~------~--~---
