jQuery 1.4.2 buggy
in [JavaScript]
|
From: David Mark on 24 Mar 2010 11:16 Hans-Georg Michna wrote: > On Sun, 21 Mar 2010 19:13:39 -0400, David Mark wrote: > >> ... The OO documentation is truly lagging >> behind at this point though. I'm working on that. > > Guess I'll wait for it. You don't have long to wait now. Ethan B. generated a framework for it and I am filling in the blanks today. There will now be two references: one for the API and one for the OO interface. > >> In the meantime, see >> the TaskSpeed tests. Comparing the various sets of test functions for >> each library should be enlightening. Despite its many documented flaws, >> the suite does at least allow you to compare techniques for common tasks >> between the various libraries (hover over a result to see the function >> that returned it). > > Actually I am not all that interested in the performance itself. > I consider the most valuable side-effect of performance tests > that the give an indication of the general quality of the > software. Yes, but the test functions used, when compared to those of the other libraries give you a good reference for the OO interface (as that is what I used for all of them). > >> It is My Library (the "my" refers to you). Start with a minimal build >> (optimally the features that interest you) and see the examples on the >> builder test page. There is even a console to view the internal >> machinations in real time. Then there is the Examples page, the >> discussion forum and the (there is you look) documentation. Barring >> that, you can ask me. It's worked for a lot of people so far. :) > > I don't work that way. I need definitions. I need to know in > advance what exactly each function does. Examples are certainly > very welcome too, but ultimately I need the definitions. Coming! :) > > I'll wait patiently. I believe there is a widespread need for a > software layer between the programmer, particularly the average > programmer, and the web browser, with several quirky browsers > around. There is definitely a demand (if not a need at this point) and inexplicably it has never been filled in competent fashion. > > A few examples of functions that differ from browser to browser > and that could do with some software layer help are Ajax > functionality, positioning, innerHTML limitations, events, and > several other functions that differ among DOM implementations. All that is certainly covered in spades. > The almost too popular CSS selectors may also be a good thing. > The more experienced professionals will probably not overuse > them anyway. > I don't consider them a good thing, but other than a few minor issues (e.g. QSA add-on should pass elements through to the original, operating only on documents) and a couple of missing CSS3 selectors (e.g. not:), I am confident I have the best one. Hell, none of the others has I had left off a few weeks ago working on XML support. I'll wrap all of that up before the release.
From: David Mark on 24 Mar 2010 11:31 David Mark wrote: > Hans-Georg Michna wrote: >> On Sun, 21 Mar 2010 19:13:39 -0400, David Mark wrote: >> >>> ... The OO documentation is truly lagging >>> behind at this point though. I'm working on that. >> Guess I'll wait for it. > > You don't have long to wait now. Ethan B. generated a framework for it > and I am filling in the blanks today. There will now be two references: > one for the API and one for the OO interface. > >>> In the meantime, see >>> the TaskSpeed tests. Comparing the various sets of test functions for >>> each library should be enlightening. Despite its many documented flaws, >>> the suite does at least allow you to compare techniques for common tasks >>> between the various libraries (hover over a result to see the function >>> that returned it). >> Actually I am not all that interested in the performance itself. >> I consider the most valuable side-effect of performance tests >> that the give an indication of the general quality of the >> software. > > Yes, but the test functions used, when compared to those of the other > libraries give you a good reference for the OO interface (as that is > what I used for all of them). > >>> It is My Library (the "my" refers to you). Start with a minimal build >>> (optimally the features that interest you) and see the examples on the >>> builder test page. There is even a console to view the internal >>> machinations in real time. Then there is the Examples page, the >>> discussion forum and the (there is you look) documentation. Barring >>> that, you can ask me. It's worked for a lot of people so far. :) >> I don't work that way. I need definitions. I need to know in >> advance what exactly each function does. Examples are certainly >> very welcome too, but ultimately I need the definitions. > > Coming! :) > >> I'll wait patiently. I believe there is a widespread need for a >> software layer between the programmer, particularly the average >> programmer, and the web browser, with several quirky browsers >> around. > > There is definitely a demand (if not a need at this point) and > inexplicably it has never been filled in competent fashion. > >> A few examples of functions that differ from browser to browser >> and that could do with some software layer help are Ajax >> functionality, positioning, innerHTML limitations, events, and >> several other functions that differ among DOM implementations. > > All that is certainly covered in spades. > >> The almost too popular CSS selectors may also be a good thing. >> The more experienced professionals will probably not overuse >> them anyway. >> > > I don't consider them a good thing, but other than a few minor issues > (e.g. QSA add-on should pass elements through to the original, operating > only on documents) and a couple of missing CSS3 selectors (e.g. not:), I > am confident I have the best one. Hell, none of the others has I had > left off a few weeks ago working on XML support. I'll wrap all of that > up before the release. Oops, got distracted by a call. That unfinished sentence is supposed to read: "Hell, none of the others has attribute wrappers that are close to competent, which certainly impairs their ability to query documents in reliable or consistent fashion."
From: David Mark on 25 Mar 2010 14:25 Stefan Weiss wrote: > On 19/03/10 19:51, David Mark wrote: >> Thomas Allen wrote: >>> On Mar 19, 2:42 pm, David Mark <dmark.cins...(a)gmail.com> wrote: >>>> Root mean square? >>> At the risk of having no sense of humor, RMS is a common abbreviation >>> for Richard Matthew Stallman, founder of the GNU project and all- >>> around FOSS (free, open-source-software) fanatic. >> No worries, Thomas; I don't know him from Richard Dean Anderson. :) > > At the risk of having no sense of humor, IMHO, some deeper understanding > about the free/open software movement and the philosophies and > individuals at its center could greatly benefit your project. Much as I > loved MacGyver as a kid, RMS has a rather more profound message than > "all I need now is a hairpin and some bubblegum". > > You are now the author of an open source library, and you apparently > want people to participate in your effort. Some anecdotal feedback from > one potential user (=me): I think your library has a lot of potential, > but I wouldn't dream of actually using it unless and until it is truly > open. And that means open sourcing all of it, including the builder and > the raw source it uses. It also means having a public bug tracking > system and a public code repository (with versioning and releases) which > contains all of the code required to build and maintain the project. The > current builder code may be messy, and ASP may not be the way to go, but > having it out in the open is way better than having no source available > at all. > BTW, there is a public bug tracking system. It is built into the repository. I am going to announce in the forum that users should post any issues to it. http://code.google.com/p/ourlibrary/issues/list Thing is, since the initial push to expose feature detection breakdowns in ancient browsers, there haven't been a hell of a lot of bug reports (though I know some bugs exist). I'll start posting known issues there as well. Fair enough? I still don't see how the builder code matters much (most users seem to favor the full build), but there is talk of porting the builder to PHP. So you've got that to look forward to. :)
From: Hans-Georg Michna on 26 Mar 2010 09:12
Sounds very good. I'll be reading about the progress here. Thanks for doing this work. I'm hoping for your success. Hans-Georg |