Fine, whatever. I'm tired to arguee about this issue. I am still planning in improving Your code further. Adding delay in milliseconds to call the ajax funciton so that fast data entry may call only once to the server and refactor onblur event so that it will select the last autocomplete result if one choice only was returned by the ajax call.
Also Scrollbars and scrollTo functions so that keyboard arrow events would trigger the srolling followup. f(t) On 7/27/07, Janne Hietamäki <[EMAIL PROTECTED]> wrote: > > Heh, I don't care what you do with my code, it's nice to see it's > used, but I don't want to have duplicated code/bugs in Wicket. > > If you compare this: > > http://papernapkin.org/pastebin/view/1095 > > and this: > > https://svn.apache.org/repos/asf/incubator/wicket/trunk/jdk-1.4/ > wicket-extensions/src/main/java/org/apache/wicket/extensions/ajax/ > markup/html/autocomplete/wicket-autocomplete.js > > There are for sure a load of duplicated code. Do you think i don't > recognize my own code? :) > > > Janne > > On 27.7.2007, at 1.30, Francisco Diaz Trepat - gmail wrote: > > > I see you still claiming copy except for the comments. at least you > > gave me that. > > petty you fail to specify as I ask. > > > > On 7/26/07, Janne Hietamäki <[EMAIL PROTECTED]> wrote: > >> Hi, > >> > >> I did not mean to criticise you in any way, but for sure there is a > >> quite a lot of duplicate code/functionality in both javascript and > >> java. When I compare these two versions of javascript, most of the > >> code is exactly equal (ok, you have added some comments). Duplicated > >> code means always duplicated bugs, and that is always a bad thing. > >> Specially with javascript is bad, because a new browser version tend > >> to break things that have earlier been working just fine. When almost > >> the same code is in two places, the work needed to fix all the bugs, > >> test everything etc. is duplicated. Less is really more. > >> > >> I see no point of having alternate version of autocomplete for doing > >> almost the same thing as the old one, that will only bloat the API > >> and increases the amount of code. I would like to see the > >> autocomplete extended so it can handle the use case you have. That > >> should be quite straightforward thing to do, and really, you have > >> already done that. Now we have to find a way to integrate you work > >> into Wicket. > >> > >> Patches are a good thing, because now it seems you have worked with a > >> version from the 1.2 branch, but trunk already has some fixes which > >> your code does not contain. It's easier to handle the changes with > >> patches, so we can see what you really have changed. > >> > >> Janne > >> > >> > >> On 26.7.2007, at 21.39, Francisco Diaz Trepat - gmail wrote: > >> > >> > Janne I thought some more about this mail. And I wanted to say a > >> > couple of things: > >> > > >> > 1) It is not my intention in any way to replace the code for the > >> > auto-complete you and Igor developed. > >> > > >> > 2) It is not my intention to criticise it either. > >> > > >> > 3) I had to develop an alternative, as I wrote in the code, I based > >> > the extensible autocomplete code on the one that you and Igor > >> > developed. > >> > > >> > 4) Here are some Print Screens taken to the Open Source WinMerge > >> > application on the tree key files which are the JavaScript, the > >> > Abstract extension of the AbstractDefaultAjaxBehavior and the > >> > extension of that that implements the renderer interface. As you > >> > might notice the codes are far from similar and far far far far > >> > away from a copy paste. > >> > > >> > 5) If a patch is the proper thing to upload instead of whatever I > >> > did, then please tell me and I'll be more than happy to follow the > >> > procedure, although it was not my intention to submit a patch to > >> > anything, just a new feature based on an existing one. > >> > > >> > To finish up, If some where on the way of contributing this > >> > autocomplete version I steped on anybodies toes then I am sorry and > >> > can bet that it was not intended at all. I am not here to compete > >> > in any way with anybody. > >> > Wicket struck awesome to me and to my team mates as well. And > >> > inspire me and them to help out whenever possible without > >> > compromising company policies. We hope to be able to keep > >> > contributing but don't want to get wrong impressions. > >> > > >> > Respectfully, > >> > Francisco Díaz Trepat > >> > > >> > > >> > On 7/25/07, Francisco Diaz Trepat - gmail > >> > <[EMAIL PROTECTED]> wrote: > >> > Would you be a bit more specific please? > >> > > >> > > >> > I thought I rewrote most of it. Perhaps I am mistaken. > >> > > >> > > >> > Which part? > >> > > >> > > >> > Best Wishes, > >> > f(t) > >> > > >> > > >> > On 7/25/07, Janne Hietamäki <[EMAIL PROTECTED] > wrote: > >> > Hi, > >> > > >> > It seems you have copy-pasted a quite bunch of code, and if you're > >> > going to develop this further there is no point for me to try to > >> > remove the code duplication and try to integrate the code with the > >> > old autocomplete. Also, patches would be nicer. > >> > > >> > Janne > >> > > >> > On 25.7.2007, at 23.14, Francisco Diaz Trepat - gmail wrote: > >> > > >> > > I'm such a user... So anxious some times... > >> > > > >> > > Thanks, I have now attached a file named AutocompleteTestCase.rar > >> > > (badly > >> > > named, but...) with the code and a page with 3 samples. > >> > > > >> > > Until I am a master in maven and contribute some more, please be > >> > > patient > >> > > with details, like not putting all in the quick-start given > >> sample > >> > > (as igor > >> > > suggested correctly). I'll get there. > >> > > > >> > > ++ > >> > > > >> > > I'm still working in the code thinking now I am going to add two > >> > more > >> > > features to it: > >> > > > >> > > *AUTOSELECT_ON_ONE_RESULT* = true / false configuration; > >> > > *Selects the choice without further need of interaction if > >> only one > >> > > choice > >> > > is returned by the ajax call. Good for zipcodes, areacode, and > >> > > general code > >> > > data entry.* > >> > > > >> > > DELAY_AJAX_CODE_MS = number of milliseconds to delay the call to > >> > > the server > >> > > to look up choices. Again good for zipcodes, and general code > >> > > entering data > >> > > entry slaves who usually enter data very fast. > >> > > > >> > > regards, > >> > > > >> > > f(t) > >> > > > >> > > > >> > > On 7/25/07, Eelco Hillenius < [EMAIL PROTECTED] > wrote: > >> > >> > >> > >> On 7/25/07, Francisco Diaz Trepat - gmail > >> > >> < [EMAIL PROTECTED]> wrote: > >> > >> > Wait a minute, I am creating the jira but when do I upload the > >> > >> code? > >> > >> > >> > >> Right after you created it. > >> > >> > >> > >> Eelco > >> > >> > >> > > >> > > >> > > >> > > >> > >> > >
