Announcement

Collapse
No announcement yet.
X
  • Filter
  • Time
Clear All
new posts

    Improved Logging

    When a user application executes a smartgwt api method expecting something to happen, and that method decides to do nothing as a result of sequense or contextual related issue, I would like to see it logged at a warning level.

    Giving users more information about these kinds of scenarios will reduce tech support threads.

    #2
    You've found one single method, out of around 8000, that is not doing what you expected in a particular circumstance, and the cause is unclear and could be a usage issue (it hasn't been looked at over here yet). This is not a general problem, so please stop posting about this as though it were a general problem unless you have several other examples.

    Comment


      #3
      Perhaps I am continuing to post because even after providing code, logs and lots of description info - my posts have been completely ignored.

      My critisism is completely constructive and intended to beter things for everybody - your attitude however is a mystery.

      Comment


        #4
        BTW: ListGrid.hideField() and ListGridField.setWidth() are two examples.

        hidefield() does not appear to work unless called after draw and setwidth() only when called on a ListGridField that was explictly created (Not a ListGridField obtained from the datasource)

        I dont know what the magic number of samples I have to provide to meet your requirement for a valid post - but since I have managed to find two examples having worked with less than 1% of the smartgwt api... I find your assertion that this was just 1 example out of 8000 methods highly dubious... to say the least.

        Comment


          #5
          Your assertion that there is a pervasive problem is not plausible simply because if it were true, there would be hundreds of other posts about it. So please read the FAQ and create standalone test cases for anything you believe is a bug. This is the proper course of action for any framework that is obviously working so very well for so many people.

          Comment


            #6
            Ok so you want to argue semantics... I'll let you have the last word. Clearly the people posting all the issues on this forum are posting about some other framework and improved logging and documentation is not something this framework could benefit from.

            Comment


              #7
              ???

              Documentation can always be improved. Logging can always be improved. Features can always be added. If you want to be constructive, you do not complain about documentation at large, which is not actionable criticism. You provide test cases for bugs, patches, specific documentation improvements for specific under-documented methods.

              But more specifically, you do not take one method which you *speculate* is broken and generalize that into a critique that all other methods might be similarly broken. Not only is that not constructive, it's just incorrect.

              As you may have realized, we basically must respond to unfair criticism of the product to combat misconceptions. But we do so in lieu of answering actual questions to the detriment of all other users. So if you want to participate in the community and not detract from it, a starting point would to be avoid wasting support's time in this way.

              Comment


                #8
                1. Neither of my posts to this forum were so vague as to say simply "Improve documentation" - your remark is unfair. I specified what aspects of the documentation I would like to be improved (You may see that as splitting hairs but I am certainly not going to research and provide an example of every situation like this I come across - I was referring to a general policy of the kinds of things that should be logged).

                2. I did provide 2 examples of apis that have this problem and I believe that my explanation is sufficient. Given... as I have already said that I have found 2 instances in such a small sample size - I think the idea that there are more examples is not beyond the realm of possibility - although "pervasive" was your choice of words not mine.

                3. And most importantly... Dude... we may disagree on exactly how a problem report should be submitted but please note: this is the "Wish List" forum. My initial post in this thread simply documented a general observation. I jotted it down for the record because I was starting to see a trend that disturbed me.

                It was you that sparked all this by telling me "not to post" after ignoring all my previous posts... you cant have it both ways...

                And yes it has not escaped my notice, that complaining gets a response when posting a question (regardless of the quality) may not get anything. I ask you to consider, that a short reply asking for more info would probably have cost you less time than engaging me in an ongoing debate.

                Originally posted by Isomorphic
                ???

                Documentation can always be improved. Logging can always be improved. Features can always be added. If you want to be constructive, you do not complain about documentation at large, which is not actionable criticism. You provide test cases for bugs, patches, specific documentation improvements for specific under-documented methods.

                But more specifically, you do not take one method which you *speculate* is broken and generalize that into a critique that all other methods might be similarly broken. Not only is that not constructive, it's just incorrect.

                As you may have realized, we basically must respond to unfair criticism of the product to combat misconceptions. But we do so in lieu of answering actual questions to the detriment of all other users. So if you want to participate in the community and not detract from it, a starting point would to be avoid wasting support's time in this way.

                Comment


                  #9
                  Your first post provided zero specific examples, so it was neither actionable nor constructive. If you *later* provide concrete examples, this does not bear upon your first post, or on Isomorphic's reaction to it. This is all very obvious.

                  Instead of circuitous justifications of your criticism of the product, can you please just recognize that several prominent community members have reacted negatively to your critiques of SmartGWT and/or your expectations of free support.

                  It's very simple: in the absence of a clear test case, don't claim there are problems, that's a waste of time and may force Isomorphic to defend our product against unjustified claims.

                  Comment


                    #10
                    You know I am honestly getting tired of responding but as long as you keep baiting me with these inaccuarate and exagerated responses I feel inclined to continue.

                    Once again... it was a GENERAL response to a wish list forum... it was never intended to be a specific actionable item - so please drop that.

                    I will certainly accept your comments about providing test cases for even remotely complex support issues - and just as documentation and logging can always be improved - so can the level of information provided to a support person.

                    I dont know how many "prominent" members have reacted negatively to my posts - I only know of 1 id called Isomorphic (and one other person whose comments I did not find overly negative). I deal with people based on the content of what they post - The numbers are proof of very little to me.

                    With regard to your last paragraph... I've seen my posts in several places characterized with a lot of very negative words like "deeply flawed", "pervasive", "critisism", "defend our product" etc - thats an awful lot of bile - and is largely out of proportion with my words or intent.

                    I dont see any need for you to "defend your product" - as there was no attack. I have noted two places where a method was called and silently did nothing - Yes I could provide a test case but at this point I have lost all interest in doing so.

                    Comment


                      #11
                      toggle57,
                      Stop bickering and bantering and move on to more productive tasks and ask specific questions if you run into issues, and follow the rules of the forum as indicated in the SmartGWT FAQ sticky.

                      Constructive feedback is welcome but you seem to be valuing your feedback more that what others (myself included) think of it. So stop being argumentative just for the sake of being argumentative. Doing so will only make people less inclined to answer your posts. You are just one of many users of the free LGPL product and being aggressive and argumentative will not get any special attention or priority response.

                      Now can we please end this thread, and any other thread where you have provided "feedback". You were pointed to the javadoc issue in tracker so add your comments in there if you want to see any improvements.

                      Sanjiv
                      Last edited by sjivan; 3 Dec 2009, 04:38.

                      Comment


                        #12
                        I do so happily, I simply could not let that last post stand unanswered.

                        Comment


                          #13
                          Can't believe I actually took time to read through this whole thread... :)

                          Comment

                          Working...
                          X