sgolayfilt no longer support SINGLE
in [Matlab]
|
From: Walter Roberson on 11 Jun 2010 12:33 Jan Simon wrote: > The new list of open/solved bugs is a very good step in the right > direction. A tiny history including new/changed/lost features and > open/solved bugs in each M-file would be more user-friendly. "Very" > serious backward(s?) compatibilty demands for a full documentation of > changes reachable by users. I would also recommend that the documentation for each command include an indication of when it was introduced, and when new option are introduced, that the release for them be documented as well. I would recommend that this be done for each command, not just new commands.
From: Steve Eddins on 11 Jun 2010 12:35 On 6/11/2010 10:54 AM, Jan Simon wrote: > Dear Honglei! > >> We take backwards compatibility very seriously. > > I have the impression, that TMW takes backward(s?) compatibility > seriously -- without "very". > Most of all I miss a list of changes in the functions. E.g. if some asks > in the newgroup for the FIND command and uses Matlab 2006b, I cannot > find a message in my 2009a doc of FIND, when the 'first'/'last' option > has been added. I cannot find the doc of 2006b in the net, therefore I > have to search in all List-Of-Changes pages from 2007a to 2009a, if I > want to find the version including the enhanced FIND. Jan, I'm personally very interested in this comment and I want to respond to it. But I'll preface my response by saying first that we hate, Hate, HATE to talk publicly about anything that's not already completely finished and released. That said, I'll tell you that user feedback in recent years has been loud and clear on this point. We are seriously looking into it. > [snip] > > Nevertheless, it is easier to port Matlab source from Matlab 4.2c to > 2010a than a function calling the API of Windows 3.1 to Windows 7. > Another example: Try to port source code from K&R-C to C89 to C90 to C95 > to C99. > > [snip] Thanks for making this point. In discussions like this, many folks seem to assume that languages such as C or C++ are more stable than they actually are. But even within the existing version of C++ standard, just upgrading our C++ compilers to the latest vendor release always causes us a lot of headaches in our product builds. (Don't get me wrong - I'm not trying to make excuses for when we cause similar headaches for our own customers. I'm just trying to clarify the "compared to what" question.) > I think, TMW can be proud of the reached status of backward > compatibility, and could be encouraged to improve it until the degree of > perfection is reached to become the "ultimate computing environment" and > claim to have an "outstanding service to our customers". My perception of the attitude within development is this: We know that we have improved the way we approach compatibility issues, but we also know that there's still a lot more that we should do. Thanks for your thoughts, --- Steve Eddins http://blogs.mathworks.com/steve/
From: Jan Simon on 11 Jun 2010 17:00 Dear Steve! > But I'll preface my response by saying first that we > hate, Hate, HATE to talk publicly about anything that's not already > completely finished and released. That said, I'll tell you that user > feedback in recent years has been loud and clear on this point. We are > seriously looking into it. "hate, Hate, HATE"?? Where do these strong emotions come from? Getting feed*back* only after new features are released does not allow to use the strong power of a feed*forward* control. How do TMW determine the needs of the users, if they/you avoid discussing about not finished tools? Kind regards, Jan
From: sscnekro on 11 Jun 2010 17:44 >That said, I'll tell you that user feedback in recent years has been loud and clear on this point. We are seriously looking into it. As Steve Eddins describes it, it seems that in reality at TMW they 'd receive a mass of feedback from a broad audience of ML users. As for myself, I have not encountered a standardized possibility (a poll, questionnaire or similar) to discuss my experience with ML and these websites. I know that I can make an e-mail to one of the TMW contacts or write to this newsreader, but that is not standardized communication. For instance I'd like to be asked about help documentation and in particular, on these websites. Does anybody know such possibility? BTW: I discovered happy company to have a PR dept 3 (!!!) clicks from main site: http://www.mathworks.in/company/pressroom/contact.html Jack Little has last posted to the newsreader >10y ago: http://www.mathworks.de/matlabcentral/newsreader/author/245 What are your guesses, Folks, how many legitimate posts will need to be added to this thread until he 'd respond? Hey, just pls FEED IN more ideas on what we could undertake.
From: Jan Simon on 11 Jun 2010 18:04
Dear Sscnekro, > Jack Little has last posted to the newsreader >10y ago: > http://www.mathworks.de/matlabcentral/newsreader/author/245 Sorry, I missed something: Why do you try to contact Jack Little? Jan |