sgolayfilt no longer support SINGLE
in [Matlab]
|
From: sscnekro on 10 Jun 2010 13:06 Jan, very good. Maybe, step by step, we get there. I agree that piece is superior to negative emotions, unfortunately, in business one often needs to be assertive. If pieceful and compromising style was helpful to you, the better. Now, as you were so kind and gave your answers to my questions, I'd try to build on it a bit by a few YES / NO questions. Please read it first to the end (to the point), you will see that there is no need to write in answer to each question. [1] You work on software for clinical decisions. Do you find it important to involve your clients in software development? [2] You have much less clients than TMW. Suppose you'd go big. Your firm would not be able to communicate with each client individually, except for LARGE ones. Would you find useful, should one of permanent tasks of your *PR dept* be reading posts newsgroups to filter out information for management and product development decisions? To make the sites *represent statistically meaningful sample* of your highly diversified customer base, would you see usefulness of intelligent web design - so that as many customers with different licence, age, area of application as possible would be involved and keep returning to the sites? Would you find useful to run polls? > The polite guy forwarded the message to the "developper board" and there it waits and wastes diskspace for the last 18 month. But whenever I had a direct and personal contact [3] The problem called organization. Do you agree, that the software developers and engineers at TMW are simply right guys, trying to get the product quality and usefulness up to edge of own possibilities? Do you think if they would be receiving info from ML users as under [1] and [3], would they not take it into account? Do they receive such info, and of what quality? [4] You 'd make rather frequent product upgrades, adding new features to your software. Would find it mutually good to offer 24/24 technical assitance to your clients while they are implementing new realese? Woud you find it important to apply legal disclaimers with new releases? [5] Do you think, the software developers at TMW are less smart than you? If you are able of seeing that "A long-term-support version would be a great benefit and I'd expect that the experiences growing from such a version would help to improve the whole product.", 'd they not bee able of seeing the same? I would expect your answers 'd be: [1] Yes [2] Rather yes [3] Yes | Maybe not. [4] Yes | Yes [5] No | Yes. Fine, so coming to the point ant that is at the same time answer to your question: > What specific problems did you have with which tools? What are your experiences with the TMW support team? I like ML as a product. In general, I'm a fan of the guys who develop and engineer it. In particular, I respect Steven Lord, the others I don't know. I also find individuals at TMW helpful and quick (e.g. Shari), unlike "official ways". So, what is my problem? The TMW as an organization, the way processes involving customers are managed, the websites including newsgroup that simply ignore the potential of intelligent web based communication, and so on. There are companies delivering less complex products with much less responsibility (impact), yet they are willing to invest much more in PR and legal issues. And admittedly, this thread got me up.
From: Walter Roberson on 10 Jun 2010 14:58 sscnekro wrote: > [1] You work on software for clinical decisions. Do you find it > important to involve your clients in software development? I _do_ work on software for clinical decisions. Our clients do not have all that much input into the software development, as we develop the software and they hire us to run the software on their dataset and give them the information they are looking for. When it eventually comes time to make the software available in the clinics themselves, the software has to be completely rewritten anyhow to ISO 900x standards, if the software is for decisions about humans; after that the medical device approval process takes 2 years and costs approximately $2 million, if everything goes perfectly; we spin off companies to take care of that kind of work. Does this approach work? Yes. We are one of the most cost-effective departments, and our spin-off companies have been amongst the largest Initial Public {stock} Offers in Canadian history. (And to be clear here: we have no connection with TMW other than as full-price customers.)
From: sscnekro on 10 Jun 2010 17:20 > I _do_ work on software for clinical decisions. Yup. I'm trying to imagine that. In the first phase, they let you to compute everyting they need and you develop the software to enable such type of computations. In the second phase, the software is internalized, so that clinics personnel can run computations. So, what are the differences compared to ML and TMW? Do end-users of your software actively program in it? If not, they need to be involved in the software development to a lesser extent than ML users. The whole process is (much) > 2y, upgrades are not that frequent, every bug is eliminated before the clinics is "left on its own" with your product. Your product is tailored specifically for the client. TMW faces a much more challenging situation. The customer base is highly diversified. Your software is a "ready to use" end-product whilst Matlab serves as an input for further programming. There can go much more wrong with ML than with your software. Etc. But, TMW spends much less time / money / efforts *with the clients* being involved in *their processes* than you / the spin-offs. Faces more challenging - - spends less. And that's the point. In my view, the quality and depth of customer related processes at TMW is somewhat inadequate to the nature of the products. I can even imagine that TMW developers / engineers would like to be more involved with clients, prolong the testing phases, work on a stable version etc., but such decisions depend on management / other depts. If they prefer to do it that way, good, but then they 'd at least need a strong legal dept and smart disclaimers.
From: Jan Simon on 10 Jun 2010 19:02 Dear Sscnekro, > [1] You work on software for clinical decisions. Do you find it important to involve your clients in software development? I'm not really sure about this question. Of course the clients are involved in the design of the software, because the tools should solve their problems. But the development details are my part of the work. It is even hard to define, who the client is: The physio-therapists who perform the measurements with the patients, the surgeons who need the results, the patients or the health insurance who pay? > [2] You have much less clients than TMW. Suppose you'd go big. Your firm would not be able to communicate with each client individually, except for LARGE ones. I cannot go big. For the actual scientific work in my software it was necessary that I understand the problem. I had to perform the measurements by my self, I had to look onto the bones during the surgical operations, I had to be in touch with many patients personally. We measure e.g. gait patterns of 3 year old children and I had to learn that the best software design is worthless, if the person, who instructs the child to walk a specific way, wears white cloths or the child is hungry. A personal and physical contact to patients and software users costs a lot of time, so I cannot go big. As I've said before, I think that anonymous communication is inefficient. Therefore I decided not to do the same. > Do you agree, that the software developers and engineers at TMW are simply right guys, trying to get the product quality and usefulness up to edge of own possibilities? Some of Matlab's toolbox functions are amazingly ugly and inefficient. There are a lot of tips from MLINT in Matlab 2009a. Therefore I have the impression, that the quality of Matlab functions has a Poisson distribution. > [5] Do you think, the software developers at TMW are less smart than you? If you are able of seeing that "A long-term-support version would be a great benefit and I'd expect that the experiences growing from such a version would help to improve the whole product.", 'd they not bee able of seeing the same? The developers do not see my special needs, if I do not tell them. This is not a question of smartness. > So, what is my problem? The TMW as an organization, the way processes involving customers are managed, the websites including newsgroup that simply ignore the potential of intelligent web based communication, and so on. There are companies delivering less complex products with much less responsibility (impact), yet they are willing to invest much more in PR and legal issues. And admittedly, this thread got me up. Let me ask you again, because I'm really want to know: What is *your* problem with Matlab or TMW? You point to problems of TMW. All big organizations loose the contact to the users: Microsoft does not offer the tools I need, my goverment does not care for my specific problems, and it is a hard bureaucratic work to get some hours for observations with the Hubble space telescope. Human brains are designed for group sizes of 12 persons. Unfortunately it is obviously impossible to build big things with 12 persons: computers, cars, skyscrapers, newspapers, helicopters, or a programming framework as Matlab. This is the same problem since we tried to build a tower in Babylon. Anyhow, this has no relation to Savitzky-Golay-Filters anymore. I think, we should wait to solve the problems of TMW until they employ us, pay enough money and allow us personal communications with users and developers. Good night, Jan
From: Joaquim Luis on 10 Jun 2010 19:41
Honglei Chen <Honglei.Chen(a)mathworks.com> wrote in message <hq2pim$599$1(a)fred.mathworks.com>... > Hi Bruno, > > We take backwards compatibility very seriously. You do??? So may I ask, where is the Maltab compiler right now? |