CommonSense

QueueYueue, I forgot to mention that I have looked at the OO Plugin Framework that uses the Factory Pattern. So many patterns to choose from ... CS

QueueYueue, I've watched Trevor's and AQ's presentations, and have built some basic OO code and now know quite a bit more than I did, but this is a continuous learning process of course. One other question to you is that while I am looking at and trying to continue the design process that you suggested, what design pattern(s) that have been mentioned here on LAVA and on the NI website, is closest to your suggestion? This will greatly help my understanding of the design implementation process - at least for this particular use case. I have been reading the forums on the various pro's and con's of the current version of AF, versus say MessagePump and Lapdog, and I am leaning toward Lapdog a little as I would like not to be limited in the kind of customizations down the road. Just my 0.02$ at this point. Thank you, CS

QueueYueue, Could you point me in the direction to what OO example project that I should take to learn before going on to tackling the AF ? Thank you, CS

QueueYueue, One more clarification / point per a paragraph I wrote above; "When the TCP Client starts up, it reads a text file that says how many of each type of tabular and XY pages the operator has configured, the name of the tab page, with the associated channels and attributes on each page that he/she would like to monitor. The operator can edit this text file only via a deparate dialog / vi program that opens up from the menu on the Main vi. " I meant to say that when the tcp client starts-up, it reads a text file as to how many of the maximum amount of tab pages are displayed (not hidden). With the OO eventual design, the new tcp client application will still read a text file to the quantity and type of test pages that are of interest to the operator. Maybe after reading the text (*.ini) file at start-up, the Main vi has the available data displays that he/she can select from a menu list via the menu bar above. I am assuming now from the old/current design that they can only show one data display at a time (i.e. one subpanel). A new and somehwta more complicated design may allow the operator to show several data displays at one time that are separately spawned / re-sizable Windows? Anyway, thank you again and I am now returning now to read & properly digest your original post, CS

QueueYueue, Thank you for taking the time to formulate a design framework per my, yes, vague requirements/use case. Before I begin to digest the class structure, I'd like to say more about the TCP client use case: -Connection Use Case: 1.) When the client application starts up, it is not connected to a (the) Server. When the operator decides to connect to a certain Server, they must enter the IP address, the Port, a configuration alphanumeric value, and another optional configuration piece of information. 2.) When the client connects to the server, the server first sends a list of the available data channels, plus channel attribute information. 3.) There may be several test servers that are running concurrently in different labs, but the client can only connect to one Server stream at a time. 3.) When the tcp client operator disconnects from the Server, they'd like to still look at the data on the display (tabular or time history) and not have it go away until they close the program. Sometimes, the Server has to go down for h/w changes, then come back up in 10 minutes to 2 hours and while this is happening, the test engineer may need to still analyze the data that was acquired and streamed up until that time. Note: The Server type is a non-LabVIEW written program that is also a data acquisition system with a separate module that handles the Server-Client connects / disconnects and the streaming of data along with stream status commands in certain XML tags. The Server hardware is usually a Win7 Dell precision desktop PC. -Current LabVIEW Client Design w/ single Tab Control: The tab control is designed to have (for example): 1.) A maximum of five (5) pages that display tabular data (2 columns, 25 rows each). 2.) A maximum of five (5) pages that display time history data (Qty=1 XY chart, with up to 10 channel traces per chart) When the TCP Client starts up, it reads a text file that says how many of each type of tabular and XY pages the operator has configured, the name of the tab page, with the associated channels and attributes on each page that he/she would like to monitor. The operator can edit this text file only via a deparate dialog / vi program that opens up from the menu on the Main vi. Note #1: The time history data is handled via a single LV2 style global and not with the XY graph. I pre-allocate x amount of space at startup. Note #2: The current list of customers that are using the tcp client are looking at data that is acquired at 1 to 5 Hz, for anywhere from 30 minutes to 24 hours. The above use case information I was hanging on to until / if someone responded to my OO design query. From what little I've read of your original design thoughts, I'm pretty sure that most of it still can be used with some modifications. I've probably have gone now from vague to overload of requirements! Thank you again, CS

Hi, I am making a determined effort to migrate a particular application of mine over the fence to OO, but I would like to be "herded" towards using the correct or proper design pattern(s). This application of mine is a TCP Client app that allows the user to connect to one (1) test server as to display test data in various forms (i.e. tabular, time-histories, gauges, etc.) using a single tab control. There can be up to ten (10) such Clients logged in at one time to the same Server. I have read recently that several folks on NI & LAVA forums advise to move away from the tab control and to use sub panels instead. Combine that with this web page's neat UI plugin example, I seem to be getting close to finding the proverbial "light switch" in the dark room, but ... it could be something else .... Some notes: I do have LabVIEW 2012 at my disposal. The test server is written in C++. Down the road a bit we may need to encrypt the XML stream so I may need to put a LabVIEW web service in between the Server and each Client using SSL - if that's a future design bit to note here but not as important as the basic environment or use case of a TCP Client that displays data via several types of displays - which the display types will only grow in the near future as we all have experienced. So adding different displays should be as "easy" as possible. Note: My question was originally posted here, but I was advised to post it here on LAVA for a more experienced architect to respond. Thank you very much in advance !!! CS

Hi 'yall (some NI Week rehearsal here ...), Its been several months now since my last post, so here is the obligatory update: I have implemented a TSVN Client and Visual SVN server architecture - including JKI's TSVN plugin which works great ! - for about 3 months and everything has been running very smooth I must say. We determined that we needed a centralized repository (RAID6 supermicro server) for now, but we will re-analyze this architecture every year to see if we need to move to something like Mercurial which looks like another big step forward. We pretty much work on our own code, but have left the door open for multiple people working on the same project. Once we got the just of how to setup the repositories on the server and on each client, it has been one of the best moves we have ever made. VSVN is such a low-overhead breeze to setup and maintain, the bar has been set very high. Maintenance is so very low and we don't have to rely on someone in the company I.T. group (i.e. you know, could be in a cube on Mars) to get a VOB going or with new PC's and O/S changes occurring now, wonder where the VOB patch is or ... you get the point. The bar has swung greatly back to retaining knowledge so one can be self-sufficient and not rely on others to get the job done. I can't emphasize this enough. Thanks to everyone here with your suggestions !

All, I just met a Linux text programmer who has used svn/cvs/rcs/sccs and ClearCase and says that git is as good as SVN. ClearCase was a mess considering the amount of servers/services that need to be setup (i.e. vob servers, view servers, vob admins, modified kernels, etc.). He also passed along the following git favorable benchmark website : http://whygitisbetterthanx.com/ Anyway, I've now completely run the original topic (vehicle) off the road so to speak and now I am also looking in the git direction. Signed, A little confused but seem to be heading in the right direction ...

All, Thank you for your responses but it looks like I need to clear up a few details : 1.) ClearCase DOES cost us $$$. 2.) I know of a small group that uses Perforce and witnessed it in action and the process of checking out / in was MUCH easier than ClearCase !!! .... but Perforce is not free. 3.) The Open Source (i.e. free) allowable choices are : CVS / TortoiseCVS, SVN / TortoiseSVN, git, or Bazaar. These are the boundaries that are set for me. So, I don't want to hear about any other SCM tools out there - but thank you for mentioning the other products. The focus of my topic is that I'd now like to hear from SVN / TortoiseSVN users out there. Thanks !

Hi, I am looking for any and all comments on this topic from the LAVA community : I am a LabVIEW programmer who has been "forced" to use Clearcase for several years now by the text based coders in the dept. It seems that there are now more LabVIEW programmers that use TSVN and SVN than any other SCC tool out there in industry (my perception), and I am sure for good reason and not just due to cost savings !!! I would like to present a functionality / ease of use comparison case between ClearCase & TSVN / SVN to justify the potential migration to TSVN & SVN. Thank You in advance !!!

Back from the holidays, Happy New Year All !!! Seems a little cold around the world these days ... hmmm (checkmate .... again). Now back to the lecture ... Crelf, you have proven that being a good LabVIEW programmer (not a fighter ...) does not carry over to having a rational debate about the GW scare, and the related real topic of re-distribution of wealth. Your probably love socialism as you are too young to have experienced anything else. A young socialist that has been nurtured to depend huge government for all of your needs. One tip: Any country with a constant unemployment rate (over say 10-15 recent years time) of above 4-6% is a system that is not working. Hmmm, lets look at some European socialist countries like France & Germany for example .... do I see 7, 8, 9 10 and above ? So to uncover your (un) consciouness some more, you also have sipped the coolaid into believing that the early 1910's style American Progressive movement, rooted from facism and what the globama folks want to see here is a good one. Most young and old progressives (read re-gressive & communist/facist) when I quiz them don't know the real roots of where it came from and that it has been around for about 100 years in the USA. But it fails each time because too many US cisitizens have read and understand what the Framers warned about that a degradation of freedom and increased gov control withs its associated tryanny will start in this type of form (like a bad fungus). I truly hope you haven't drank the progressive coolaid !!! Hopefully your not this Seond tip: look up Cloward & Pivan which is along the same lines "Rules for Radicals".. Yes, the big transformation that was announced by globama is this. See who he sourrounds himself. See what domestic & international (cap n tax) policies fit the design of the collaspe of the US. How does this fit in with the Founding Framers designs? Answer: it doesn't - at ALL !!! As for the Solar project being canceled, the reason was that this project was out of scope from division's main product line which is propulsion power. As for Capitalism and innovation among the private sector, many other companies in the US have and are pursuing this, like Boeing's owned sub which currently has the most efficient panels. Yes, Capitalism does work !!! But you already knew that ? I sure hope so .... if not, then I wonder about your other beliefs ... Summary: The GW scare is badly broken scientific process. Just like several hundred years ago "The Earth is Flat" mantra of "now lets don't question us Elites" and lets move zillions of money around the globe so we can FEEL better. Funny, during Moore's Law, at the same time common sense in the scientific community was 1/2 ing at the same rate. Which side of current history are you going to tell your children that you decided to be a part of. Relativsm (SF style even - yes I have relatives in the Bay area who are finally coming back from Planet Loon, but some aren't). The 60's experiment has failed. It fails the Math (ask the State Legislature in California). This is just another related topic. Hopefully soon (like beginning November 2010), true Math will prevail as does common sense and good processes!

All, I agree with Bob Hamburger that we need to get away from using oil via safer energy sources via Manhattan like Project effort. Not only to have clean energy, but also as a National soveriegn security measure to remove relliance on folks who devise daily to remove us from the planet. I've was involved in a project some 7-8 years ago that was a LabVIEW solar panel testing system (Control & DAQ) but the plugged was pulled due to other funding priorities at our large Aerospace company. This topic is not tied to (good) science, its all about re-distibution of wealth. Before Al Gore pushed his own scare button in the mid 1990's, there has always been via the "very moral" UN an effort by certain countries to redistribute wealth around the world. This is just another effort in that vain. None of these countries I am speaking about here and have since my mid 20's or mid 1980's are true free market capitalists (note much different that a corporatist). The pre GW alarmists were / are all from non or very limited Free Market, Socialist countries. The ironic truth is that 80% + of the source money will come from of US and other capitalist countries. Hmmm, sounds like a race condition in the code... Before the hacked e-mails were made public (note the huge lack of press coverage) 95% of the nails were put into closing the coffin on this hoax. I have debated many on what forms of government is best (I received a minor in Poly Science in '84) in an analog measurement fashion, and have shown free-market capitalism has done more for the world than has anything else. Beginning way back in Greece, but I see Greece has lost its historical past and is having some fiscal problems. Not a coincidence of course. I never have said in that amount of time that Capitalism is perfect, but its far better than the rest (Churchill, bless his soul). Those who have tried to bring up other forms always fail in their proof. Proof must come with facts via reason and rational thinking. Please get me right, I do see some great ideas from various European countries, but overall the USA has the best design. There is a movement here to follow Europe but businesses will begin to vansh and/or prices and inflation will go through the roof! Retirees on fixed income are just loving inflation about the 4-6% we've had since the mid 1980's when the personal wealth grew at the quickest rate ever. Now that free-market wealth is being sucked away to the Obama government and to overseas. We must have been bad free-market wireworkers !?! In summary, a lack of freedom (read increased government control) means less innovation. The National Instruments company grew from Free-Market Capitalism. We can't change this historical fact. And yes, CRelf, since re-distribution of wealth is the real topic here (not sane science), the types of government and their perspective productive outputs (fiscal nature and level of freedom being the criteria) are within the scope of this topic. Now my wife and I are off to celebrate Christmas with our families. We are expecting the usual conversations, some topics like this one are sure to surface with the most vocal folks being the ... ones who have forgotten history. Funny how that works. We are forced to become more vocal with facts in hand. I am now about pack the gifts in the car along with some books and other info (in case of fire) such as the US Constitution and a handy pocket book of Federalist Papers. Good topic Paul G !!! Back to wireworks and Merry Christmas to All !!! Karl

Lets all take a timeout and take a quick quiz to see what our levels of understanding truly is on this HUGE topic that needs final conclusion : Short Quiz #1 : http://www.geocraft.com/WVFossils/GlobWarmTest/start.html Short Quiz #2 : http://icecap.us/images/uploads/Globalwarmingquiz.pdf Further reading : http://www.oism.org/pproject/ http://rankexploits.com/musings/ http://climateaudit.org/ http://icecap.us/index.php http://iceagenow.com/ http://wattsupwiththat.com/ http://www.friendsofscience.org/ Stay properly informed folks. Please. The most inconvienant truth in the USA is the Constitution. We've had one type of framework since 1776. Look at how many many other countries have had since then. Capitalism and free markets were / are part of that plan. Each country has its own free will of choosing. The real topic here is about redistribution of money and the choices people make to what type of government they have. A Conditionally-led Republic (USA) type of government as designed by the Framers is far better than anything out there (with all due respect to other LAVAer's). This is fact. Marxism, Socialism, Communism just don't work. How many potatoes could you find in a market in the USSR ? One maybe two ? Lets be conservationist but lets not impose wacky, leftist mob rule on common sense people. The tide is turning in the US against the 60's leftist nonsense. Europe needs to return to its senses too (I love your food, but wonder what happen 2-3 hundred years ago with your government paths ...). The leftist in the last 40 + years have wanted to push out God from the culture. Now they want to push out civilty, common sense, and proper debate in science - because it "feels good". When the numbers don't go their way they still can't admit their wrong. Sounds like my 17-year old slamming doors because he can't have the family car keys. He is doing better now (heading for West Point). So, during the timeframe of Moore's Law where computers where doubling in speed every x years, common sense, reason and rational thinking has lessening at the same or greater rate. That is why we are having this debate. The warmest year in the past 100 years was and still is 1934. Ouch, facts do hurt ! Can we erase that ? No. Sorry. Maybe in 1934 they ate too much meat that year, or maybe they were exhaling too much. Hmmm. Sounds like a One Flew Over's convention discussion. In the end, humans want one thing the most and that is Freedom. That would be Liberty over Tyranny. Freedom from voodoo science trying to CONTROL our lives. Now lets return to a higher level of reason and rational thinking ... even while we are coding in LabVIEW. Oh, BTW, I am using NI 8353 (Quad-Core), PXI express (two 6535's), LabVIEW 2009, & Real-Time fo the first time. What fun ! Merry Christmas to All ! -Karl

QUOTE (Tomi Maila @ May 16 2007, 10:28 AM) Tomi, I am starting to build a LabVIEW v8.5.1 interface for HDF5 1.8 and I was wondering if it is possible (with the LabVIEW Vi's available from http://www.me.mtu.edu/researchAreas/isp/lvhdf5/ ) to have 1 writer and several (5 max) readers concerning the same hdf5 file ? From reading the HDF5 literature it seems to be able to support this scenerio but I have noticed some caveats that the LabVIEW Vi's may not support 100% of the functionality. I would greatly appreciate any pointers/lesson's learned when you were building your interface. Thank you, Karl

Mario, How can I get my hands on your wrapper? -Karl