WOC

Jaha, då är VM över. Vilken underbar vecka i Strömstad! Själv var jag där som IT-coach via SOFT tillsammans med Åke Larsson. Alla la verkligen ner ett fantastiskt arbete och allt var en succé! Jättekul verkligen! Själv behövde jag knappt göra någonting då alla andra var så duktiga! Jag högg in där det behövdes, och ironiskt nog för en IT-coach så hjälpte jag till att skriva ut start- och resultatlistor samt häfta upp pappersresultat 🙂 Skriva ut och häfta upp pappersresultat… det har jag inte gjort på 4 år 🙂 Det var ett antal personer som hade rätt kul åt det och jag kunde ju inte annat än instämma i den muntra stämningen 🙂
Under veckan fick jag också tid att hjälpa Åke att få in stöd för hans D-link-modem på ROC:en. Av någon anledning ställde det sig i disablat läge som default… Vad ska man då med ett modem till tänker jag… Men nu funkar det i kommande version 6.4. Det kommer även stöd för millisekunder på önskemål av Sportident i Sverige och England. Dock har SI-enheterna bara stöd för 1/256 sekund, inte 1/1000. Så det är en upplösning på ca 4 millisekunder.
Så är man åter i de vanliga rutinerna och my day time job är igång igen 🙂
Nästa event för mig blir som IT-coach på de 2 SM-helgerna.
//Oskar

11 tankar kring ”WOC

  1. Christoffer

    Vi hade det kul i sekerteriatettält på tävlingarna, det blev nästan som en workshop med allt snack om teknik 7

    Svara
    1. Oskar Berg Inläggsförfattare

      Ja, verkligen! Klart kul!! Och lite utveckling hann man ju med också 🙂 Och jag blev en mästare på att printa startlistor och vika dem 🙂

      Svara
  2. Simon

    Hej! So can you give some details on how the subseconds (that’s what SPORTident calls their 4ms-resolution) will be delivered from ROC to event software? Will there be a new API, or a parameter to activate subseconds, by default off, or what? I’m hoping you will keep compatibility of the existing punch-interface so that existing software doesn’t break… Thanks!
    (PS: and, as 1/256 cannot be mapped exactly to 1/1000 seconds, will you round the values to the nearest 1/1000 or truncate them down to the next lower 1/1000, as would be required by IOF rules for seconds or 1/10 seconds, depending on timing resolution requirement?)

    Svara
    1. Oskar Berg Inläggsförfattare

      Hi Simon!
      I will not break the compatibility, no worries! It’s a new interface to get ms and I’ve updated the API-doc that you can receive when it goes to prod if you like.
      As I have done now I am rounding it. What would you suggest would be the best? I’ve not heard of any rules of ms before actually, so any advice would be appreciated!
      Cheers Oskar

      Svara
  3. Simon

    Hej! There are no IOF-rules for milliseconds, as the IOF only specifies seconds (most comptitions) or 1/10s (only sprint at WOC or something, can’t remember 100%). But for both of those, the times are supposed to be truncated down. So as a logical continuation, one *could* truncate the 1/256 down to 1/1000. On the other hand, the 1/256-steps are large enough so that a mapping from rounding 1/256 to 3 digits after comma (creating 1/1000 with approx 4ms steps) back to 1/256 is unique (this would be needed if you have to re-create a SPORTident message from the ROC-data to send it to software that only understands SI messages). Therefore, you can probably do the rounding and everything will be fine 🙂
    Actually, if you are extending/adding a new API, could I ask you to also add the ”mode” byte from the SPORTident D3-extended message format (from SIACs) and/or after reading the mode-byte from the SI SysMem (for directly connected stations where you can query the SysMem)? For stations that you cannot (or should not) query (for example SRR-stations and BSM-clusters behind Sunjo-boxes) or old-style SI D3-autosend messages that don’t contain the mode-information, you could send the mode-byte as ”unknown”=0x00. It would be good to have that mode-information, as many o-softwares don’t have the codenumber-to-function mapping that OLA and MeOS have… THANKS!

    Svara
    1. Oskar Berg Inläggsförfattare

      Hi Simon! I had to take a look at the rules of IOF and the only thing I could find was this:
      ”23.6 The timekeeping systems shall measure times of competitors in the same class,
      relative to each other, with an accuracy of 0.5 seconds or better (or with an accuracy
      of 0.05 seconds if timing is to tenths of seconds).”
      …it doesn’t say anything about rounding, but I didn’t read the whole PDF I must admit 🙂 Laaaarge PDF 🙂 But as you say… I must be ok to round at ms… And why didn’t SI add another bit to get real ms? 🙂
      If you need the original SI-message there is a parameter in the new interface to get the raw data. If it exists of course… Memory read… Other softwares sending punches to my website can be times when I don’t have the raw data.
      That is a good suggestion. I’m so used to OLA that I get blinded some times, sorry 🙂
      I suppose that Legacy mode will send 0x00 then… the 53-message.
      And when I read from memory it’s no problem to get the function. I already read that out to send to web status page.
      Good suggestion, I’ll try to get it in Simon!
      Cheers Oskar

      Svara
      1. Simon

        Hej Oskar,
        you were soooo close when you quoted 23.6. If only you had looked at 23.5.: ”Times shall be rounded down to whole seconds.” 🙂
        And regarding extra bits, I need to play my usual role as a nit-picker: adding 1 bit would not be enough, you would get 1/512 second. You would need to add 2 bits to achieve 1/1024 second…
        Yes, the 53-message has no mode-byte, and also the ”standard” D3-message from BSM-stations has no mode-byte. Only the ”D3-mod” coming from SIACs has the mode-byte, replacing the MEM2-byte normally indicating the address at which the punch is stored in backup memory. If you need the exact doc about this, drop me a quick mail and I’ll send it to you.
        Cheers,
        Simon

        Svara
        1. Oskar Berg Inläggsförfattare

          Hahaha, ofcourse it is 2 bits 🙂 I blame a have day at work for the mistake 🙂
          I asked IOF and they don’t have any rules for radio times, only for finish times. ROC only sends preliminary times, not official times. But I think I’ll do the same to truncate the time so that preliminary finish time and official finish time is the same.
          For 53- and normal D3-messages I can send the function read from SI-unit I suppose. Does BSF8-SRR send normal or D3-mod message?
          Cheers Oskar

          Svara
          1. Oskar Berg Inläggsförfattare

            Hmm, it seems BSF8-SRR doesn’t send D3-mod in normal punching mode… So when I receive punches from BSF8-SRR in normal punching mode I will not be able to supply function for you… Or do you have any way of doing it?
            ” “Modified” D3 records are sent by SRR Dongle in Air+ mode when receiving back-channel punches from SIAC or any SRR dongle if > 1 punch sent from SIAC (e.g. radio readout). Standard D3 are received from autosend BSM stations if not in legacy mode (command 0x53 received), and through a dongle when receiving from SIAC back-channel not in Air+ or from a direct punch in a BSF8-SRR.”
            Cheers Oskar

    2. Oskar Berg Inläggsförfattare

      It’s only the function, bit 4-0 from mode that you are interested in? Right?
      Cheers Oskar

      Svara
  4. Simon

    Hej!
    Yes, I am only interested in bits 4..0. BS8-SRR sadly do not allow reading SysMem and do not send D3mod. So yes, you can only provide the mode from either SIAC-messages through SRR-dongle or from stations where you can query the SysMem. I’m mainly interested in mode from SIAC-D3mod-punches.
    Cheers,
    Simon

    Svara

Lämna ett svar till Oskar Berg Avbryt svar

E-postadressen publiceras inte. Obligatoriska fält är märkta *

Please answer this question (antispam) * Time limit is exhausted. Please reload the CAPTCHA.