Pass this on to the developer of Oceanlog...bug downloading VT 4.0 memory past 512Kb

[AndyDragon]

I have a shiny new VT 4.0 and have been loving it...imagine my (non) surprise when Oceanlog failed to download properly once the profile memory passed the 512Kb mark in the computer's memory. The profile is there, but Oceanlog fails to download properly. And I know why.

When the downloader in Oceanlog reads the internal log book entry for the profile data, it's not using the proper byte masks for the addresses.

Here's the list of profile addresses (and the original bytes) for my profiles. I use 2 sec profiling, so likely I'm the first one to hit this (or to notice it):

Dive 1: 0x00a50 ('A5 60' @ memory location 0x0424) to 0x03550 ('55 23' @ memory location 0x0426)
Dive 2: 0x03560 ('56 63' @ memory location 0x042C) to 0x05A20 ('A2 25' @ memory location 0x042E)
Dive 3: 0x05A30 ('A3 65' @ memory location 0x0434) to 0x089F0 ('9F 28' @ memory location 0x0436)
Dive 4: 0x08A00 ('A0 68' @ memory location 0x043C) to 0x0A330 ('33 2A' @ memory location 0x043E)
Dive 5: 0x0A340 ('34 6A' @ memory location 0x0444) to 0x0D710 ('71 2D' @ memory location 0x0446)
Dive 6: 0x0D720 ('72 6D' @ memory location 0x044C) to 0x10E90 ('E9 30' @ memory location 0x044E)
Dive 7: 0x10EA0 ('EA 70' @ memory location 0x0454) to 0x148A0 ('8A 34' @ memory location 0x0456)
Dive 8: 0x14830 ('83 74' @ memory location 0x045C) to 0x16E90 ('E9 36' @ memory location 0x045E)

The downloader worked for the first 6 dives, but after that, the profiles (and exit time and a few other things) at completely wrong. It seems that instead of using 1F as a mask for the second byte of the address values, the developer is using 0F. This means the address is being truncated to 512Kb when the downloader reads memory instead of 1024Kb. This leads to corruptions in the downloaded data.

Please correct this, I didn't pay for a $1000 computer which cannot be downloaded properly...

 

[Insta-Gator]

Andy, I hate Oceanlog. Sounds like you know what you're doing. Write a better version, we will come ($).
Need a way to selectively format printed log page (A5) output.

 

[AndyDragon]

There IS an alternative (actually a couple). Diving Log uses the libdivecomputer work by Jef (TnT) in it's 'Universal Downloader'. The VT 4.0 download is not quite ready (at least I get a protocol error right now). Nick (don't know if he has an account here) has also worked on this for his MacDive downloader. I worked with Jef in the past and I gave him some info on the VT 4.0, but I just don't have the time to spend on this (I work full time for a major software company and they keep me too busy!)...

 

[Jim Lapenta]

At least you have an idea of what was wrong. Many of us with Oceanic computers - which are great computers by the way- found out after many tries that the best usage for the download cable was as a cat toy, a way to make a few extra bucks by reselling it to some other poor schmuck, or as a good thing to tie up unruly kids, and that the software disc makes a neato coaster on the desk. Other than that not much more could be done with Oceanlog. I started trying in 2004 and it's still never gotten fixed. Even Microsoft isn't that bad.

 

[AndyDragon]

Well that is the ultimate problem...I had a sneaking suspicion because I've played with the memory dumps from many of the Pelagic dive computers, but for most they would assume the computer is defective, not the software. This would cause a huge logistical and monetary snafu for them. I love my Oceanic computers...all 5 of them (not all mine, everyone in my household that dives, dives with an Oceanic on their wrist) and I'd hate to see the computer suffer because of a easily correctable software bug. I hope they actually get this (I posted to Doug on Facebook as well) and avoid costly confusion and erosion of the confidence in their hardware. Their computers rock, their software stinks. I just want to see the bug corrected (for now), but ultimately would love to see the protocols and memory layouts published - even if it's not "public", but in the form of a NDA-type agreement. I can understand the reluctance to make this open source...opening themselves to liability claims is a deadly game, but if they published ONLY the information that can be read (and only the read protocols) so that only information already recorded during a dive is available. This then leads to the problem of firmware updates when in third party applications. It's not as simple as just making is open source. The original source is VisualBasic (and for developers here, you'll know that means it's likely a damn mess) which makes open source even more difficult. In the end, the best course would be for Oceanic to license a third party application (one that would be willing) and rebrand it and add their specific hardware access (and remove others of course). Or contract someone to write a better solution for them. Either way, it's obvious (to me) that the current course is one that erodes peoples' confidence in what is awesome hardware. Look through the posts here and it's pretty obvious that Oceanlog is the ugly stepchild...

 

[AndyDragon]

For those with a VT 4.0 facing download problems, the beta download (2.3.5 beta) from the Oceanic site fixes the download issue. I was able to download all the dives from my VT 4.0. Damn this computer kicks butt, love it!