Damn POS 921!

[tm22721]

Also, no matter how much software is tested in a lab (we go through 3 phases of testing (standalone, integrated, independent)) users will find a bug by doing something so unuusal that you can't build a test case for it.

You forgot the most important test phase for a mass market consumer electronics product with complex software - beta test. The more complex the product (especially GUI features), the more beta testers and longer beta test period are required. Dish's product development process is still immature after eight years worth of products.

 

[John Kotches]

tm:

I agree that beta testers can be a valuable tool, but there are some products where you can't run a beta test.

Regardless, there are always going to be latent defects.

Cheers,

 

[SimpleSimon]

tm:

I agree that beta testers can be a valuable tool, but there are some products where you can't run a beta test.

Regardless, there are always going to be latent defects.

Cheers,

Well, yeah, it's hard to beta-test aircraft carriers or WMDs, but NOT consumer products.

 

[rad]

Well, yeah, it's hard to beta-test aircraft carriers or WMDs, but NOT consumer products.

I think that NASA designed, built and launched the Saturn V rocket in less time and with less problems then E* and the 921, and the Saturn V worked correctly the first and every time it was used.

 

[SimpleSimon]

I think that NASA designed, built and launched the Saturn V rocket in less time and with less problems then E* and the 921, and the Saturn V worked correctly the first and every time it was used.

Yeah - they put the necessary money into the project, and had people that actually knew what they were doing.

 

[John Kotches]

Simple Simon says:

Well, yeah, it's hard to beta-test aircraft carriers or WMDs, but NOT consumer products.

Hmmm... I seem to recall the US and other countries running numerous atmospheric and underground nuclear weapons testing.

The beta-test of an aircraft carrier is the shakedown cruise

I will say that Dish needs to have more engineers with 921 boxes at home, and a larger stable of external beta-testers.

That still won't cure the issue of latent defects, but it would improve the quality of the initial release.

Cheers,

 

[angiodan]

My 921 has been perfect with SAT, however OTA is a whole different ball of wax. Basically, I can't watch it. Missed the Stanley Cup finals in HD on ABC, which pissed me off more than anything.

Pissed me off so much, that I went to Game 7. If I couldn't watch it in HD, I watched it 5 rows from the ice.

And yes, the game and experience was fantastic!

A couple of pics for you.

 

[SimpleSimon]

Hmmm... I seem to recall the US and other countries running numerous atmospheric and underground nuclear weapons testing.

The beta-test of an aircraft carrier is the shakedown cruise

I will say that Dish needs to have more engineers with 921 boxes at home, and a larger stable of external beta-testers.

That still won't cure the issue of latent defects, but it would improve the quality of the initial release.

Cheers,

You're right - but I didn't say "impossible", just "hard". Prior poster implied some things can't be beta'd. I picked the toughest ones I could think of, and it still can be done.

 

[tm22721]

tm:there are always going to be latent defects.

At L184 we are still seeing masses of timers either not firing or being mis-labelled. These problems are not 'latent defects' - they are easily reproducible and are being reported by dozens on this and other forums, and we have gone through a number of software upgrades so they are not oversights.

Somebody please explain what is going on with the flagship product.

 

[John Kotches]

tm:

Latent means it passed through testing undetected, only to be discovered in the field.

That says that their testing needs improvement (duh!). See above about rounds of testing.

It would appear based on the results that they aren't applying much in the way of a structured, software engineering approach to their releases but I can't say that conclusively.

Cheers,

 

[Cholly]

You forgot the most important test phase for a mass market consumer electronics product with complex software - beta test. The more complex the product (especially GUI features), the more beta testers and longer beta test period are required. Dish's product development process is still immature after eight years worth of products.

I'd like to agree with you, but beta testers can't even begin to scratch the surface of the pile of bugs that can exist in a complex piece of equipment, especially if it's based on unproven tech. The folks at DirecTV lucked out when they joined forces with TiVo. By the time DirecTVo came into existence, TiVo, Sony and Philips had shaken the worst of bugs out of the TiVo PVR. I have my second TiVo (80 hour series 2), having given my original Philips unit to my son. Yup, rock solid, and it does a nice job in recording SD content from my Dish 811 receiver (once I got the right remote code entered for the IR blasters).
Enter Charlie Ergen and Echostar. They decide to develop a DVR of their own. Oops -- TiVo has patents on PVR concepts. Gotta license them or come up with something different. I'm sure they had a big can of worms to work with in developing the several different recorders that they've brought to market. The real test comes when the unit gets in the hands of the public.
I've done both alpha and beta testing on a number of software projects in a prior life, and still saw lots of bug reports once the software was handed over to the end user.
I don't have a 921 yet (I have one on order). In the meantime, I'm dealing with the foibles of my 811, and wondering if the OTA tuner on the 921 is better than the one on the 811, and if the picture output from the 921 is as dark, etc. etc. Time will tell, I guess.

 

[tm22721]

My 921 was working (more or less) from early January until recently, ie L184 when all hell broke loose, missed timers, re-labelled timers, etc requiring frequent re-boots and timer re-entry.

In its current condition it is virtually useless. After all the reason we have a PVR is to save time and the current software negates that goal.

I exhort E* to roll us back to pre-L184 software.