If we extrapolate from the directions they were pointing in when they failed
will we find an alien craft?
8318 publicly visible posts • joined 11 Jun 2009
We had trouble with Exchange Server 4.5. I remember running some DB repair program that took 4 or 5 hours to scan the DB half a dozen times to get it into a state that Exchange Server was happy to load.
I was pissing about with some VB for some web app and discovered VB would allow you to dump a whole exchange DB and read everyone's emails and I think I could have used that to rebuild a corrupted DB better than the MS repair program but never have the courage to try it in action. Some nice reading while repairing the DB mind.
While doing the annual sums for the tax avoidance I found a receipt for a 256G card. It took me months to find it. When I had finally stopped looking for it I decided to bore the kids to death with a computing demo and found the bloody thing stuck in the back of the tv while trying to use it as an extra large monitor.
So I could have recorded the golf after all!
There's already tensorflow on the Pi and on a Pi3B+ with a heatsink it can run the 'starter' Mnist train and verify in under 10min IIRC.
I've seen a PiZero doing realtime object recognition on a pretrained model using the RaspberryPi camera.
My only objection to this addition to the AI collection is saying its somehow easier to program in than say tensorflow. I'd bet all they are doing is just pre-setting some parameters that you will want to vary later so its not really any easier than tensorflow.
And if you really want to master AI Weka takes a shit load of beating given you dont have to write a single line of code as the GUI can be used to set just about anything.
They're only giving them away free for the Xilinx stuff - that would be at levels where the accountancy/licensing cost probably exceed the revenue. But if they can help you get started then when you go custom they get a share of that.
I wonder if you can get 8 Arm cores and highways on a Artix7 gate array for less than a production equivalent though?
but we had a GEC 4000 (as used to run Prestel) and that used to fall over if there was a lightning strike within 30 miles or so. In the days before the internet people used to be impressed by our weather forecasting abilities as we'd pop out for an early lunch on a sunny day with a seemingly unnecessary coat.
I worked as a chip designer for a while and towards the end of that MBAs started imposing their 'business practices' to technology. In order to reduce static damage to chips as MOSFETS are particularly succeptable due to static from uman ands going down the lead to the chip pad onto the teeny weeny gate of a transistor the voltage resulting is inverse to the size of pad so bigger pads mean lower voltages so MBA decides minimum pad size for all chips designed here.
Now as our group was making ultra fast (for then) ECL the pad size imposed meant that our chips could not work - simply driving them at the speeds we were going at would take more current than the leads could carry! And ECL doesnt normally suffer from static problems.
It took several meetings and threats of violence before we got an 'exception' so we could make chips that would work.
Either 'I cant work out how life started on a warm wet chemical rich earth so it must have formed in cold barren space' or 'it was somehow created on another warm wet chemical rich planet and was somehow blasted into space in a way that somehow preserved the life from the massive acceleration required to do so to travel across space unaffected by radiation or time and magically land on earth without burning up in the atmosphere or suffering massive deceleration that regularly causes shocks that can shatter quarts let alone the remaining life in the missile!
I think this is another US myth, as is the 2 year one. The 3 weapons detonated during the war exhausted the stocks of fissile material available, The next test wasn't until July 46 by which time Russia would have probably taken over most of Europe by then as, in all honesty, Russia's weapons production had reached such phenomenal levels that there was no stopping them.
A six point plan is one point short of being able to do anything about the problem at hand. I've found that point 7) Before you start get promoted to a level where you are actually allowed to actually see all the data you will be transferring.
Nothing harder than trying to verify shifting data and associated semantics when you cannot verify that certain activities* are outside your purview.
*when I say activities this includes fiddles that people quite high up never expected to be uncovered so an addendum to point 7 would be "I dont want the whole place to close down and all my friends to lose their jobs when most of the board are locked up so a decent reference and I'll be on my way if you dont clean up your act because the next one who tries to sort your IT may not be so generous.
I used one of those along with 20 other engineers and things could get slow - a 3 minute jobbie would take half an hour or more. Until the day I discovered that a program I had would crash and leave me in whatever superuser debugging mode and then I could lift the priority of my batch job to 2 below max (any more and the OS hung) and my job would be done in 3 minutes and the system managers never found out why everything else ground to a halt.
I never did it on bigger jobs because they meant trips to the really good library we had there to further my knowledge of obscure computing ephemera.
Why bother though? I've got £2000 worth of roof tiles that dont get wet because they've got PV mounted over them. Why not just stick PV up instead of tiles - it would be a lot quicker and save a lot of (possibly literal) debugging when shit gets in the tile contacts!
One of the hard things about seismology on a global scale is the wave front of a shock is seriously blurred and diffused by the time(s) it gets to the other side of the planet. Even something as massive as Chicxulub would only be a medium to large earthquake and it would be very unlikely to act as a trigger to a volcanic eruption.
I think Intel cant afford ARM. I voted against the sellout to Softbank as I though ARM would be worth 4 or five times the offer price in a few years. Seeing their post buyout release of info on there NN software I thought I'd underestimated. Seeing their ML stuff I KNOW I underestimated.
I was working with some fast Bipolar in the late 80's and someone had published a design for a 600 gate 16 bit machine and I simulated a version of that at 2.4Ghz on the process I was using. It only had 7 or 8 opcodes* but at the time it would have been the fastest CPU out there by a large margin.
*IIRC but Baby only had 7 and that was 'complete'.
@mage. I am a fan of ARM (I fucking wish I was still a shareholder though) However for nearly 40 years I have always thought that if IBM had chosen the 6809 computing would have been 10 years further on than it was. But would the future, ARM*, have still been around?
*I say the future is ARM - if there ML stuff comes out at a Raspberry Pi level they WILL be the future.
I've often wondered if I should get some form of above cloud transport for events like this. Then I remembered noctilucent clouds.
And as for Eta Carinae being above the clouds would probably mean the GRB would render you blind anyway, either permanently or with a shit load of Cherenkov.