Freesteel Blog » 2006 » February
This morning I received a document chiselled on a clay tablet and delivered by passenger pigeon. Actually, it was a letter from Mary Ann Brozowski, General Counsel of Surfware Inc. and sent by DHL rather than by email or through the regular postal service who’s deliverers actually know where I live — not at the other end of the Cathedral.
It says nothing useful, contains no email address, and reads:
“As you are undoubtedly aware from Surfware’s publications and promotions in the industry, including its web site, Surfaceware’s SURFCAM product includes patented technology for automating the rough machining process, which Surfware calls “Step Reduction Milling” or “SRM”. The SRM process, as well as machines employing the process and articles of manufacture for generating commands or creating three dimensional images in accordance with the process, are all covered by US Patent Number 6704611B1… which is assigned to Surfware.
“A Patent application was also filed with the European Patent Office…”
She then included a copy of Patent Number 6704611, which you can’t read on an ordinary browser.
What could this mean? Firstly, the letters I sent to that company were about the Truemill Patent Application US2005246052 (A1). Secondly, she seems to be informing me about a software patent, since it’s supposed to cover something which makes 3D images, and Europe hasn’t knuckled under this very bad American idea so far. And thirdly, I’m baffled at the number of versions of this trivial idea I’ve seen patented under this name. The webpage I have read this material from is: WO03019454 or EP1419474 (A1). The whole archiving and numbering system for patents is so terminally screwed it must be deliberate.
I don’t know what a General Counsel for Surfware Inc does, but they probably have a lot of time to get up to mischief. I’ve no idea of their plans, if any. But a well-paid and hard working profession who endeavour to make things more obscure the harder they work, and believe that by depositing some smudged pictures about an idea that’s equivalent to sticking those extra steps between the rows in a lecture theatre with the Californian State Government, they can steal someone else’s work which they don’t understand.
Anyways, even if it was sound — which it isn’t — it doesn’t apply. We don’t create a first set of Z-levels and then “determine a second set of Z coordinates for machining a second set of Z level planar slices with the first tool”. That’s the naive first draft method, and it lasted a couple of months in the early versions of the code. What we actually do is find a set of Z-levels at the most optimal spacing and positions (just above flat areas, for example), and then batch them into groups up to the depth of the fluke of the tool. Nothing similar to SRM, although it might look the same to a lawyer with too much time on their hands. You never know what they are thinking. Or what they are reading…
Hi there Ms. Brozowski. If you have got this far with using the web, why not drop a comment onto the blog and start a conversation? Who knows, I may say something that you can take out of context and use against me.
Meanwhile, I had promised myself to read more of Groklaw at some point.
Mary Ann Brozowski is also on the Board of Directors of the Association of Corporate Counsel Southern California Chapter, in whose newsletter I read this stunning advice on page 4:
“Many business are operating in or moving to a so-called “paperless” environment, such that all the information and records concerning a business are stored in some electronic format. If a natural disaster [like a hurricaine] results in the total loss of electronic data, the cost of trying to recreate that information can be quite significant. Thus, businesses who indeed rely primarily on electronic data should look into the possibility of purchasing [insurance] coverage against such a loss.”
Nope… No mention of electronic backups to a warehouse in Timbuktu and many other out of the way places. So-called electronic data is, to a corporate lawyer, less robust than paper. Maybe that’s because it can’t be shredded so easily. You’ll find a good deal of whinging about the “Sarbanes-Oxley Act” in the corporate legal press. No mention that this Arthur Andersen incident was what precipitated it. They don’t want you to remember that. They want the good old days back.
Getting the Online Adaptive Clearing done seems to have captured our imagination. A splash of colour and some careful design so that you do not need to read instructions to work it out has made it look real good. Anyone here who reads this blog is invited to check it out.
It’s still not done. We can’t plot sections of the toolpath or select positions on it yet, and we’re still lacking a post-processor. Also, the job-sheet (the window you get when you hit the “Start Machining” button) needs some serious finishing off. The idea is that the diagrams and controls should be self-explanatory enough to get you most of the way without reading any instructions.
The post-processor is going to let you set the feedrates. Anything else it needs to do? It’ll be on a separate pop-up page like the job-sheet so we won’t have to cram all the settings on one line along the bottom of the main page.
Any other suggestions will be greatfully received.
We’ve promised ourselves that we’ll buy a mini-machine tool (one that works this time) as soon as we see photographic evidence of someone actually cutting a toolpath generated through this webpage. This may take days, or this may take years. This may even take us borrowing a machine tool from someone and doing it ourselves. Who knows?
Spent the second half of last week finishing the Lords on publicwhip with Francis. Lots of little twiddly bits. We even got a scattergraph of the voting behavoir since June working, and it came out convicing. The mathematics is all wrong, but it looks pretty.
More relevant for a useability point of view, I’m going to forget the keyboard shortcuts for previewing the STL file in freesteel and make the interface the same on our web version of adaptive roughing.
Then for the the weekend I had to drink too much wine on Saturday night and then go for an 80 mile cycle ride with Becka down to Llangollen and back without any lunch. We hit an all-you-can-eat buffet in Chester before we caught the train home and didn’t feel good after that either.
I can’t believe the amount of spam about gambling, porn, and drugs that gets posted at the comments of this blog.
I’ve given a random guy who emailed me just now a download link for the Adaptive Clearing demo version. It’s unlikely that our instructions are good enough to make everything clear. So rather than risk losing all his questions and my answers in my spam-infested email box where they won’t be found when it comes to rewriting the instructions, I’m going to ask him to post his comments here.
Thursday, February 9th, 2006 at 4:53 pm - Machining
Lots of behind the scenes working going on with the Adaptive Roughing online. We’ve finally cracked the problem of getting data back from the server without reloading the webpage. (It’s outlined in this page.) This enables us to implement the “View Max” function on an STL file. That and the all important “View Length of Toolpath Calculated So Far”. You should be able to see the fruits of this in a day or so.
Meanwhile, I got to get back to some tedious scallop work. No sooner does it start producing results not entirely full of glitches, it needs to go faster. There’s a good tool called AMD Code Analyst which works really well if your PC has an AMD processor, which mind doesn’t. However, our friends in Copenhagen have got one and have identified four lines of code which takes up 50% of the time.
This is the point where you choose how to optimize things. Either rewrite those few commands to make it run a few percent faster, or find out a way to avoid them being called several trillion times per minute. As they are searching for the horizontal gap between two vertical fibres of varying length in an unordered dynamic list, there’s a good chance it can be done if I put my mind to it. Which I haven’t yet. Probably it’s going to be this weekend’s work since all we’ve reserved for doing is to go out looking for the famous Formby footprints further up the coast.
So, we went south to Bristol and ran a number of errands whilst trundling out crappy Proxxon converted machine tool which we are no longer proud of on trains, busses, and taxis. In particular, we were looking for a replacement in which the parts wouldn’t be bent.
So we visited Peatol, which is in an ordinary house in an ordinary street in Birmingham and saw the imported American Taig Mill that he is selling to people. I can now recognize it in the picture: huge half-inch sided solid metal rods along the axis beds and a total weight of about 40 kilos. He has the coupling kit for CNC, which looks more convincing than either glue or grub screws, but we have to find out more about the CNC conversion ourselves since he doesn’t do electronics.
One interesting tip-off is that most of his customers are model engineers. There’s a big model engineering show in Harrogate from 5 to 7th May which we now want to visit. There’s a whole world of anoraks out there doing stuff who like cheap CNC software (maybe even free stuff calculated through the internet) who may have some useful comments to make. We crave technical feedback more than anything else.
There’s a model engineer exhibition in Liverpool on 29, 30 April this year.
And here is the Guild of Model Wheelwrights dedicated to the “precision scale modelling of horsedrawn vehicles… by means of construction from scratch.” Unfortunately, they don’t include too many pictures because I was hoping to see a photo of a hamster pulling a chariot.
Martin has just now ordered three months of junk mail by book us into Mach 2006 on 15-19 May. This will give us a look-in on the so-called “high-end” market.
It’s all about software for us. For cars, music stereos, shoes, jewelry, and bacon, I can understand the difference between “High-end” and “Low-end” on account of the care, expense, and precision with which the material is manufactured, and corners are cut. But with software the distinction and stratification is gratuitous. There’s good and bad software, and old and new software; just as there are good and bad novels, and old and fashionably new novels. No one is prevented from reading A Canticle for Leibowitz because it’s reserved only for the elite who can pay. No, only SF reading nerds, like dedicated 2.5 inch gauge enthusiasts, get to know of these gems. The rest of us get mesmerized by the gaudy lights and hand over good money for latest and greatest new music, magazine, software, car, kitchen, flavour of crisps, or whatever else, and that’s fine. Just as long as there’s room for people who care to still get at the good stuff. Everything spends more time out of print than in.