Slashdot Mirror
Slashback: Blender, Paly, Dragon
A cleaner UI would be nice in the next round. Qbertino writes "Blender 2.28 - the first major release after it was GPLd after a $100,000 community source-code buyout in October last year -- is finished. It's now got a wide variety of added features such as Audio Sequencing (as mentioned earlier) and a complete redo of the built-in Python engine for your 3D scripting convenience and import/export empowerment. It runs on Linux, FreeBSD, Irix, Solaris and that other OS :-). See the full changelog here and get the new version binaries here. Cheers to the Blender folks and: Happy Blending!"
Just close the curtain on your way out, citizen. utunga writes "After recent claims that their voting systems were grossly insecure, Diebold has issued a rebuttal which has in turn been panned. One question this raises : Do programmers now have worry that their comments ... such as - 'Reimplemented MMIO functions, as MS is too effing lazy to provide them under CE. Most of this is cribbed from the Wine Project.' - might wind up in the media (or worse, in court) as evidence for one side or the other ?"
Correspondence school? chipace writes "The newly released Dragon-V CPU could have deeper roots in Austin, Texas than in China. The Alchemy Au1500 (AMD) displayed at Comdex 2002 has a lot in common with the new CultureCom Dragon-V cpu (or is it the other way around?). Both have identical MIPS32 cores (16k instruction + 16k data caches), Ethernet MACs, USB 1.1, PCI 2.2, SDRAM controller ... same power consumption. I'm not saying they are pin-compatible... just that this is by no means an original chip (seeing as the Au1500 has been available for over a year). Is the Dragon-V a ground-up development that CultureCom is describing, or is this just another case of a Chinese company doing reverse engineering?"
They can swim out and try, though. Complete Bastard writes "The Australian is reporting today that Aussie corporate Linux users, including AusRegistry, which runs Australia's domain name registry, are also starting to say no to SCO's licensing scheme. After reading the recent /. roundup of corporate ire, it would seem the business world is starting to truly make it's opinions known in this issue..."
The wisdom of the free market. skwang writes "Do you think John Poindexter should keep his job? The head of Pentagon's department responsible for Terrorism Information Awareness (formerly Total Information Awareness) and most recently known for his Policy Analysis Market, which would allow investors to buy future's contracts in middle east events such as the overthrow of King Adbullah of Jordan, has himself a futures contract on Tradesports, as reported by CNN.
Investors can now buy futures contracts to speculate on whether or not Poindexter will keep his job after August 31st. Since Poindexter's contracts are new, they do not represent an accurate indicator of his job security."
Could be too late: Eponymous Coward writes "CNN writes "Retired Adm. John Poindexter, who created a firestorm this week with his plan to create a futures market that would capitalize on predicting terror attacks, will resign in coming weeks from his post at the Pentagon, a senior defense official said Thursday. The official said the research that Poindexter and his Total Information Awareness program (TIA) were conducting had become just too 'unorthodox'." Ya think?"
No good deed goes unpunished. Anonymous Coward writes "In regards to the June 25th Article 'WiFi Exposes Sensitive Student Data': The School district has decided to boot all volunteers, the story is here..."
Seems like a harsh way to treat long-time volunteers with expensive skills.
'Reimplemented MMIO functions, as MS is too effing lazy to provide them under CE. Most of this is cribbed from the Wine Project.'
Wine has been under the LGPL for a while now... anyone know more about this 'cribbing'? What exactly was copied, and from what tree?
I haven't heard that the voting system code is available under the LGPL... in fact, I've heard that secret source code is quite important to keep ahead of competitors.
Poindexter has already lost his job.
--
What is the sound of this sentence?
Maybe it works for the experienced user, but it's still a stellar example of bad UI design. Tiny buttons with cryptic icons, a GUI interface that works in an irritatingly nonstandard fasion, and so forth. Fixing these would go a long way towards making it accessible to new users, and would not hurt the experienced users one bit. Given that the poor interface is by far the biggest complaint people have about Blender, you would think that some thought would be given to fixing it.
How can we continue to believe in a just universe and freedom to eat crackers if we have no ale?
When they were first annoucing that the Chinese would be producing the dragon chip, there was a lot of speculation here and other places that it would just be a reverse engineered version of something else.. guess that's what happened.
This looks like a job for the FSF. How far are the binaries being distributed? Since they contain GPL code, it shouldn't be too difficult to make a case for source code release, which would open the whole app to peer review (and, if the article is even halfway right, hilarity).
Since they are distributed to the state, the source code should be available via the freedom to information act or similar laws.
The real history here is not the foundry, but the architecture. RISC is simpler, not only to run fast but also to design, manufacture and perhaps reverse engineer.
Leandro Guimarães Faria Corcete DUTRA
DA, DBA, SysAdmin, Data Modeller
GNU Project, Debian GNU/Lin
I haven't used blender and have only dabbled with other 3d programs, so maybe this is obvious, but why does a 3d program need audio sequencing capabilities?
Their rebuttal basically says: we have no
software security because we rely on hardware
and procedural security. If a machine is not
meant to be connected to the network and has
only one means of user input which has a limited
number of point and click options then the
underlying software security would indeed seem
irrelevant.
Recently I've learned to write 3ds max plugins, and a good 3/4 of the time I spend is just trying to get around in max. The actual 3d handling code (C++) is really simple to deal with.
Point being, 3d modelling is Hard, and I would expect any reasonably capable 3d modeller to be difficult to learn, even with a well designed UI.
Blender doesn't necessarily have a horrible UI. It's just a targetted one, meant for power users who need fast access to a lot of functionality.
Not to say it couldn't be improved.
-John
Maybe it works for the experienced user, but it's still a stellar example of bad UI design. Tiny buttons with cryptic icons, a GUI interface that works in an irritatingly nonstandard fasion, and so forth. Fixing these would go a long way towards making it accessible to new users, and would not hurt the experienced users one bit. Given that the poor interface is by far the biggest complaint people have about Blender, you would think that some thought would be given to fixing it.
The people who complain about Blenders UI wouldn't manage with any 3d application. If you read any of the numerous tutorials on Blender, you can get the hang of the UI in less than a half hour. I'm not saying be efficient and quick, but at least use it without difficulty.
3d modeling applications with more power than TruSpace can't have that user friendly of an interface because of the sheer number of functions it has to have in quick access. If you look at Maya, they have spent tons of time in the UI, and their biggest contribution was the pie menu. Using Maya for an hour, vs. Blender you will notice Blender has faster keystrokes (while Maya is more "usable") but after 10 hours, Blender is more usable.
Their GUI also works in fairly standard fashion, with menus and hotkeys. The button tray at the bottom (default) is easy to see, and after you know what the icons mean (5 minutes of reading) it makes sense. I'm going to reiterate this point: Most people that complain about Blender and it's interface haven't read any of the documentation on it and spent 30 minutes trying to figure it out.
It isn't a mail client, it's a 3d modelling application.
Dacels Jewelers can't be trusted.
I've downloaded Blender and tried to run it: After 40 seconds, a blank window appeared, but nothing else. After 7 minutes, I was tired of waiting, killed the app and trashed it. This is the same experience I had on other machines half a year ago with Blender. I just wanted to fiddle around with it a bit, as I regularly do with randomly downloaded Mac apps, but this App seems to not adhere to the usual standard of "start the app in 15 seconds or less and everything works".
I think that this market could have been very effectively gamed by those very same self-serving agencies. Want extra money in the budget? Buy "V.P. assasination" futures. Similarly, wealthy terrorists could manipulate our "intel" by buying many, many shares of "everything will be fine". The market is fine for some things, but people have an irrational faith in it. It really boggles the mind (and pisses me off because these people are in charge).
The people who complain about Blenders UI wouldn't manage with any 3d application.
Not so. I was able to pick up Maya, 3DS Max and Lightwave and start editing meshes pretty quickly just by playing with the interface. I have read a couple of Blender tutorials and it still seems like too much work.
Given that the poor interface is by far the biggest complaint people have about Blender, you would think that some thought would be given to fixing it.
Actually, many people praise blender for its user interface. Yes, it is difficult for someone who hasn't read the manual to use, but once you actually read the manual, you'll soon realize that the interface is pretty easy to use. You can get a lot of things done very quickly because of how the mouse and keyboard interact. No, it does not conform to most user interface guidelines, but that's a little irrelevant for a specialized app, don't you think?
"It isn't a mail client, it's a 3d modelling application."
Agreed. Most of these arguments come from the "everything is an appliance"[1] crowd. If it doesn't have a big red "push me" button, then it's too hard. Thing is for those of us whom "time is money" (and make good amounts too), the Blender interface (much like the wordperfect interface) is right up our alley. Remember our tools are for "work", not for "play".
[1] I'll leave it to the audiance who got that trend started.
User input works by the voter putting a smart card into the machine and making selections. A non-cheating voter is supposed to use a smart card provided by the election officials, but a cheating voter can bring her own maliciously programmed card. The security paper described how such a card could be programmed. That is a serious vulnerability. I think they should run a "Black Hat voting" election at DefCon. They would announce in advance that they're going to use Diebold machines to elect the Evil Overlord of the Cracker Universe, with voters encouraged to try to cheat the machines, and the election would be run with the same so-called safeguards as a real election. I bet the results would make Diebold's "rebuttal" look pretty silly.
The makers of every other decent 3D app don't think it's irrelevant.
I last used an early version of blender...so my opinion may be dated...but the documentation sucked as badly as the interface. With most 3D apps, you can just look at it and know what you're seeing. With blender, screw actually manipulating objects, just figuring out what's in front of you can be a pain.
It can be awkward as an employee, but particularly as a volunteer - where you don't have the rights of an employee. The boss is completely non-technical but wants things done exactly his way despite your protests that it wont work, is insecure, doesn't meet their needs etc.
Using the same blind logic, the boss locks the only people that know the system out of the system (change all root passwds, change locks on doors etc) and then 'make do' with a poorer quality system that they pay more for. Normally in these cases with small schools, there simply isn't the budget to employ a sysadmin and deploy nice (read expensive) network topologies and so people volunteer.
A sad case of biting the mouth that feeds..
You failed to address the relevant point: Blender's UI is needlessly cryptic. Even assuming that you really can pick it up in half an hour like you claim (obviously false; I and many others have spent far more than half an hour trying to learn to use it), there is absolutely no reason to simply ignore the years of research and work on user interfaces. A well-designed interface would be just as powerful of the "power user", and still not be as opaque to the new user.
If you think that having a text entry widget that looks like a button is good user interface design, you're insane.
How can we continue to believe in a just universe and freedom to eat crackers if we have no ale?
Bell is now in jail, supposedly for stalking an IRS agent, but the trial was something of a cause celebre for cypherpunks.
Type "Jim Bell" and "Assassination Politics" into Google for more details. Muldrake was the first I know to point out the similarity between Bell's scheme and Poindexter's terror casino.
Notice that there's no mention of the Paly servers in the first article -- the WiFi security hole was in the district offices, and at middle schools, and not at the high school (where the volunteers were administrating computers).
I don't think this project had any merits at all. The people who have been defending it in the press have been using domestic examples as success stories like the Iowa Electronic Market. Well, considering how small a sample size that is and the nature of the two party system AND the more or less predictable workings of the electoral college, prediciting the president is a lot easier than dealing with dozens of governments, rogue organizations, etc.
Also, these traders would mostly be Americans. Pardon my french here, but most traders I know couldn't find Iraq if I fucking pointed it out to them on a map. Not to mention anyone with any insider info could easily be targeted (as in for being killed) for spilling the beans too early.
The truly scary part is the eventually the government would be making decisions on these "trends" that are more or less expressing the lowest common denominator's opinion. That's hardly informative and a recipe for disaster.
Take a look at the Iowa Electronic Market's current state. There is no clear winner this early in the game and it looks like its doing nothing but reflecting the polls.
I call bullshit laisze-faire ideology on this and good riddance to Poindexter. Really now, if the market noosphere truly was that precognitive then Enron and WorldCom would not have been such a surprise.
1) CISC chips are a RISC chip with a built-in ROM and microcode decoder. Makes very little difference when compared to adding useful things like caches. The distinction between CISC and RISC in embedded systems is usually a question of instruction cache usage, code size, and/or memory bandwidth.
2) Intel and AMD have many lines of microcontrollers and simple CPUS. In fact, AMD has a much wider line of these low-end programmable devices than Intel, as that has historically been it's forte. (Ever own a GUS sound card?) So I'd say it was probably because AMD had a wider selection of things to model after.
Or maybe AMD security isn't as tight.
There are lots of potential reasons...
THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
It was not approved on that cycle, but is still one of the approved ideas in the local power elite and will most likely be pushed through eventually.
A very bad idea indeed - but appealing to those who want to instill the Proper Notions of Goodness and Niceness in students. (Think Political Correctness.)
Poindexter's idea should be canned because it won't work.
.)
There are two ways that the market could provide a serious indicator of the future.
One is that Osama bin Laden, or some co-conspirator, will be so greedy he will place the correct bet before the event happens, and that our special algorithms will detect that anomaly in all the trades and react accordingly. This is far fetched enough that we shouldn't be wasting money on it. Can you accurately predict events like Enron and WorldCom by watching insider trading ? No. (There was insider selling, of course. But given all the stock market data and no hindsight, you still can't make accurate *predictions*
The other is that numerous experts who normally just gaurd their jobs by telling the bureaucracy what it wants to hear will suddenly trade on what they actually KNOW instead. Thus, we will find out what they really think. This might work in some sense, but you'd have to force the guys to really gamble, limit the market to cut out noise, etc. This is what some of the contractors proposed and Poindexter or someone else as DARPA shot it down in favor of an open market.
I believe what John Poindexter was actually trying to do was neither of the two semi-plausible methods. He actually thought that a market of people essentially all with no real knowledge, (i.e. no Osama's ), betting on each others fads, buying futures contracts hoping more that everyone else would hope on that band wagon and they could re-sell it rather than cash in the contract itself, would somehow produce a view of the future out of the churn of numbers.
It's as rediculous as believing that the mass of investors all reading WSJ and betting on their own inflated sense of insight will produce a market that accurately predicts stock dividends instead of just being a massive transfer of wealth from the foolish to the inside traders. These guys whorship the idea of that churn of numbers and bids; morons betting on things about which they know nothing are not stupid, they "taking the necessary risks" etc etc.
As rediculous as it is, it is plausible that Poindexter really believed that, given the sort of stock market cheerleader girls he hangs out with.
The futures market idea is interesting, but it only works when a large number of people have enough information to at least have some opinion.
With terrorists attacks, only a select group of people would have valid information, namely terrorists and people connected to (or monitoring) them.
If a system wasn't totally anonymous, then no sane person 'in the loop' would ever make a bet, because they would be immediately arrested after an attack.
If a system WAS anonymous, then you have much larger problems, mainly a terrorism business model. It's quite simple.
1) Plan a terrorist attack
2) Bet your attack will happen
3) Attack
4) Profit
So now not only do we have insane terrorists trying to do shit, but all kinds of criminals. There are a lot more bad people in the world who love money then love Allah, and a lot of smart ones too, just look at some of the stuff the Columbian drug cartels have done so far, like data mining phone records to find out who had been informing on them since.
Also, in order to sell futures, you need to list the price. This means all the information collected by the system is known to terrorists as well as the government, so terrorists can check to see how well they've been keeping secrets.
In general, letting people profit off death and destruction is a bad idea.
autopr0n is like, down and stuff.
"Put another way, actually do it for a living an you'll find out why it is that way."
Trust me; if you do, you'll very soon buy maya, 3dsmax or softimage 3d/xsi.
Blenders UI sucks. It is not a sportscar, or if it is, it's been sitting under a midden in the garden for 10 years. The only possible reason you say Blender has a good UI is because you haven't used it to really make something.
Trust me; there's a reason why you don't see Blender in production environments. And there's also a reason why the Blender developers have put the fixing of their UI as a priority for future versions. Stop defending something even the makers see as broken.
-- Waht? Tehr's a preveiw buottn?