Altera web site inaccessible

On a sunny day (Tue, 28 Mar 2006 02:33:35 +0200) it happened Ben Twijnstra wrote in :

Possible, I dunno what was the reason, mmm I sure hope politics will not screw up the name system, and countries will not start filtering traffic from specific zones (like China is filtering). But maybe it was just something technical.

Mmm have to look it up, lemme see.. 29.95 Euro / month. I had about 8 or nine ISPs over the years, ALL of these sucked, had downtime, incompetent helpdesks...... So far I have detected 2 glitches in the last one and a half years or so... Just one minute interrups in the middle of the night. Zero errors in ifconfig always. Very good and a lot cheaper then an ISP. Speed is now 512 / 3072, they keep increasing it every now and then, fixed IP. Name server runs here too. So, if you want to do this, and want to have your own domain name, you will have to register a domain (is cheap), and perhaps run some nameservers 24/7.

Now for an interesting experiment, now I have Quartus in Linux wine win98 emulator in one screen, and Webpack 8.1i scripted in an other, on grml-Linux on the same box. So would it be possible to synthesize with Altera to EDIF, then use the EDIF in Webpack to program a Spartan? Have to look into that...... Just to get around some Webpack bug... hum... icarus seems to run out of memory when trying to make EDIF..... 600MB should be enough? LOL hehe Maybe it will work...

The problem has been rectified as of 4:00 gmt yesterday. We appologize for any inconvenience this has caused.

Tim Colleran Altera Corp.

On a sunny day (Wed, 29 Mar 2006 01:06:17 +0200) it happened Ben Twijnstra wrote in :

Well, I think not, but somebody will likely correct me. What happened now was this: I got stuck in the weekend when my video filter would not compile on webpack-8.1i. It gave some error case, and that indicated the 'bug' would be fixed in 8.2. So.... And I believed what it did say..... So I took my project, put it into Quartus, and it told me about a silly typo I made. I fixed that, then it compiled without errors in the Altera soft.... So today I took the corrected version back to webpack, and have now the best video filter I had so far, just sat there looking at the scope... cool. The lesson to learn here is one for Xilinx I think: For Gods sake give a sane error message. They completely send you on the wrong track. I know why, the X soft will synthesise, and then get stuck and tell you how it got stuck. The Altera soft did some simple checking before it started and told me about my error.

Anybody remember the old ZX81 / Sinclair Spectrum? That BASIC was soooo popular because it did error checking on each line entered. Hey even iverilog did not catch this.... It can save the user many many hours.....

So, even if the X soft was better then the A soft (I dunno, I have a different feeling), the program with the better error checking (and sane messages) will save you +++++hours and so money too.

Of course the advantage of the capitalist world is competition.... So we use the best of both. Next time in a similar case I will run on Quartus too to see what happens.

But X should fix that really.....

Groeten Jan

sounds like a good idea.

Of course, and to 'help them along', we can meanwhile promote using Quartus as a 'second opinion', should a designer ever hit an unhelpful WebPack error message.

I have changed the heading to help this :)

-jg

It won be optimal, but I happen to know one Philips lab that did something similar - but they used the eqn file to convert to xnf. Does ISE still use XNF?

Best regards,

Ben

What was the bug? Can you create a test case so that iverilog could be enhanced to print a useful message?