It's probably more trouble than it's worth to sort through and separate vendors and oems from enthusiast sites

You know THIS would be a logical next project...

but what is the "quality" of the data ?
some disclaimers gonna be needed

Certainly not uniform data quality. How difficult is it to produce such graphs on your own?

no idea for the zero head flow rate, other than by arbritary definition of conditions
Cathar probably has a better idea than I
my pump flow bench is in pieces - and low priority for now

I've actually made a start towards it, pulling data off the graphs at http://www.reefs.org/library/pumps/ .

If anyone wants to give me PQ numbers, I'll find a home for 'em.

Some of my code upgrade ideas are smashing to splinters on the rocks of incompatibility. :mad:

Don't forget to make the graph big enoug for the Iwakis :)

I'll look for the .xls that that data came from in a sec

Already planning a 'zoom' button. :)

.xls is okay, but .csv is easier to import into Origin.

Okay, I re-did the image shuffling code to better suit the absence of the Show'em button. It should also be quicker and require fewer contacts with the server.

One bit of weirdness, in IE if you change a bunch of checkboxes very rapidly, the key legend can get corrupted (multithreading fun: it isn't doing the legend updates in the correct order). Let me know how serious of a problem you consider it.

This one.

I dont personally consider it a problem because I use firefox :)

Also, it's easy to import xls into origin; just choose open excel and point it to the proper tab.

LOL, way to show concern for the readers. 90% or better use IE?

It uses excel's dll's to open 'em, and since I don't have excel... They're still useable, but I have to pulled out the data as text and tabs or as a csv. :shrug: No biggie.

Not if PC's audience is anything like Ace's Hardware's.

jeez, what a profile
that is a geek site

An alpha with coordinates. Not so exciting with the water blocks, but it'll be extra cool with PQ curves.

Resolution needs improvement, and it only works in IE at the moment. It looks like I can no longer avoid browser detection and tailored code. Ah the fun of specs that are mostly ignored.

EDIT-it now works with Mozilla.

can you use a pixel counter (with limits) and then convert back to the units ?

Know any good way to count pixels with javascript? I'm deep into the Document Object Model (DOM) searching for good/better ways to do this.

no, I dont write java, just a drinker

there exist (java applets ?) that count pixels, set the reference and as you move the cursor the coordinates are displayed
I used these to layout wep pages
I'll look for an example (later ??)

Java applets are a whole 'nother ball o' wax. I'll peek at an example if you find one, but despite the name similarity, Java != Javascript. I'll delve into Java if I need to, but in the meantime I'm looking for a solution using in-page scripting.

I found the correct object names to get my crude version going in Moz/Firefox. Now to come up with a way to do it better.

nice project, quite impressive =)

Okay, I've smoothly working coordinates in this Mozilla only version.

Is it just my imagination, or are the metric grids/units a wee bit off?

Anyway, if you have Moz/Firefox, give it a whirl and tell me what you think or if you find any weirdness.

In completely unrelated news, didn't you buy a Digitec 5810 or 5820 Groth?

5820, yes. Nice little thermometer.

Do you need any probes? I have a stack of nine new in boc 701s (w/ Certs) and I'll send one or two if you think they'd be useful :)

Does your 5820 have analog out on the back btw?

I bought a lot of a dozen probes, include eight new in box w/certs (no retsyn, though) 701s at the same time I bought the meter. I even offered some to you in a PM. ;) No outputs on mine, pencil and paper recording.

On topic, did that coordinate returning function work for you?

Yay! Now the beastie will output graph coordinates in both IE5.5 and Mozilla 1.7.

The metric flow numbers don't match the metric grid due to a slight inaccuracy in the metric backgrounds. BalefireX? :D

Please please test it with your various browsers so I can maximize compatibility.

Seems to work fine in firefox 0.91

It appears to be a "rounding to pixels" problem - when making the metric graph, I had to determine how many pixels made up a gallon, then convert it to liters, and then split that into 4 equal parts. I assume you did a similar thing to do your coordinate plotting, but we are one pixel different per liter - IE, at 1LPM on my graph, you get 1.01 on the coordinate, and at 7LPM you get 7.07 on the coordinate. Perhaps there is some way to drop it by that .01 per liter on the coordinate side, as taking off one pixel from the graph would mean that the 4 equal parts would no longer be equal.

time for an upgrade to 0.1 pixel resolution

I based the coordinates on the gallons, since they were the original units, and set the scaling and offset to work with them. For the metric, the pixels are still translated to gallons, then multiplied by 3.7854118 to get liters.

It'd be a lot cooler to have the metric measurements match up to the american ones, rather than fudging the metric coordinates. Every other liter moved over by a pixel?

Edit-Hey pH, where can I get one of those one tenth pixel resolution monitors?

I believe I multiplied by 3.78 (imprecise, I know, but it was good enough when working with 72 pixels per inch) you might want to see if that gives a better matching result.