@TaliesinBeynon oh yes, what Szabolcs said: general format (cross-platform) for exporting stuff. Now, to send data from one machine to the other I use Compress@... but that is very slow on large expressions.

Forces me to do post-processing of data via the terminal interface on remote machines. Does Wolfram Research really want the shiny front end going to waste like this?

of course. the standard iterative procedure involving generators. And god forbid that the actual problem is mentioned in the "question"

nope. copy it, paste it, run it, work out what is unsatisfactory about the result, fix it, report back.

@Szabolcs MX is actually cross-platform as of 10 (the documentation is out of date). The only limitation is that it's not cross architecture (by which I mean, 32-bit vs 64-bit). But does anyone still use 32-bit Mathematica? That's fairly niche... As for the versioning, I believe it to be backward compatible, in that if you write a version in 10 it will be readable in 11, but I don't know that for sure.

@Szabolcs so MX is almost already the solution you're looking for

@Nasser What does it do? I mean if it increases i from 1 to 5 why does it have to check if i is less than 15?

@Pickett here it is math.uiuc.edu/Macaulay2/doc/Macaulay2-1.6/share/doc/Macaulay2/… just strange syntax

@TaliesinBeynon Query seems similar to the Parts you are suggesting

@Nasser It looks complicated because it does so much more than a normal for statement. In another language you would need a lot more to do what they do:
list = array();
for i in v {
	if( p ) {
		list.push(x);
	}
	else {
		break;
	}
	z();
}
print_array(list);

Hi all. An OSX user running v10 is reporting a problem with my code that I can't debug:

A little help, please?

@Nasser If you just want a normal for loop:
for i to 5 list i^2
(* Out: {0, 1, 4, 9, 16, 25} *)

@Pickett I think your code is much easier to understand. Still think M is so much more clear. #^2&/@Range@5 :)

@TaliesinBeynon Count me in as a supporter of having graphical models

and WikipediaData

Range[n] should have been made to start from 0 by default. I think this is a design error

Range[3] should be 0,1,2,3 not 1,2,3

@Mr.Wizard Your 2+2 example works for me at least on OS X.

@Pickett Thanks. Are you running version 10?

Of Mathematica I mean.

@Mr.Wizard Yes, version 10 with the latest version of OS X.

@Pickett Would you post that as a comment in reply to Thomas? It would at least let him know that the problem might be local.

@Mr.Wizard ok

Thanks.

@Nasser Can you give an example where this would be better? Since lists start at index 1, I feel that it is more natural for counters to start at 1 as well?

@Nasser Why do you think that's a design error?

@Pickett @halirutan yes, Range starts form 1 because M uses index 1 as first index. But that is also a design error :). If things were done right from the start, it should be index 0, and Range[n] should start from 0 and all things will fit right. But too late now to fix things. so we go on...

I found another strange if not a bug in DSolve. And internal error but still a result is given
eq = (D[y[x], x] == (Log[-1 + y[x]]*y[x]/((1 - y[x])*Log[x]*x) - Log[-1 + y[x]]/((1 - y[x])*Log[x]*x) - f[x])*(1 - y[x]));
DSolve[eq, y[x], x]

makes no sense to me.

but I am tired now, so will go sleep. later very one...

@Nasser But why exactly is it a design error?

@Nasser I strongly disagree. Discrete things should be counted from 1, it's much more natural. No one says: I have 3 balls, 0, 1, 2 and 3, just in program languages this happens. I love MMA design for this in the way it is :)

@TaliesinBeynon This is great, I'd worked out (or Szabolcs told me) that it is cross-platform by trial and error but worried about it being undocumented and did not use it.

@TaliesinBeynon I guess my point is, this is a big deal for some people (eg those who run code remotely and need to transfer largish amounts of data). It should be documented!

@Murta @halirutan I think it is more natural to start counting from 0 but this is just my opinion ;)

Any way, there are more important issue with M. I just finished a report on Kamke differential equations for fun. Run the 900 ODE's. And M performance is not pretty actually.

I do not know if V10 has gotten slower or what happened.

They have enough on their plate with documented bugs already and they should focus on them. I don't want them to hide their unofficial stuff only because we can't grow up and live with its bugginess.

Is anyone aware how I can search the chat messages for a message of myself where I posted an imgur image?

It seems urls are not searched.

@halirutan I faced a similar problem yesterday. I ended up brute-forcing it by using Import to get the last two weeks worth of pages from the chat transcript and pattern-matching the hyperlinks. I couldn't seem to get the Google "link:" operator working for me.

@WReach Using google and the site option was my second try but it seems it is not possible to get the whole transcript with this.

I hope it is OK if I post several bug questions because there I can easily include images and screencasts and it is visible to everyone what I report to Wolfram.