« first day (315 days earlier)      last day (4170 days later) » 

12:00 AM
@Szabolcs I had been wondering myself. She has logged in, though, so probably just too busy to post at the moment.
 
@Szabolcs She took a new job sometime this summer and that's about when she stopped posting. I'm guessing work's keeping her busy, although she visits almost everyday
@OleksandrR. do you have a quick link to the page?
btw, is an update to the WB expected anytime soon?
 
@Brett are legends still done using Inset? I guess yes as that's a natural choice. But then is there any improvement to help deal with misaligned Insets when resizing graphics? Or generally, any improvement in handling conversion between absolute, scaled and graphics coordinates?
@rm yes she was saying that she was leaving academia
I'm also curious if the new front end supports HiDPI mode on OS X..
 
@Szabolcs It better...
 
12:18 AM
@OleksandrR. This is fairly accurate, actually.
 
@BrettChampion :D
 
12:30 AM
@Oleksandr the symbolic tensor stuff must have been in development for quite a while. I remember I saw it mentioned, maybe on MathGroup, before v8 came out, but then was a bit disappointed that it didn't appear in 8
 
@Szabolcs I can imagine... it looks very complete and general (although TensorReduce could possibly have been called TensorRefine or TensorSimplify). For this and other headline features such as theorem prover or FEM ("PDE special functions") I can imagine Wolfram doesn't want to let it out the door while only half baked.
 
Is there any FEM stuff in this release? I was looking for it but didn't find it
 
Maple has had the symbolic matrices and vectors for a while, and although I don't use Maple or know anyone who does, from the documentation it looks like Mathematica is now way ahead in that respect.
@Szabolcs doesn't look like it (at least not documented). Maybe ask Oliver Ruebenkoenig... if he's not sworn to secrecy...
 
@OleksandrR. I used to use Maple regularly in 2009/2010. One reason why i remain disappointed that Mma is still behind in units and plotting (ease of use)
 
12:53 AM
I can't understand what is the reason for producing the NDSolveValue and ParametricNDSolveValue functions... they save a measly ReplaceAll, big deal...
And they have no equivalents for the other *Solve functions
 
@OleksandrR. what's this about a "theorem prover"?
 
@OleksandrR. This is definitely written for beginners.
Look at the examples:
 
Mathematica 9 documentation on line!.. Cool! So I can talk about it in the forum! I was in the conference this year. The text highlight and auto complete are the best part.
 
@OleksandrR. With a list of interpol. functions no one would do that:
Plot[Evaluate[Table[sol[[i]][x], {i, 5}]], {x, 0, 10}]
 
@Murta yup
 
12:56 AM
when we have Through
 
@AndrewMacFie Yes, very nice. The date functions are nice too, there is a lot of new things.
 
@OleksandrR. Hmm, on the other hand they used the same approach in the NDSolve doc..
 
17
Q: What does "EquationalLogic`FindCounterexample[]" do?

belisariusI found the undocumented EquationalLogic`FindCounterexample[] by browsing the lists of available symbols, but I have no clue about its purpose. The name is intriguing! So far I found the following: Accepts 2 or 3 arguments Returns an Integer (1, 10, 15, 31, 100 ...) For the first two argumen...

It is not certain that this is under development, though... it might never advance beyond where it is now, which is sufficient for the kind of things added to FullSimplify in version 6.
 
@OleksandrR. nice find!
 
@halirutan the *SolveValue functions' documentation are total copypasta.
 
1:02 AM
@OleksandrR. Yes, I see this now too..
 
Any way of finding all the new turorials?
 
1:25 AM
Did you see the ImageFile* functions? They seem to be able process large images on-disk even if they don't really fit in memory
This (having to read everything into memory and process in one go) has been a significant limitation for me in the past. But not with images.
 
@Szabolcs yes, when I read "out-of-core" I had hoped for something like memory-mapped file support. Instead it is just these image functions limited to a few filetypes.
 
Another interesting looking find: asynchronous tasks. reference.wolfram.com/mathematica/tutorial/…
It's it entirely clear though if this does anything more than download asynchronously. I'm browsing an an ipad in a bar, sometimes its a little difficult to read. My wife is playing piano here sometimes.
If the functionality is more general, I wonder how its implemented and if it can be a replacement for multithreading (dynamic isn't)
There's now a way to define custom input and output stream methods reference.wolfram.com/mathematica/tutorial/StreamMethods.html
 
@Szabolcs seems it is not general... only certain predefined asynchronous tasks can be used (and in M9 you can download a file from a URL and that's it). Probably these are just threads but someone has to implement a thread-safe kernel function to do anything with it.
@Szabolcs nice!
 
1:50 AM
Now there is a Matlab inspired ArrayReshape. This used to be available as an undocumented, non-crash-safe function under a different name
 
Glad to have that... Internal`Deflatten was a bit of a liability
 
The TimeSeries package is now embed into M9. With ARIMA, SARIMA, ARMA and etc. It was the TimeSeries package in the past.
 
2:10 AM
No ISOweek number suport!...
 
acl
I admit that after v8 I was somewhat underwhelmed by this release. There's nothing wrong with it, but librarylink and compilation to C (and documented bytecode!) allowed me to almost ditch C completely
well, legends aren't bad of course
and I am not sure what I'd have asked for if I got to decide what is to be implemented, so... faster numerics? hard to see how that would happen though.
 
@acl I also think that to an extent we managed to hype it up for ourselves and built up unrealistic expectations. This is a solid release just like version 7 was, but it isn't as ambitious as version 5, 6, or 8.
 
acl
Well honestly I am not that sure what I'd expect that would have the impact compilation had (for me I mean). FEM maybe? I don't know. Even if I try to come up with a list of things I'd like to have, I don't know what it would contain! At least if I'm realistic about it
 
Version 7 solved a lot of small problems in version 6 that overall made me question whether that ambition had been worthwhile at the time. What we don't see yet with M9 is how many M8 bugs and oddities have been fixed... it might be a case of improvement and consolidation as much as adding new features.
I don't have M6 installed any more and haven't for a long time. If M9 is to M8 as M7 was to M6, I think it's going to be another straight switch for me.
@acl message passing between parallel kernels would be a nice addition. Definitely achievable yet also quite a major undertaking.
 
acl
@OleksandrR. yes that would probably make a lot of difference for my work
@OleksandrR. bugs? what bugs?
:)
 
2:26 AM
ICA transforms would also be nice to have
as would fitting functions that don't puke on complex data/models and Savitzky-Golay filtering... but I already wrote packages for those; it's not hard
 
acl
@OleksandrR. I don't know much about that. I mainly use mathematica for things like constructing and diagonalizing large sparse matrices, then operating with the eigenvectors. so, the closer I get to C performance for that, the better
the complicated stuff and interfaces are nice to play with but not what I do most of my working day. I am guessing I am a minority though
@OleksandrR. as an aside, since you're handing in the thesis soon, how come you're not sweating over that and are instead chatting about the new version of mathematica?
or have you handed it in and have your viva on fri?
 
2:52 AM
@acl well you're a proper physicist. I'm only a pretend one.
 
@acl: I would love to see compiled and vectorized statistical distributions. The symbolic stuff is good, but basic performance requires compilation and listability.
 
@acl I am working on it... just writing my conclusions at the moment. I've already resigned myself to the fact that whatever I don't finish in time I will just cut out. I already messed up by wildly overestimating how much data I could feasibly fit into one thesis... if I included all of it I'd have had a 1000 page thesis and taken 2 years to write it.
 
acl
@asim I don't use those much so I also don't know much about them. But increasing performance overall is my main wish (especially since parallelizing is limited by license availability--we only have 50 here, so the 3-4 of us that use mathematica on the cluster keep interfering with each other)
@OleksandrR. hardly an original predicament... During my viva I was asked "so, how come you didn't expand and publish this? ran out of time, did you?"
I never did expand and publish "that", and have since switched to other things. time is more of a limiting factor than we think!
 
@acl the first 3 years of my PhD were a total write off. Everything I did, I did in one year.
 
acl
@OleksandrR. it's experimental right?
 
2:58 AM
yeah
 
acl
I have heard similar stories
well, mine was overall two years, so not much time for writeoffs. But I did waste significant amounts of time doing stuff that led nowhere
it did teach me a lesson though
 
I tried some calculations too but they were not practical at the scale I needed them to be.
 
acl
namely, "no, all those people that failed to do X are not idiots; it's just too hard"
 
Similar in experimental. Two groups have done experiments I also did; they published whereas I found that the results were a mistake
Seems to be a perennial problem that nobody necessasily knows if the next guy's instrument worked properly or not unless the system studied is very well known and the results are clearly erroneous
 
acl
@OleksandrR. Well during a phd I suppose we're allowed to get it wrong a few times. Not so much afterwards though
 
3:06 AM
Well I suppose one thing I've learned from all this is to recognize the signs of failure early :P
 
acl
so, are you staying on at your mysterious Russell group institution for a while?
(see, I'm proud of knowing what the russell group is, as I learnt of its existence after I left imperial)
 
I am, yes... my supervisor offered me a short PDRA appointment. Just enough to write up my PhD work and teach the next student what I was doing.
 
acl
@OleksandrR. so you are already worrying about what to do next, I guess
sounds fun
 
Yeah but at least I have some time to find a job that suits me now. Problem is I'm not really sure what field I want to go into, because I've enjoyed most things I've done. Will have to get some better ideas about that within the next couple of months.
 
acl
@OleksandrR. you'll probably not think about it until 1 week before you absolutely need to, if experience is anything to go by
finance would be a lucrative direction I guess
 
3:16 AM
@acl ugh. Boring. (At least IMO.) Also I'm not very motivated by money. I want to do some postdocs but not sure about the details yet...
 
acl
@OleksandrR. well, if you're planning to do something more theoretical, now's your chance. later it'll be hard if not impossible to switch.
 
@acl I suppose so. I don't know; I know a few experimentalists who are also very good theoreticians (not so much vice versa though). But then what I have in mind by "theory" and what you do are probably not so similar.
 
acl
@OleksandrR. right, our definitions likely differ, but what I am trying to say is that if you do wish to spend more time on theory, now is the time to make that move. after that it'll be more difficult. however, if you were to ask for my advice, I'd say not to do it (move to theory).
 
Why do you say that? My concern would be that there are too many people much more intelligent and creative than I am and it is difficult to make a serious impact.
 
acl
@OleksandrR. that is one problem (people tend to be very smart and motivated).
there are others that I am afraid I won't get into (would take too much effort to articulate)
 
3:31 AM
How about difficulty in securing funding to do what you want in a changing economic climate and without necessarily any obvious application at the outset?
 
acl
and anyway a serious impact is hard to make in either experiment or theory. in experiment you also need more resources (financial and otherwise)
 
yes, agreed... although I just meant enough of an impact that people are willing to give you money on a regular basis :P
 
@OleksandrR. how close are you to being done?
 
acl
@OleksandrR. that's a different issue. funding, at least in my area, tends to come from basic research institutions (eg epsrc) so that does go down, but not as much as you'd think. I don't know how it goes in experiments. but there are other factors, eg, the REF (research excellence framework) causes positions to open a year or so before deadlines
 
@rcollyer a couple of days away :P
 
3:36 AM
@OleksandrR. wow, me too. I just have to finish writing the damn thing, and a calculation or four ... :P
@OleksandrR. Best of luck.
 
@rcollyer ah, I thought you were being facetious at first. Thanks, and good luck to you too!
@acl yes, EPSRC here too. The money perhaps is not going down but the competition is getting increasingly fierce.
 
Evening, @BrettChampion.
 
@rcollyer Evening!
 
@BrettChampion just got a reply to the job ad. 11 days waiting and all i got was a please send us your details -- no information about the actual job or the company or the location etc. etc.
 
@Brett I just saw that M9 documentation is online. Since when? Is there a scheduled release date yet?
 
3:42 AM
@MikeHoneychurch Double-blind job interviews, maybe?
 
@BrettChampion no smiley so are you serious? More like someones good idea without considering the logistics -- who will administer it, how much time do they have to respond, etc. etc.
 
Night all. The thesis calls.
 
@magma Since sometime last night, apparently. Release dates are one of the areas of my NDA that I'm most careful about. :-)
@MikeHoneychurch Mainly just forgot the smiley. :-) I think it's more likely it's a combination of still working out the details of how to run the program, the Thanksgiving holiday here in the US, and the forthcoming Mathematica release. It seems to me that there are more positions now than the last time I looked, which I take as a positive sign.
 
@rm-rf Ah! A question where those gauges will come in useful!
0
Q: How to manipulate gauge theory in Mathematica?

user6818I want to know if there is a way of typing into Mathematica an expression like the following, $\epsilon^{\mu \nu \lambda} f^{abc} A^a_\mu A^b_\nu A^c_\lambda + g\epsilon^{\mu \nu \lambda} A^a_\mu \partial_\lambda A^a_\nu + \bar{\psi}(\gamma^\mu(\partial_\mu + gA^a_\mu T^a))\psi$ $ (..repeate...

 
Thanks @BrettChampion. I did not really hope to get a specific date.
 
3:57 AM
(And to be clear, I have no idea how the job board is being administered at all.)
 
@rm-rf Joking, of course... although this seems like a perfect fit for the new symbolic tensors in M9. I'm not familiar with this area of physics but someone who is can probably put together a nice answer.
 
@OleksandrR. Indeed! :-)
 
@BrettChampion as you know, I know how some people there are quite stretched with their workloads. It is one of the reasons the Wolfram Library was having so many delays. I guess my point is that the resources need to be made available rather than just add another task to someone who already has many many things to oversee.
...and bypassing the Wolfram intermediary step this increased workload for some employee could be eliminated
 
@BrettChampion posting documentation before releasing a new version was done before? Or is it a new idea?
@MikeHoneychurch are you applying for a job at WRI?
 
@magma no
 
4:05 AM
@magma I think it's been done before, especially since there are a lot of docs pages, and you want them to be up before you update the front page that points to them (after a few clicks.)
 
@BrettChampion I see. I was caught by surprise, like many others here. A pleasant surprise, I may add. Well guess M9 will be out soon
 
4:28 AM
@Szabolcs How's the new position?
 
I like it a lot, the topic is to my taste. But still getting used to the new country :)
 
What's the topic?
 
Just dropped in for a minute. I think I'm just too excited waiting for M9. Need to sleep soon
It's networks
 
@Szabolcs where are you now?
 
More specifically sampling random graphs with constraints
@magma in the U.S., Indiana
 
4:33 AM
@Szabolcs a big jump from here :-)
 
Time for me to go. See you all later!
 
@Szabolcs your uploader works also with other SE sites?
@Szabolcs what happened to the new wri forum that we could preview some weeks ago? It is open to everybody now?
 
 
4 hours later…
9:04 AM
@P.Fonseca There are now three methods specifying the type of user interaction with 3D plots
 
@magma Yes, it works because it does nothing more than to upload an image to a server and let you paste the url of it. The question is, is the server of imgur dedicated to SE or to Mathematica.SE only..
@magma Here you have your answer
Jeff Atwood on August 17, 2010

Thanks to our good friends at imgur.com, we now support native image uploads on all Stack Exchange network websites.

Yes, that means adding a picture to a question or answer is now as easy as …

clicking the Insert Image toolbar button selecting an image from your computer, or the web clicking Upload

selecting an image from your computer, or the web clicking Upload

clicking Upload

… there is no step 4, you’re done!

These images are kindly hosted by imgur.com.

Alan Schaaf, the man behind Imgur, generously provided us a network-wide “pro” account that keeps any images hosted through our websites around indefinitely. …

@magma It's StackExchange wide. Therefore, you can upload images for other SE sites too without being afraid of wasting Mathematica.SE resources
 
9:26 AM
@halirutan thank you @halirutan
 
 
2 hours later…
10:58 AM
on FEM in V 9: If some has V9, may be they can try ?"NDSolve`FEM`*" and see if there is actual function definitions show up when clicking on the topics that gets displayed. In version 8, these symbols that gets printed seem to be all place holders at this time.
 
11:44 AM
@SjoerdC.deVries I'll wait for my download, to try it out.
 
@P.Fonseca It's here in the documentation.
 
@SjoerdC.deVries OK. The play animation is not very explicit (since they don't rotate it...). So I don't know what is the difference between Globe, ArcBall and TrackBall. I'll have to try it out (and since I'm not one of WR oldest/biggest clients, I'll probably only receive a link within a few days).
 
12:31 PM
@NasserM.Abbasi at least NDSolve`FEM`AssembledMatrixData, the first I tried, seems to be read proteced and readable once you remove that attribute
 
 
2 hours later…
2:04 PM
Can we customize the predictive interface? That is, add or remove "power toys" from the list, triggered by the patterned that is in the current cell? It can be interested for adding our own theme of work, with our own “power toys”.
 
2:16 PM
Does anyone else get this window flashing when Mathematica starts up?
 
2:32 PM
@Szabolcs No., I just started Mathematica on windows. Never seen the above message before.
 
acl
2:46 PM
5
Q: Any way to prevent Mac firewall from asking to allow connections all the time?

Tim KempWhenever I start Mathematica 8 on my Mac the built-in Mac OS firewall asks if I want to allow JLink to accept incoming connections. Then, during running, I'll get the same question about Mathematica.app. Sometimes the requests occur several times during the day, though not always. It's a minor ir...

and
4
Q: Why does OS X need permission to allow JLink connections each time it starts?

raxacoricofallapatoriusI understand that JLink loads each time Mathematica launches or each time a kernel starts, and that this can require permission from the firewall, if it is enabled but OS X provides settings the provide this permission by default, so that I don't have to provide it each time Mathematica starts...

I don't know about mathkernel. you can just switch off the application firewall for instance
(there are also two other firewalls on os x, ipfw and pf)
 
Hi all!
@acl Did you manage to run my installer?
 
3:05 PM
@LeonidShifrin I used it on a mac and it works well out of the box
I even installed it using it
 
@rm-rf Thanks for confirming, that sounds good. Re: installed it using it - this is exactly what I meant for it to be - the basic bootstrap installer. Any specific project format would impose more limitations on what can be installed. I plan to soon produce a more specific installer that would be more directly useful for code exchange, based on this basic one.
Hi @Andy!
 
Hi @LeonidShifrin
 
@AndyRoss How is everything?
 
@LeonidShifrin not bad, just checking the boards to see what people are saying about 9.
 
@AndyRoss Isn't is a little early for that yet?
 
3:10 PM
@LeonidShifrin well the docs have been up for a bit anyway :)
 
@AndyRoss Yep, that's right.
@AndyRoss What is your take on the most significant addition in V9 (from your perspective)?
 
@LeonidShifrin for my part I'm obviously excited about the new stats functionality (including RLink). I particularly enjoy the autocompletion. I feel crippled without it when I switch to an M8 kernel.
 
The auto complete feature (aka 'Code Assistant') is very nice
 
@ArnoudBuzing thanks, I never knew what I should call it :)
 
@AndyRoss Yes, autocompletion is a great thing. I remember switching from V5.2 to V6 with the code highlighting and I could barely use V5 after just a few days with V6. I guess this will be a similar effect.
 
3:15 PM
I think the official name is 'Context-Sensitive Input Assistant' ... wolfram.com/mathematica/new-in-9
 
Hi @ArnoudBuzing!
 
@ArnoudBuzing Yes it is. I also like the bracket highlighting.
Morning @Verbeia
Time to go.
 
@AndyRoss I have just been looking at the docs. The time series features look interesting. I am trying to work out the applications for the new TemporalData object.
And hello everyone... Quite a party in here.
 
Hi @Verbeia
 
@Verbeia I think that TemporalData will become more and more useful as time goes on. It is particularly useful when you have multiple paths (or time series) and you want to slice them in various ways.
 
3:22 PM
@AndyRoss will it be possible to do date-aligned arithmetic with this, so you can add, say different components of GDP together to get a subtotal, even if the data don't have exactly the same dates, dropping the ones that don't have matching dates (not sure if I'm making myself clear)?
 
@SjoerdC.deVries @SjoerdC.deVries: the trial link is now up.
 
If it does, you just got access to much of Excel's market.
 
@ArnoudBuzing Thanks for the heads-up. Will be downloading in 5 min!
 
@Verbeia perfectly clear. Lets just say that this is possible but hasn't been made super easy yet. To get an idea what sorts of things are possible already I suggest you look at the Options section for TemporalData docs.
 
@Rojo yes, but you forgot to specify “straight and available”
 
3:27 PM
I have premium service, still in effect. Do I qualify for free upgrade? Or do I need to buy version 9? I am student. I logged into my portal account, but do not see V 9 there. But I see one can actually buy it now for students also.
 
@ArnoudBuzing Clicking on "Featured Examples" in the documentation takes me to the web, not its equivalent page in the docs. I guess the URL was hardcoded there...
 
@ArnoudBuzing downloading now, but it seems to be stuck
Oh, going on again. Apparently I paused it. 4 minutes to go, jay!
1.1GB, It's getting bigger every time
 
@rm-rf: this is intentional, the file size of those featured examples was enormous.
 
Oh, ok. Would it be possible to download some of them for local use (like downloading stuff from *Data functions) without doing so individually from the webpage?
 
@rm-rf: I will ask if they can make this available as an additional zip file.
 
3:35 PM
Thanks :)
 
Installing
Halfway...
@ArnoudBuzing If I get a full license for v9 (working on that), do I have to uninstall the trial, or can I enter the new activation key?
 
@SjoerdC.deVries: It's the same download file
 
I keep getting a The front end encountered an error while processing a "NotebookPredictions" packet. message
 
@ArnoudBuzing Ok, so I don't have to download it again, but will I need to reinstall?
Restarting the PC
 
@SjoerdC.deVries - no need to reinstall, just a different license/activation key.
 
3:47 PM
iPadding now: vol verwachting klopt ons hart...
 
@rm-rf @rm-rf: I am checking on that. Can you send email to support@wolfram.com and possibly cc:arnoudb@wolfram.com?
 
@ArnoudBuzing Sure
Also, I now can't use my version8 after activating version 9. Mathematica just quits unexpectedly
 
Guess I'll wait with my activation for a bit...
 
@Ajasja Yeah, don't know what's up with that... will send that to support@ too. I had v7 and v8 running side by side, but then both had the same license and activation key
 
acl
@LeonidShifrin yes, works fine
 
3:58 PM
@rm-rf this seems odd. I've never had any trouble running them side by side.
 
@rm-rf there must be something special in your configuration. If you can contact me at arnoudb@wolfram.com we can take a look.
 
@acl Cool. thanks for confirming. I will likely (although not yet fully decided) drop the Java dependence soon, since there are functions like ExtractArchive available in M (I missed that one), but it's good to know that it works.
 
@rm-rf @ArnoudBuzing I have v8 and v9 now running side by side. So far no errors, but my v8 now bears the v9 trial banner.
@ArnoudBuzing It seems the installation has overwritten my v8 product key
 
Sounds like I should move ~/.Mathematica out of the way before I install
 
4:02 PM
Oh, an unrelated question: How come the logo is identical? (not that I don't like it:).
Also like the redesigned webpage.
 
@Arnoud At the moment, is it only the trial that is downloadable?
 
OK, Reactivating the original V8 license worked.
 
@Ajasja it's not really identical... the v9 logo has a denser "texture"
 
@Szabolcs Mathematica 9 is available from the store. Trial downloads are available as well. Upgrade emails have not been sent yet.
 
@rm-rf Sorry, hard to see here here :)
 
acl
4:09 PM
OK so let's see if 9 demolishes 8 on my laptop
 
I mean, hard to see the change in the icon.
 
@Szabolcs Your image upload palette gives me this:
General::unavail: ExportString is not available in this version of Mathematica.

Java::init: Java is not running. You must call InstallJava[] to start the Java runtime.

Java::init: Java is not running. You must call InstallJava[] to start the Java runtime.

MakeJavaObject::arg: MakeJavaObject cannot convert ExportString[\!\(\*
GraphicsBox[TagBox[RasterBox[CompressedData["
1:eJzsvYeXI9d17qu3/O69Xn7ytb1sP1sWZSr4SmIQJZKyJdmS6SvJlkSRFEXO
UCSHnNCNHKoKhRwKQOecc0BOFZBz6oROkwMpyV7v/jF+51RCAT1DkbZkckgc
 
@SjoerdC.deVries evaluate InstallJava[] before trying to use it. Does it work then?
 
Trial has Export(String) disabled
 
OK. I see, no problem
 
4:12 PM
@acl btw, how do you install a different version on your mac? I just rename one of the apps (or use a different folder)
 
acl
@rm-rf same
 
The first thing I tried was Texture. I had reported problems with transparent textures before. It seems to work now!!!! Fingers crossed that it stays that way.
 
@rm-rf I use /Applications/Mathematica 8.app and /Applications/Mathematica 9.app etc.
 
Is it worth getting the trial for someone who's expecting the real deal? How long will it take?
 
@rm-rf @acl Have you also been trying the trial version?
 
4:14 PM
@Szabolcs yes
 
acl
@Szabolcs I'm copying it to my hard disk now (haven't actually ran it yet)
 
@Rojo Depends on your internet connection... it's a huge 1.something GB
 
@rm-rf so it grew
 
I'll try to control myself and wait for the full version ...
 
I'm wondering if the real deal is expected in hours, days or weeks
or minutes
My connection at home is about 35Mb/s :)
 
4:16 PM
@rm-rf Could you please test if this bug is fixed?
 
@Rojo You can always install the trial and activate with the real key when you get it. No need to reinstall (source chat.stackexchange.com/transcript/message/7043418#7043418)
 
@Ajasja But will that enable the disabled exportString and whatever else?
 
@Rojo I would suppose so.
 
@Ajasja Remote connecting to my home to download it then :), thakns
 
Broken link in online docs: more info link here links to the development site
 
4:19 PM
Downloading...
5 mins left :D
 
Ok, just to make sure... we're not tagging stuff as , right? I think we decided that by default questions are always about the latest
2
 
@Verbeia Yup, I noticed that too.
 
4
 
@rm-rf Yes.
 
@Szabolcs I now get True instead of False. I assume that's good, right?
 
4:22 PM
Yes. It's fixed then.
 
acl
ok the new input stuff is nice
 
just couldn't resist, downloading
 
1
 
Hello @DanielNewman
 
any one knows when those with premium service can download version 9? I assume one gets free upgrade if one has valid premium service. thanks
 
4:25 PM
@Verbeia should be fixed in a minute. Thanks.
 
Compiling to C does not work in the Trial version.
 
Hi @asim - is this the most people we've had in the chat room at once?
 
I am certainly new...so that shows that you may be right!
 
acl
@asim indeed. maybe it uses ExportString or something similar
 
I opened an existing notebook. Apparently there is a new default stylesheet. After automatic conversion I get these messages:
INTERNAL SELF-TEST ERROR: winFonts|c|992
Click here to find out if this problem is known, and to help improve
Mathematica by reporting it to Wolfram Research.
 
4:28 PM
@NasserM.Abbasi This should be available before the end of today.
 
@ArnoudBuzing, great! thanks.
 
@acl: It says, Export does not work in this version...
 
acl
@asim right, that's how I concluded that
 
Hello
 
There appears to be new slideshow stylesheets.
 
4:35 PM
Can I undo more than once in 9?
 
WRI web site is taking long time to load. Too many requests
 
@Szabolcs Hi!
 
@Giovanni Probably not.
 
It was definitely a good move to rename the "More information" section and make it the same style as other parts. People tended simply not to expand it in previous versions.
 
@Giovanni no indication from the documentation that they've added it, but I don't have it yet
 
4:39 PM
@Giovanni no
 
22 in the room
 
@Giovanni No.
 
Is it good to send bugs I find in trail version to support@wolfram.com?
 
@ssch What did you find?
 
@Szabolcs First run of SocialMediaData["Facebook", "Friends"] shows a broken sign-in button
(looks broken only, it still works)
 
4:41 PM
Image3D[RandomReal[1, {10, 10, 10}]] doesn't render well here, but then it's an old graphics card
 
@Szabolcs It's a bit flickery for me
first I thought it was randomizing each update, so now ofc I have to make such a thing
 
@ssch I don't have that problem but it doesn't work anyway because ExportString is not available in the trial
@Arnoud Maybe this is something to take note of: the social networks API is one of the highlights in 9, so people are likely to try it out. But the first doc example, SocialMediaData["Facebook", "Friends"], won't run because it seems to need ExportString
 
acl
@Szabolcs indeed. in addition the first time it killed the kernel
 
posted on November 28, 2012 by Stephen Wolfram

I’m excited to be able to announce that today we’re releasing Mathematica 9—and it’s big! A whole array of new ideas and new application areas… and major advances along a great many algorithmic frontiers. Next year Mathematica will be 25 years old (and all sorts of festivities are planned!). And in that quarter century we’ve [...]

 
My contribution to the version 9 blog post frenzy
How many MATLAB Toolboxes make a Mathematica 9
 
4:54 PM
@Szabolcs You mean it's blurry or it's slow? I've got the former not the latter
 
@WalkingRandomly This is pretty cool. I remember reading your similar post for V8, and revisited it also now. The change in the number of toolboxes is quite impressive.
 
@WalkingRandomly Nice. The MatLab compiler can make stand-alone executables though
 
@SjoerdC.deVries I mean that it renders all gray, without the different shades being visible
 
@WalkingRandomly great post. Now one just needs a price comparison ;)
 
This is what it's supposed to be, according to the default image:
But I get this:
Admittedly, this machine has an old graphics card, but I still believe it is a bug because the 3-channel (colour) version renders fine
 
5:00 PM
I get the first one.
 
Check out this little example in the docs: wolfram.com/mathematica/new-in-9/built-in-integration-with-r/…
 
First one here too
 
@Verbeia The file paths give away who wrote it :)
 
@SjoerdC.deVries But Matlab compiler is not available to students. In addition, many toolboxes and some functions are not supported with the compiler. Actually, the Matlab compiler is not really a compiler. Same 'm' files are used. It is just that the Matlab run-time is what is shipped with it, and all wrapped up nicely. I think WRI can make a compiler also for Mathematica if they wanted.
 
5:04 PM
@Szabolcs not necessarily. They just give away the last WRI employee to evaluate it :)
 
Ok, it probably does not make sense to hide it any longer, since this will become quite obvious soon. I am the developer of RLink, and I did that example
8
 
acl
RLink is nicely documented
oh
ha
 
@LeonidShifrin about time :)
 
@AndyRoss Yep :(
 
Awesome, Leonid
 
acl
5:06 PM
@LeonidShifrin did you write the docs too?
 
@LeonidShifrin I knew it when I read the docs. I didn't even need the "Preamble" heading
 
@Verbeia Well, time will tell, but thanks :)
@acl Yes
 
@SjoerdC.deVries haha
 
@SjoerdC.deVries Regarding MATLAB compiler--Yes it can make standalone executables but you could argue that we can do the same with Mathematica and CDF files...kinda...sorta
 
acl
@LeonidShifrin nice job (well I guess RLink itself is a good job too, but I haven't tried it in detail yet)
 
5:07 PM
@acl Well, thanks :)
 
acl
@WalkingRandomly do the matlab executables consume licenses?
 
Ack! What's a guy gotta do to score a copy of version 9 around here? I am a site license administrator and I don't get to get one :)
@acl no they don't
 
@DanielNewman Congratulations for the predictive interface. After trying it it seems like it will really be practically useful.
 
@acl good job too because that's how we run MATLAB jobs on our condor pool
 
@Szabolcs The blurryness is intentional. It's part of the rendering function used and can be changed
 
acl
5:09 PM
@WalkingRandomly that's a significant difference then: one can't submit more jobs to a cluster than licenses
(50 in our case)
 
indeed
 
gtg, will be back soon
 
We've had 3000+ jobs on our condor pool...all using MATLAB Compiled application
 
Users of the Mathematica 8 student version may encounter a startup issue when they try out the Mathematica 9 trial version.
 
acl
@WalkingRandomly yes. Here I'm basically forced to rewrite things after a while (to C or Python) to avoid this limit. not very pleasant after mathematica
 
5:11 PM
When I have more information I will post it here.
 
Going to prepare dinner now.
 
There's that little floating window that comes up when selection tool is activated on an image. I'm wondering if this kind of UI element is user-accessible (can we make something like this too)?
 
I'm off too. Bye all
 
@Leonid I think this will finally get me to learn R properly :)
I need to bike home for lunch too
see you later
 
@Szabolcs Thanks... lots of hard work from lots of folks on my team and others throughout the company :)
 
5:17 PM
One thing I like abut Mathematica is that several features which don't really look that exciting when just reading about them turn out to be very well designed and practically useful
Like Manipulate and the Dynamic functionality. It's not obvious before using it, but it is a very useful tool. The predictive interface is similar, it's not just shiny, it's something I think I'm going to use
but well, I'm saying that after playing for half an hour only :)
 
When I type something in the doc center and press cmd-k to autocomplete, the suggestions appear at the top left corner instead of where the cursor is... Does this happen on windows/linux too?
 
acl
@Szabolcs yes it's really nice. works for options too
@rm-rf hm yes same here
 
I can't reproduce that
 
acl
well not at top left, but some point not coinciding with the posn of the cursor
 
@acl I would've really liked string autocomplete for options :(
 
5:21 PM
You mean type Plo somewhere in a doc page, press Cmd-K?
 
@WalkingRandomly site update notifications are in the process of going out, but this process takes a few days usually.
 
acl
@rm-rf yes...
 
Hmm, can't download the trial from my Premier account. Guess I'll have to wait
 
@Szabolcs Well, if you really need R, this might help :)
 
@LeonidShifrin I don't really need it, but it looks exciting and I started reading the docs a few times. But I always stopped.
@rm-rf Oh I see now, you mean when typing in the search field. It appears in the middle of the screen for me, also not where the cursor is. I never used Cmd-K in that search field before
 
5:24 PM
@Szabolcs R is a fine language, if you ask me. Much better than Matlab, to my taste.
 
I really must go now, my wife will complain that lunch gets cold ...
will be back in an hour
 
@rm-rf On my Mac it's down a few "lines" from where it should be.
Vitaliy's put up a couple new ads -- one for the V9 main page, and one for a series of "Live" events.
 
5:43 PM
@BrettChampion @BrettChampion where are those ads?
 
@AndrewMoylan Hey Andrew! How's it going?
 
22 people in the room, wowo
 
@Rojo there were briefly 26 people, about half an hour ago
 
@Szabolcs How was your lunch?
 
@Leonid quick! :-)
 
5:51 PM
@Szabolcs Yep, that's exactly what I meant :)
 
but very good
 
@Szabolcs That's the best combination in some cases)
 
I live close enough to campus that it takes me 9 minutes to get home by bike
 
thanks @BrettChampion
 
@Szabolcs You are lucky. When I was in the US,it took me about 20 mins by bike to get home from campus. But actually, once I moved to that place off campus, I had to buy a car, since all stores and other places were really far.
 
5:57 PM
@LeonidShifrin congratulations for RLink. I have a question though. I know nothing about R , so I wonder: does it do something that MMA9 cannot do? If yes, then why WRI does not try harder? If no, what is the point of linking?
 

« first day (315 days earlier)      last day (4170 days later) »