ImproveMathEquationEditor/IRCMeetings/May2010/13rd May2010

This page is the IRC log of the 13rd May 2010 IRC meeting ( #education.openoffice.org ) with students for Google SoC 2010.

Back to May 2010

[18:03] ericb2: hi

[18:04] * Spiso (~chatzilla@stu405.sjc.ox.ac.uk) has joined #education.openoffice.org

[18:05] Spiso: hello

[18:05]  hi

[18:05] meeting time ?

[18:05]  now

[18:05]  yes, it is

[18:05] * kohei_ (~Kohei@opensuse/member/kohei-yoshida) has joined #education.openoffice.org

[18:05] just one minute

[18:05] ok...

[18:08] ok, I'm back

[18:08] * ericb2 see we didn't define an exact agenda

[18:08] so I'll propose one

[18:08] Agenda for 13rd May 2010

[18:08] 1) Point on the tasks

[18:08] 2) Objectives for the coming week

[18:08] 3) Misc

[18:08]  do

[18:08]  agree

[18:08] Point 1 : point on the tasks

[18:08] sounds good...

[18:09] who starts ?

[18:09] kohei_: rodo_ hello btw :)

[18:09]  hi there

[18:09] What do you mean, ? tasks identified ?

[18:09] jopsen: yes

[18:09]  hola ericb2

[18:10] For the record, the log of the meeting will be put there : http://wiki.ooo4kids.org/index.php/ImproveMathEquationEditor/IRCMeetings/May2010/13rd_May2010

[18:10]  ericb2: so we should do some kind of recap, what we've done?

[18:10] I haven't really changed mine since last time... that is:

[18:10] # Determine new representation of cursor position

[18:10] # Figure out how to draw the cursor

[18:11] for the moment being... (there's more task later) :)

[18:12] Since last time, I've been analysing how the nodes are composed and written a small patch to dump equations to graphs using graphviz

[18:12] jopsen: interesting

[18:12] Examples are linked to here: http://www.freedesktop.org/wiki/Software/ooo-build/SummerOfCode/2010/Jonas#Notes

[18:12]  I did some tracing for the values as you told me, I just have to rewrite it so it's comprehensible text

[18:13]  I found some minor error, but it shouldn't be that bad

[18:13] Spiso: did you find some anomaly, say some potential bug ?

[18:13]  yes, as I just said, but it should cause the misallignment

[18:14] Spiso: if so, I'd suggest you to write a patch, we could test all

[18:15]  it wont make much difference

[18:15]  one is that in arrange function of Expression node there is a trick of imagining that there is letter 'a' written in that

[18:16]  and later the node is extended by it's actual content

[18:16] Spiso: you mean sort of "initialization" ?

[18:16]  the problem arises when the actual content is smaller then the letter

[18:17]  kind of, the reason for doing it is that when you leave the expressionNode empty it's useful if it has nonzero size

[18:18] jopsen: some tracks about cursors :  http://wiki.ooo4kids.org/index.php/AddNewCursors

[18:18] <Spiso> this is a comment in the code:

[18:18] <Spiso> 00845         // provide an empty rectangle with alignment parameters for the "current"

[18:18] <Spiso> 00846         // font (in order to make "a^1 {}_2^3 a_4" work correct, that is, have the

[18:18] <Spiso> 00847         // same sub-/supscript positions.)

[18:18] Spiso: you're right

[18:19] Spiso: does it work for you ?

[18:20] <Spiso> haven't tried actually

[18:20] <Spiso> I'll do, now

[18:20] ericb2: by cursor I probably mean caret...

[18:20] Spiso: yes, please. I'm curious to see what happens

[18:21] jopsen: ah, ok, so the carret is calculated in the sallayout (vcl), but there is probably something interesting in starmath and writer too

[18:21] jopsen: yes, I read something about the carret. Hang on

[18:22] in starmath

[18:23] * Spiso has quit (Remote host closed the connection)

[18:23] ericb2: as far as I've concluded so far it's currently drawn as a rectangle...

[18:24] jopsen: yes it is

[18:24] ericb2: but about position representation of the caret I'm thinking a struct of SmNode* and bool BeforeOrAfter

[18:25] * Spiso (~chatzilla@stu405.sjc.ox.ac.uk) has joined #education.openoffice.org

[18:25] <Spiso> :) my computer crashed

[18:25] <Spiso> and I lost some of my notes :(

[18:25] Spiso: too bad :/

[18:26] ericb2: Using such struct for caret position representation... All I'd need is a bunch of methods to compute the next position... I'm considering a visitor-pattern for that...

[18:27] jopsen: I'm not design patterns specialist, but I'll follow you there

[18:28] ericb2: I imagine every node have a method for determining what the next position to the left of it is...

[18:28] ericb2: and by left I mean... Left as in left of it on the screen...

[18:28] <Spiso> ericb2: I can't see whether it works, there are too many rectangle :), do you know how to change the size once the formula is created?

[18:29] Spiso: in the menu -> Format -> font size or something similar

[18:29] Spiso: when the editor is active

[18:30] Spiso: other way  : change the zoom value

[18:30] ok, can we summarize, and go to point 2 of the agenda ?

[18:31] 2) Objectives for the coming week

[18:31] good..

[18:32] <Spiso> in the a^1 {}_2^3 a_4 the middle indexes are lower then a^1, a_4

[18:32] <Spiso> so it doesn't work that good

[18:32] I plan to make some representation of the caret... and see if I can get it moving...

[18:32] Spiso: looks like. Please log screenshots and cases you tested

[18:33] jopsen: that's a good idea

[18:34] jopsen: do you think this will be done for next week ?

[18:34] hmm... depends on my sleep needs... I still have a lot of school work...

[18:35] jopsen: ok, so we'll consider you'll continue to work on that. Ok ?

[18:35] but think some naive sort of movement is likely...

[18:35] yes...

[18:35] <Spiso> ericb2: I can't zoom it when changing the object, it's disabled don't know why

[18:36] <Spiso> ericb2: I wanted to resize it once it's done, like a picture

[18:36] Spiso: and can you modifiy the font size ?

[18:36] <Spiso> ericb2: yes, but only up to 200%

[18:36] Spiso: ah, not that big

[18:37] <Spiso> ericb2: how did you make your screenshots?

[18:37] <Spiso> ericb2: because the mine would be kind of useless so small

[18:37] Spiso: it depends : Windows, Mac Linux .. on Linux, there are bunch of applications doing that. Else, a dump of the video memory > into .jpg should do it ;)

[18:39] Spiso: other sugestion : comment what is disturbing, e.g. only draw rectangles, but not some lines, or the opposite. Use environment variables to achieve that, instead of rebuild everytime

[18:39] <Spiso> ericb2: but you have to make OOo to display it larger, otherwise it's going to hav esmall resolution anyway

[18:42] Spiso: normaly, when you quit the Equation editor, the zoom can be modified a lot

[18:42] Spiso: one example to use environment variables (please adapt to your case)

[18:42] BOOL AquaSalMenu::VisibleMenuBar

[18:42] {

[18:42]   static const char *pExperimental = getenv ("AQUA_NATIVE_MENUS");

[18:42]   if ( (pExperimental && !strcasecmp(pExperimental, "FALSE")) )

[18:42]  return FALSE;

[18:42] return TRUE;

[18:42] }

[18:43] Spiso: there is an else before return TRUE

[18:43] sorry

[18:45] Spiso: to summarize, can we consider you'll continue to work on that for next week.

[18:45] ?

[18:45] <Spiso> yep

[18:45] jopsen: Spiso from my side, I'd  like to see you on IRC in meantime, say some evening

[18:46] <Spiso> I'm just not sure where am I supposed to write the thing for environment variables

[18:46] <Spiso> should I try to cateorize the arrange methods for the next week?

[18:46] Spiso: in the starmath code, if DEBUG, some colored lines are drawn

[18:47] <Spiso> besides the two small bugs I found, it looks ok

[18:47] Spiso: you can decide to select what will be drawn (in DEBUG mode that is), using environment variables

[18:47] ericb2: you want us to appear on IRC anytime specific? or just more in general ?

[18:47] Spiso: if I remember correctly, there are red colored rectangles, and other green and blue information

[18:48] <Spiso> well, I know where is the code for drawing

[18:48] jopsen: on #go-oo, or there or #ooo4Kids , no problem

[18:48] <Spiso> but I'm not sure where to take the env. variables from

[18:49] Spiso: before to run gdb, you can set / unset those variables

[18:49] <Spiso> but this doesn't help much anyway

[18:49] <Spiso> alright

[18:49] <Spiso> I want to make it large

[18:49] e.g. doing : export COLORED_RECT=FALSE

[18:50] <Spiso> I understand this now, but still, for the screenshots it's better if everything is large

[18:50] and in the code, you can define a static bool, e.g. returning true or false, depending on the variable value. then you can protect the part where something is drawn, or not

[18:51] Spiso: I'm sure this is something easy to do

[18:51] <Spiso> you said you can zoom it when you are editing the object or after you close SM?

[18:51] Spiso: after you close SM

[18:51] Spiso: the zoom is the writer one

[18:51] Spiso: the ruler is on bottom right

[18:51] <Spiso> once I managed to do it, but now it's always disabled

[18:51] s/ruler/slider/

[18:51] Spiso: strange

[18:52] <Spiso> the ruler works, but I mean resizing it like a picture

[18:53] * fardad has quit (Quit: \\// LL&P)

[18:53] * lgodard has quit (Quit: Leaving.)

[18:53] <Spiso> I couldn't drag it, but because of some error I could set it in the object menu

[18:53] <Spiso> btw, now the case a^1 {}_2^3 a_4 works

[18:53] jopsen: Spiso can we agree the next meeting date : next tursday 20th May , same hour  ?

[18:54] <Spiso> it was wrong just because I changed the size of letters

[18:54] Spiso: please log every test, and I'll have a look,  + I'll test myself

[18:54] <Spiso> yes

[18:54] ericb2: fine with me...

[18:54] ok ^^

[18:54] * ericb2 has changed the topic to: "Welcome on OpenOffice.org Education Project IRC channel !  Mailing list: dev@education.openoffice.org  Wiki: http://wiki.services.openoffice.org/wiki/Education_Project. Please visit http://wiki.services.openoffice.org/wiki/Education_Project/Effort, next  IRC meeting: Thursday 20th  May 2010 16:00 UTC, 18:00 CEST (Par [18:54]  <Spiso> can I write the diff?

[18:55] <Spiso> the thing is I can't commit it

[18:55] Spiso: you can paste the diff if you want, but please keep a trace somewhere

[18:57] <Spiso> alright so should I start for next week look at the less simple cases?

[18:57] Spiso:  that's fine with me

[18:57] <Spiso> btw, could you recommend a graphic debugger?

[18:57] <Spiso> it's a real pain in gdb

[18:57] Spiso: I don't know much on Linux ... maybe DDD ?

[18:58] Spiso: this is a gdb frontend in fact, but more appealing maybe

[18:58] <Spiso> jopsen: what debugger do you use?

[18:58] Spiso: I use beaverdbg

[18:58] http://code.google.com/p/beaverdbg/

[18:58] <Spiso> alright

[18:58] <Spiso> I'll have a look at is

[18:58] <Spiso> it

[18:58] * ericb2 too

[18:58] Spiso: it's slow... but easy to use... :)

[18:58] jopsen: thanks : I'll go to bed less stupid :)

[18:59] Ok, then end of meeting

[18:59] <Spiso> yep

[18:59] yes...

[18:59] * ericb2 children asked me to cook crepes :)

[18:59] later

[18:59] <Spiso> :)

[18:59] <Spiso> see you