scribia-developers — Discussion of Scribia development.

Re: [Scribia-dev] small quirk in 0.1.1

[Lark <lar...@un...>]

> Apparently, successive breaking will annihilate previous amount of
> successive chainings if any and be recognized as one breaking.

Fixed it.

Problem was when a chain broke on first use, it overwrote the stats for the
previous chain (the chain counter was not updated). It's not in the CVS tree
yet, I have some more things that I want to fix before I commit.

I'll generate another (0.1.2) alpha this weekend, if you like any more bugs
to be fixed before then, send mail :-)

Lark
-- 
Lark <lar...@un...>

Re: [Scribia-dev] small quirk in 0.1.1

[Ludovic N. <lni...@wa...>]

Ok. Got at least a reproducible sample on my machine. May have to test it
in case newer code changed but I doubt it.

take that as your log file:
11/23/00 12:57:20a You start dragging Tish
12/9/00 3:33:25a The chain breaks as you try to use it.
11/23/00 12:57:20a You start dragging Tish
11/23/00 12:57:20a You start dragging Tish
12/7/00 3:34:25a The chain breaks as you try to use it.
11/23/00 12:57:20a You start dragging Tish
11/23/00 12:57:20a You start dragging Tish
11/23/00 12:57:20a You start dragging Tish
11/23/00 12:57:20a You start dragging Tish
11/23/00 12:57:20a You start dragging Tish
12/9/00 3:33:25a The chain breaks as you try to use it.
12/9/00 3:33:25a The chain breaks as you try to use it.
11/23/00 12:57:20a You start dragging Tish
12/9/00 3:33:25a The chain breaks as you try to use it.
12/9/00 3:33:28a The chain breaks as you try to use it.
12/9/00 3:33:35a The chain breaks as you try to use it.


Scribia will see 4 chains with 1,2,0,0 chainings.
Apparently, successive breaking will annihilate previous amount of
successive chainings if any and be recognized as one breaking. I suppose
the scanner actually creates a "chain element" and successive chainings
somewhat distroy that overwrite that entry. That's my best guess so far.

Geotzou

Re: [Scribia-dev] small quirk in 0.1.1

[Ludovic N. <lni...@wa...>]

>Hmm. I'll have to look into that. Problem is, I only ever owned a few
>chains, so while it seems to count correctly on my logs, that isn't really a
>conclusive validation.
>
>One thing about chain counting: I start to count chain events for a "new"
>chain, when I see one of these things in the log:
>
>    1. your chain breaking
>    2. you selling your chain
>    3. you giving your chain away (selling for 0 coins)
>
>i.e. if you give your chain away (and maybe get it back), that will start a
>new series of counts. Could that explain what you see?

Since chains are apparently undistinguishable from each other, I am not
sure selling/giving/buying should be a trigger for that new chain thing. I
would prefer to have a count from one break to the next, and that's it.

But since in any case I am Scribing (Scribing is a registered trademark of
the Great Zou ;-) ) 2 chains where there were in fact 3 chains, my problem
is not overcounting but undercounting right now. I'll try to create a
single fake log file showing up the problem. If I can, i'll try to debug or
transmit it to you.

Any idea for the time issue?

Geotzou

[Scribia-dev] FW: Scribia with CW6

[Lark <lar...@un...>]

Bones (a new Scribia developer :-) sent me these instructions on compiling
Scribia on CW6. I'll make these changes to the CVS tree over the weekend,
but I'd like to post it here in case people would like to play with CW6
beforehand.

It looks like we could simply use a second project file for CW6, and our
code could compile under both 5.3 and 6. That'd be nice.
-- 
Lark <lar...@un...>

----------
Date: Mon, 18 Dec 2000 02:57:03 -0800
To: lar...@un...
Subject: Scribia with CW6

First error with CW6 is:

Error   : declaration syntax error
CScribiaDoc.h line 59   typedef map<string, UInt32>  TCounter;

with "string" highlighted. CW6 wants std::string, so i added

using namespace std;

at line 58 of CScribiaDoc.h, which makes it happy. Or you could make
them all std::string. Same effect.


....

After that, the next error is:


Error   : ambiguous access to overloaded function
'StResource::StResource(unsigned long, const unsigned char *,
unsigned char, unsigned char)'
'StResource::StResource(unsigned long, short, unsigned char, unsigned char)'
CTimeField.cp line 103   StResource timeFormat('itl0', 0);


so i changed CTimeField.cp line 103 from:

  StResource timeFormat('itl0', 0);


to

  StResource timeFormat('itl0', (ResIDT) 0);

....


To link with PP 2.1, I had to:

* Add
 - UScrap.cp
 - UPrinting.cp
 - UProcessMgr.cp
 - LAMImageWellImp.cp
 to project, per 2.1 release notes.

Amm.. i think thats it...

So, what is needed - basically to make the interface work? :)

[Scribia-dev] CVS comitting

[Lark <lar...@un...>]

A few people asked me to hint on when to commit their local changes to the
remote CVS repository. Here's what I've pulled from various earlier posts o=
n
the topic, FYI:

-----------
If you're unsure, feel free to ask me before committing. Try to commit
increments that work, i.e. don't go off for a half year with today's source=
s
and then do a huge commit. But don't commit things that crash (or don't eve=
n
compile), it pisses people off when they sync their tree and nothing works
anymore. :-)

-----------
> Hi Scribia developers! So what's the next step? I've subscribed to the ma=
iling
> list so now what? Do I register as a new user on the sourceforge site?

Depends. If you just want to grab the source and compile it, you can get it
via anonymous CVS, without the need to register with sourceforge.

If you'd like to actively develop Scribia (i.e. make larger code changes
you'd like to contribute back to me/us) I will add you as a developer. For
that, you will need to do these steps:

1. sign up with sourceforge

2. tell me your sourceforge user name, so I can add you as a developer

****IMPORTANT****
3. open a customer service item with sourceforge, and ask them to set up CV=
S
   "pserver" access to the scribia CVS repository (SSH does not work on the
   Mac)
****IMPORTANT****

After that, you have read/write access to the CVS repository.

Since I don't know if people have used CVS before, here are some short
things you should know:

=80 commit log entries are your friend, please take the time to describe
  changes made to the code

=80 never check in code that has known bugs, or worse, does not compile
  (people will get angry if they'll have to fix your bugs after they have
  CVS updated your changes)

=80 try to use the indentation style of the rest of the sources for new
  things/changes (I've tried to stick to the "Mac way" of indentation and
  naming things as much as I could figure it out; it's not perfect)

=80 include the HEADER file (Documentation folder) at the top of each new
  source file, it has some magic CVS tags in it that auto-insert the commit
  log, current version, etc. (it also has the disclaimer)

--=20
Lark <lar...@un...>

Re: [Scribia-dev] small quirk in 0.1.1

[Lark <lar...@un...>]

> From what I see, the individual chain data isn't perfectly stored right now.
> 
> In my case, it was seeing 3 broken chains before and now sees only 2 (or
> presents data for 2).
> It says 105 and 0 where it should be 105, 1, 0

Hmm. I'll have to look into that. Problem is, I only ever owned a few
chains, so while it seems to count correctly on my logs, that isn't really a
conclusive validation.

One thing about chain counting: I start to count chain events for a "new"
chain, when I see one of these things in the log:

    1. your chain breaking
    2. you selling your chain
    3. you giving your chain away (selling for 0 coins)

i.e. if you give your chain away (and maybe get it back), that will start a
new series of counts. Could that explain what you see?

Lark
-- 
Lark <lar...@un...>

Re: [Scribia-dev] Bug in the alpha

[Lark <lar...@un...>]

> Hi, just checked out the alpha and had it scan every one of my log files.
> The scanner isn't picking up Rodnus training.  Lark, if you need the message
> again, let me know.  I'll dig it up.

Here's what I scan for, is this correct for these "new" trainers?

ANGILSA   "*You notice your balance recovering much more quickly"
KNOX      "* You notice yourself doing much more damage"
RODNUS    "*You feel more receptive to healing"

Maybe there's a space between the star and the first letter for other
messages, not only Knox?

Also: I need the training messages and names of the languange trainers. So
far, I think I only have Sylvan and Dwarven.

> I think there may have been a change in some of the trainer messages between
> now and the war.  I have a bunch of undercounts and even an over count.  I
> know a few of the under counts are due to me not having my logs of the first
> couple days.  But the number they're off is to large and some of them are
> with upper circle trainers.  I'm going to rescan later tonight with scribus
> to see what he counts.  Hopefully it'll show something.

One problem that was fixed shortly after 0.1.1 was that the scanner would
pick up things in exiles' actions. That could be reason for some
over-counts. I'll generate another alpha (0.1.2) this weekend that has some
other fixes as well, let me know what that one does, ok?

In general you can help with this by scanning your log files for the phrase
from Scribia.l that was over-counted. Both CW and BBEdit can search for text
in multiple files. If you see something that should not be matched but is,
email me the line(s). (This is exactly what I do if you send me your
miscounted logs :-)

Lark
-- 
Lark <lar...@un...>

Re: [Scribia-dev] email problems

[Ludovic N. <lni...@wa...>]

Hi,

Seems to me that the share times are wrong with the build i have in some
cases. (Not the most recent version, i have to do the update and verify
again.

My current guess is that the time difference behaves badly when one sharea
acroos day boundaries or 12hours cycles. But the bug doesn't seem to be
universal so I can'T provide specific guidleines right now. I have a case
here where an hour an 17 minutes is reported as 13:17 by scribia. And i am
scanning a single file, it's not an addition problem. I'll proceed to some
more testing before investigating the code for that.

I'll be away from main dev machine for about 10 days also, so no hurry on
that bug. ;-)

Ludo

[Scribia-dev] email problems

[Lark <lar...@un...>]

It seems that iName (which manages my email address) had some problems last
week, so I'm just now getting all the mails that were sent.

I'll work on Scribia over the weekend and respond to your mails, too. (Busy
at work.)

For people who sent me mail privately: Please send to
scr...@li... - reasons:

    - it's archived, some questions have been answered before
    - other subscribers can help with questions
    - it's filtered into my Scribia folder, so I don't miss mails (normally)

Thanks,
Lark

PS: We have two new developers: Geotzou and Bones.
-- 
Lark <lar...@un...>

Re: [Scribia-dev] small quirk in 0.1.1

[Ludovic N. <lni...@wa...>]

From what I see, the individual chain data isn't perfectly stored right now.

In my case, it was seeing 3 broken chains before and now sees only 2 (or
presents data for 2).
It says 105 and 0 where it should be 105, 1, 0

I have not traced in the code to look for the error yet so I can't really
tell you what happened.

Geotzou

[Scribia-dev] Bug in the alpha

[Brad M. <bm...@as...>]

Hi, just checked out the alpha and had it scan every one of my log files.
The scanner isn't picking up Rodnus training.  Lark, if you need the message
again, let me know.  I'll dig it up.

I think there may have been a change in some of the trainer messages between
now and the war.  I have a bunch of undercounts and even an over count.  I
know a few of the under counts are due to me not having my logs of the first
couple days.  But the number they're off is to large and some of them are
with upper circle trainers.  I'm going to rescan later tonight with scribus
to see what he counts.  Hopefully it'll show something.

I'll let you know what I can find.


Brad  (aka Cygnus)
cy...@po...
_________________________________________________________________
Oh the words that he spoke
Seemed the wisest of philosophies
There's nothing ever gained
By a wet thing called a tear
When the world is too dark
And I need the light inside of me
I'll walk into a bar
And drink fifteen pints of beer                 -   Shane MacGowen

[Scribia-dev] small quirk in 0.1.1

[Lark <lar...@un...>]

I've just committed a fix to a small bug in 0.1.1 that caused text in
\action lines to not be ignored. This could lead to garbage text being
matched for item/beast names, or other things.

I've also included some initial support for consistently counting items for
which DT has changes names over time (wood, berry clusters).
-- 
Lark <lar...@un...>

[Scribia-dev] cvs updates

[Lark <lar...@un...>]

I've committed and tagged all my recent changes. The sources are now in sync
with the just-released scribia-0.1.1-alpha.
-- 
Lark <lar...@un...>

[Scribia-dev] Scribia - first public alpha

[Lark <lar...@un...>]

Please grab the first public release of Scribia at
http://download.sourceforge.net/scribia/scribia-0.1.1-alpha.sea.hqx

This is alpha-quality software. The user interface is ugly, and most
document functionality isn't there yet. However, the log scanner works
(fast!), so please run it over your logs and give use some feedback.

Send suggestions for future improvements to
scr...@li...

File bug reports at http://sourceforge.net/bugs/?group_id=6424
(Please don't send bugs by email, use the database at the URL above.)

Feel free to forward/repost this announcement as appropriate.
-- 
Lark <lar...@un...>

Re: [Scribia-dev] CVS update

[Brad M. <bm...@as...>]

On 12/10/00 11:22 AM, "Lark" <lar...@un...> wrote:

> Updated Scribia:
>   * counting drags per chain
>   * counting jury duty events
>   * counting ruined furs
>   * counting bank fees
>   * fixed a timestamp scanner bug (messed up some share times before)
> 
> I'll upload a binary to the FTP area so people can play with it without
> having to compile it. Problem is that sourceforge's FTP is down right now.
> I'll be back in town by Tuesday night, I'll try again then...

My semester ends this week, so I'll probably be able to start learning power
plant again.  Was the project moved to CW6?  I remember something being said
about it awhile back.


Brad 
bm...@as...

_________________________________________________________________

The trouble with the world is that
the stupid are cocksure and the
intelligent are full of doubt.
        - Bertrand Russell

[Scribia-dev] CVS update

[Lark <lar...@un...>]

Updated Scribia:
    * counting drags per chain
    * counting jury duty events
    * counting ruined furs
    * counting bank fees
    * fixed a timestamp scanner bug (messed up some share times before)

I'll upload a binary to the FTP area so people can play with it without
having to compile it. Problem is that sourceforge's FTP is down right now.
I'll be back in town by Tuesday night, I'll try again then...
-- 
Lark <lar...@un...>

[Scribia-dev] Re: <no subject>

[Lark <lar...@un...>]

> I have another friend  who pkayed pre-ripture who just bought back the game
> a couple days back and will start playing by christmas time that may
> actually take a look at Scribia too. He downloaded the source tonight after
> i talked to him and toyed with it like me. He was Yaruhl pre-ripture, but i
> doubt you remember him.

Great! Have him send me email so I can set him up as a developer.
 
> About chaining events, the new scanner is perfectly accurate with my recent
> test sample, which is real cool. I never thought about the being chained,
> neat ;-) I'll see later on how to handle the fine detail of what i want by
> adding to your chain callback function (if i am not mistaken, you can give
> a handler to lex for any given token).

Basically, what needs to be added is a TChainingsPerChain counter, and then
just add mChainingsPerChain["Chain 1"] = 15; or whatever the current counter
is when one breaks. I'll look into it when I lie fallen later on :-)
 
> At first, I may actually commit changes within #ifdef for you to look but
> without them being effective in standard builds. Only major changes will
> have that.

Good idea.
 
> Except that it crashes my machine almost all the time, i agree. If i ever
> get into a major problem, i'll revert to asking someone else.

Still? That's bad. I can keep sending you binaries per email, just ask me
when the import doesn't work.
 
> I may actually change the layout for the "non-UI" window slightly. I like
> to keep that window around as a summary window. It may be a bit long, but
> it can display a lot.

Go ahead :-)

PS: We should move this exchange over to scribia-developers, so the others
see it.
-- 
Lark <lar...@un...>

[Scribia-dev] Scribia updated

[Lark <lar...@un...>]

Hi,

I've added support for some events related to chaining and jury duty to the
scanner. (No Geotzou, not "full" stats yet, just some :-)

Lark

PS: We still need a real GUI!
-- 
Lark <lar...@un...>

[Scribia-dev] CW6

[Lark <la...@fe...>]

Hi,

has anyone compiled Scribia with C6 yet? Any problems?

I'm still using CW5.3, but there are two new people who may work on the GUI;
both would rather use CW6. Since we really need that GUI, I'm willing to do
almost anything to accommodate whoever will build one. :-)

In the meantime: Any new trainers/items/etcs. in the lands that need
counting support from the scanner?

-- 
Lark <la...@fe...>

[Scribia-dev] Joining the Scribia Mailing List

[Tim D. <fe...@xt...>]

[Scribia-dev] CL: Scribia - Call for Developers

[Lark <la...@fe...>]

Hello,

a while ago, I started the free open-source Scribus replacement "Scribia", 
implemented in C++/flex/PowerPlant. At this time, the log scanner engine is 
relatively complete (i.e. does most things Scribus did), and is much faster.

However, I don't know much about Mac GUI programming, and I don't have the time 
to learn it. The GUI is the one critical piece that is missing before we can 
release an alpha version of Scribia for wide-spread testing.

If you have experience with Metrowerks PowerPlant, please consider helping us 
out! The initial GUI can be very simple, just enough so people can start 
testing Scribia.

Scribia development is hosted on SourceForge.com 
(http://sourceforge.net/projects/scribia/). You can obtain the sources from 
there. I will gladly give CVS write access to new developers.

All developers (and everybody interested in keeping up with Scribia 
development) are encouraged to subscribe to the "scribia-developers" mailing 
list. This can also be done from the above web page.

Please help train Scribia! Eternal fame and glory are up for grabs! :-)

Lark

Re: [Scribia-dev] Feature Request for Scribia...

[Lark <la...@fe...>]

> There was a recent post on the newsgroup about ways to record ranks.
> Someone mentioned Scribia...
> 
> There was also a request for Scribia:
> 
>> If someone is working on the Scribus Open Source project and reads
>> this, I have a request:
>> 
>> Please allow me to drag multiple files onto the Scribus icon and
>> read them all.

Well, drag&drop isn't there just yet, but you already can add all files in a
folder, which is somewhat similar.

I'm back from my trip, by the way.  :-)
-- 
Lark <la...@fe...>

[Scribia-dev] Feature Request for Scribia...

[Pugwash <pu...@ac...>]

Hi

There was a recent post on the newsgroup about ways to record ranks. 
Someone mentioned Scribia...

There was also a request for Scribia:

>If someone is working on the Scribus Open Source project and reads 
>this, I have a request:
>
>  Please allow me to drag multiple files onto the Scribus icon and 
>read them all.
>
>  I have every log made with Merlisk since the Rip War and I don't 
>want to add each one individually.  :)
>
>  --
>  Best regards,
>  HWC for Merlisk


bye

Pugwash
-- 

----------------------------------------------------------------------
Pugwash
Dwarven Militia: http://www.dwarvenmilitia.com/

Puddleby Mailbox: 1027
mailto:pu...@ac...
ICQ: 11524889

Clan Lord: http://www.deltatao.com/clanlord.html
Hotline: 24.72.12.31

Re: [Scribia-dev] Trainer messages

[Ludovic N. <lni...@wa...>]

>I got a message about "the language of the trees" recently, and got
>experience for it. I wonder if it's worth counting, and if so, what it
>means...

It's the orga camp equivalent of the free rank you get the first time you
go to the old cemetary (fromd ead tree in marsh). Everyone gets it once and
is done after that. Limited interest for the parser except for completeness
purposes.

geotzou

Re: [Scribia-dev] Trainer messages

[Lark <la...@fe...>]

>> By the way, I noticed the other trainers have that little circle-thingy (€,
>> option-8) in front of them with no space between it and the "You". These
>> seem to have stars instead, and Knox has an extra space. Are you sure this
>> is correct? 
>
> As far as I know it is correct.  I have not trained with either Knox or
> Angilsa yet.  I asked someone who was training with them (I think it was
> Kojiro) to give me the exact line from their log.

Gotta love how inconsistent DT is with their messages. Good thing we use
flex, I'll just pattern match for all possible combinations (star/option-8,
with/without spaces). Soon.

> As far as I know, that is all that has been added that will effect the
> scanner.  I'll let you know if I think of or come across anything else.

I got a message about "the language of the trees" recently, and got
experience for it. I wonder if it's worth counting, and if so, what it
means...

I'm also considering keeping track of which quests one has solved by
scanning for e.g. the message you get when you get back to the rat challenge
person.
-- 
Lark <la...@fe...>