---Reply to mail from Marcello Perathoner about Why are bookmarks sorted?
I'm writing a custom text/plain parser.
I'm parsing a text file and as I go along I add all chapter heads to the
bookmark list using:
PluckerDocs.PluckerTextDocument.add_bookmark
[...]
I'm wondering why
After only 6 years of faithful service(!) mt Palm IIIx is on its last legs.
I'm looking at the Tungsten E2.
Any comments? Advice?
Christopher R. Hawks
HAWKSoft
-
Any research done on how to efficiently use computers has
---Reply to mail from Henrik Kjolhede about Cpluck
Yes, but in the Windows distributions I don't think the library
is included. I tried a few of the different versions, but no
luck. Only the dll's and applications were included.
I'll have to show my ignorance of Windows... I thought the
---Reply to mail from Bill Nalen about Cpluck
If someone would be kind enough to send links to the libraries that have
been added, I can take a shot at updating it.
Bill
Bill:
The regex library source is in the Plucker distribution
'plucker/tools/regexp' and Magick++ is at
---Reply to mail from David A. Desrosiers about Cpluck
I've tried that now, with MinGW and MSYS, and havent' had any real
success. So if one of you would be willing to mail it to me I'd be
very grateful.
I'm not sure it runs on Windows (yet). Someone who runs
Windows and has the
Weird!
Any more news, insights, demonstrations???
see below
Christopher R. Hawks
HAWKSoft
-
Let's call it an accidental feature.
-- Larry Wall
___
plucker-dev mailing
---Reply to mail from Michael Nordstrom about Cpluck
On Thu, Apr 21, 2005, Chris Hawks wrote:
I recently found a c++ regex library that does PERL style matching. I
think we should use it (it's GPL) as it would free us from using OS
specific 'C' regex calls for Exclusions.
Sounds like
I recently found a c++ regex library that does PERL style matching. I
think we should use it (it's GPL) as it would free us from using OS
specific 'C' regex calls for Exclusions.
http://unix.freshmeat.net/projects/cpp_regex/
Christopher R. Hawks
HAWKSoft
OK Alex:
What do I have to do to compile the viewer???
I don't use the speedup and all make does is carp.
doclist.c:37: cpuspeed.h: No such file or directory
Christopher R. Hawks
HAWKSoft
-
Any research done on how
---Reply to mail from David A. Desrosiers about Plucker in C (++)
If I did everything right, Cplucker is now in CVS.
Shouldn't that have gone into /parser/cpp or something?
You're right. Can it be moved??
---End reply
Christopher R. Hawks
HAWKSoft
Thanks David ( Alex) !!
Christopher R. Hawks
HAWKSoft
-
Anyone who says 'disk is cheap' deserves to be shot.
--Linus Torvalds
___
plucker-dev mailing list
---Reply to mail from Alex Sanielevici about Launching external app from link
in Plucker - PPI or other extension?
The app I'd need to launch is a third-party drug interaction analyzer
tool (Lexi-Comp Interact), without any arguments passed in. I think what
you're saying is I need to do a
---Reply to mail from Alex Sanielevici about Launching external app from link
in Plucker - PPI or other extension?
I'm interested in launching an external application from a plucker
document.
I've been looking at PPI, but this seems to require the user configure
the application to be
Bug 1361 is now squashed. Only ConversionParser.py left tempfiles for
Spider.py to delete and that is fixed. The extranious code to delete
tempfiles beginning with tempfile.template([EMAIL PROTECTED]) is gone.
tempfile.template has been depricted since python 2.0
Christopher R. Hawks
HAWKSoft
nextTargetByteOffset shouldn't be #ifdef'ed for SUPPORT_ANNOTATION.
So, I fixed that so it would compile without annotation and paragraphTop
too.
Christopher R. Hawks
HAWKSoft
-
I keep reminding myself of that old saying:
Alex:
Did you know that you can't compile the viewer with Word Lookup and
without Annotation?? It complains about paragraphTop being undefined in
line 2372:
offsets[ len ] = prevPosition - paragraphTop;
I'm spending 4 days a week in Illinois working and 3 days in Indiana
packing,
I just upgraded to the most recent CVS and got a NULL handle exception for
my trouble. The problem is 'no hires'. longbar.bmp is ifdef'ed (HIRES and
HANDERA), but, always used in the mainform.
These changes to viewer.rcp.in fix it.
--- viewer.rcp.in.orig Sat May 1 22:44:45 2004
+++
---Reply to mail from Alexander R. Pruss about DIA code committed
The compile problems are fixed.
Whatta guy! Thanks!
---End reply
Christopher R. Hawks
HAWKSoft
-
Let's call it an accidental feature.
-- Larry Wall
---Reply to mail from James Watkins-Harvey about DIA code committed
3) The snapshot source version (17h and 18h) seems to not compile when
--disable-palm-dia is passed to configure.
No, it doesn't. The problems are (mostly) with the #defines in resize.h
SaveResizePrefs() takes 3
---Reply to mail from David A. Desrosiers about Help with the Plucker Format
I'll work on it over the next week (getting married next week in Jamaica, so
things are a bit... weird).
Congrats!!! (I'm back in Indiana for the weekend and catching up on my
spam^H^H^H^H e-mail.)
---End reply
---Reply to mail from Chris Hawks about Bookmarks
Here's another one. the attached patch allows multiple pages/files to
be given to the parser. With --bookmarks=pages this allows unlinked
pages/files to reside in a single database with bookmarks for each page.
plucker-build --bookmarks
---Reply to mail from Michael Nordstrom about Bookmarks
On Mon, Apr 05, 2004, Chris Hawks wrote:
Checked in. The only problem I've seen is that Wired and Linux Today
name their pages '0020302305,203.html'. not easy to choose their bookmark.
Maybe we could use the title instead
---Reply to mail from Chris Hawks about Bookmarks
For external bookmarks. are we talking about supporting bookmarks like
these?? (Lifted from w3c, hope they don't mind)
(In the HEAD section)
link rel=bookmark href=#technologies title=Technologies | /
Doesn't look too bad. In TextParser.py
---Reply to mail from Michael Nordstrom about Bookmarks
On Mon, Apr 05, 2004, Chris Hawks wrote:
bookmarks.diff attached for your perusal. I'll wait 'til after the
freeze to commit the changes.
Maybe we shouldn't wait that long. External bookmarks support is
included in the viewer, so
---Reply to mail from Michael Nordstrom about Bookmarks
On Mon, Apr 05, 2004, Chris Hawks wrote:
We could call it a bug (viewer supports, but, python parser does not) G
Or a delayed feature implementation ;-)
It works for me, but, I always like to have someone else try it.
Works
---Reply to mail from Michael Nordstrom about Bookmarks
On Mon, Apr 05, 2004, Chris Hawks wrote:
bookmarks.diff attached for your perusal. I'll wait 'til after the
freeze to commit the changes.
Maybe we shouldn't wait that long. External bookmarks support is
included in the viewer, so
For external bookmarks. are we talking about supporting bookmarks like
these?? (Lifted from w3c, hope they don't mind)
(In the HEAD section)
link rel=bookmark href=#technologies title=Technologies | /
link rel=bookmark href=#news title=News | /
link rel=bookmark href=#search title=Search | /
---Reply to mail from Jewett, Jim J about Guided tours
When you say the code is already in the viewer, that suggests
a specific type of tour. What did you want it to do?
Named anchors? (Web pages can have next and previous anchors.)
Autoselecting anchors based on the text? (Opera's
Robert:
Have you checked bug 893 recently?? (Inline multiimages don't link to
full-size)
It works fine for me.
Christopher R. Hawks
HAWKSoft
-
NT is the only OS that has caused me to beat a piece of hardware to
---Reply to mail from Chris Hawks about bug 893
Robert:
Have you checked bug 893 recently?? (Inline multiimages don't link to
full-size)
It works fine for me.
Whupps! My bad, they weren't quite big enough to be multiimages.
Now fixed in CVS.
---End reply
Christopher R. Hawks
---Reply to mail from Alexander R. Pruss about PPI
I see you added support for DBicty to PPI. You negelected to change
(or commit, maybe) ppi.rcp.in to go with it.
I changed ppi.rcp instead of ppi.rcp.in locally. Oops, fixed.
Also, it would be simpler (I think) to add only 1 BDicty
Alexander:
I see you added support for DBicty to PPI. You negelected to change
(or commit, maybe) ppi.rcp.in to go with it.
Also, it would be simpler (I think) to add only 1 BDicty to the lists and
tell users of the 'free' version that they must 'paste' to lookup the word.
Christopher R.
---Reply to mail from Alexander R. Pruss about monochrome images
From: Chris Hawks [EMAIL PROTECTED]
That would negate the 'feature' discussed in TextParser.py at about
line 1556. (set a color= attribute in the IMG tag to render bw icon in
one color).
As far as I can see, that feature
---Reply to mail from Alexander R. Pruss about monochrome bitmaps
I've just noticed that indeed my version can render different parts at
different depth. This could, surely, be fixed if it indeed produces
problems.
That's what I see. Mostly it occurs when the 'background' color isn't
---Reply to mail from Jewett, Jim J about monochrome images
Chris Hawks:
Multiimages look bad if they are not all the same depth.
If purple on tile one becomes black on tile 2, I
understand.
Are you saying this happens even when black becomes
black, but we chose a lower bit-depth
---Reply to mail from Alexander R. Pruss about monochrome images
On Fri, 19 Mar 2004, Chris Hawks wrote:
I don't know from black and purple (my palm is 4 bit, better than a
2-bit machine 8^ ) But, as in my previous e-mail, the problem shows when
the 'background' color of the 'color' part
---Reply to mail from Alexander R. Pruss about monochrome images
Do you think one could commit this patch to force the background in
images to white, thereby ensuring consistency across all images in a doc,
even if some have been depth-reduced to bw? Or would that be a feature?
That
Bugs 901 and 1158 are fixed. TextParser.py now forces compliance with w3c
recommendation for XHTML on HTML agents. Tags like tag/ are
converted to tag /.
Christopher R. Hawks
HAWKSoft
-
I love the way Microsoft follows
---Reply to mail from Alexander R. Pruss about monochrome bitmaps
I am working on PalmImagePlugin.py to recognize when an image really is
monochrome to downgrade the bpp for that image.
I saw this and it may be a bad thing. Multiimages look bad if they are
not all the same depth. I had
---Reply to mail from Michael Nordstrom about bug 1167
On Sun, Mar 14, 2004, Chris Hawks wrote:
Does Mike remember why multiline.Anchor was included in the test??
Don't you remember what we did 4 years ago? ;-)
Nope! 8^) and my mail archive only goes back to Jan 2001.
---End reply
---Reply to mail from Michael Nordstrom about Auto-starting a document in the viewer
On Sun, Mar 14, 2004, Michael Nordstrom wrote:
Since we always have to update the doclist (using InitializeDocInfoList)
This will be too slow for normal use; I guess it should be enough to
just get the the
Alex:
RE: bug 1167 (justification doesn't always work). Justification won't
occur if we're in an anchor (multiline.Anchor is true at the beginning of
the line). I'm not sure this is the correct behavior (mozilla doesn't
appear to care about anchors). It looks as OK removing the test as it
Jim (J):
It appears (to me) that I've fixed bug 1135 (alt text or [IMG] not shown in
table cells). If you can confirm this, I'll close the bug.
Christopher R. Hawks
HAWKSoft
-
The ability to watch M*A*S*H on demand
'Bug' 1147 resolved.
Compile time switch added '--enable-reutrntoapp'. Library and viewer both
act the same way.
Christopher R. Hawks
HAWKSoft
-
In most countries selling harmful things like drugs is punishable.
Then
---Reply to mail from Alexander R. Pruss about multiimage format
DBFormat.html doesn't seem to describe the multiimage format, does it? Alex
It does now. 8^) It makes sense to me, but, I know how it works.
---End reply
Christopher R. Hawks
HAWKSoft
We now support in PPI for languages. (Or we will when the translators get
done.)
Christopher R. Hawks
HAWKSoft
-
In my experience, Visual Basic, Access, and Windows complement each other
perfectly - A toy development
---Reply to mail from Alexander R. Pruss about Form has editable field(s) without
Graffiti State Indicator
The latest pilrc gives me a bunch of Form has editable field(s) without
Graffiti State Indicator warnings. Is this something we need to worry
about? Alex
I got annoyed enough, that I
Recent table changes have been promoted to the GTK+ viewer.
Christopher R. Hawks
HAWKSoft
-
DOS: n., A small annoying boot virus that causes random spontaneous system
crashes, usually just before saving a massive
Table cells can now handle multiple images. Better alignment
Christopher R. Hawks
HAWKSoft
-
Windows without the X is like making love without a partner.
win-nt from the people who invented edlin.
Linux, the way to get rid
---Reply to mail from Alexander R. Pruss about transliteration docs
I added a new target to plucker/viewer/xlit/Makefile.in : docs. This
generates html documents that give a quick reference to each
transliteration. One can then generate PluckerDocs from that (though the
Makefile doesn't)
---Reply to mail from Alexander R. Pruss about xlit code
My transliteration code is in the code base. It can be disabled
with --disable-transliteration.
NEAT!
Only one (annoying) problem. The first time I ran Plucker with
transliteration enabled, I had not installed plucker-xlit.pdb. So the
---Reply to mail from Chris Hawks about Table changes
Massive changes to tables.
All calls to SafeMem(Handle/Ptr)New are in try{} blocks.
Improved error handling.
Anchors (and table icons) now displayed as 'normal' links (underline, etc)
Multiple anchors in a table cell.
Multiple sub
---Reply to mail from Alexander R. Pruss about xlit
I think I will implement this the more elegant way. This will also mean
I'll move the currently hard coded Latin1-drop-accents table out of the main
code. Unless, of course, people have better ideas or don't like this.
I like it!
---End
---Reply to mail from Chris Hawks about Table changes
Massive changes to tables.
All calls to SafeMem(Handle/Ptr)New are in try{} blocks.
Improved error handling.
Anchors (and table icons) now displayed as 'normal' links (underline, etc)
Multiple anchors in a table cell.
Multiple sub
Massive changes to tables.
All calls to SafeMem(Handle/Ptr)New are in try{} blocks.
Improved error handling.
Anchors (and table icons) now displayed as 'normal' links (underline, etc)
Multiple anchors in a table cell.
Multiple sub-tables in a table cell.
Improved align handling.
Appears to have
---Reply to mail from Michael Nordstrom about Unit test
Wouldn't it be better to just put the ifdefs *inside* the real tests?
Then the tests will still be there (and be counted), but they can't
fail,
Done!
---End reply
Christopher R. Hawks
HAWKSoft
I discovered that DoTable() in paragraph.c was drawing tables (or icons)
even if the mode was not DRAW_CHAR. This sometimes caused the table or
icon to be drawn over the toolbar.
Fixed in CVS. Could this be the cause of bugs 1078 and 942 ??
Christopher R. Hawks
HAWKSoft
---Reply to mail from Michael Nordstrom about Unit test
On Thu, Feb 19, 2004, Chris Hawks wrote:
Done! Now my only problem is my lack of 5-way/Jogdial support.
You can add the ifdefs to the test files; kind of useless to try to
test something that isn't available on the device ;-)
Done
---Reply to mail from Michael Nordstrom about Unit test
I agree that mock_{VFS,Feature}.h should include VFSMgr.h and that
paragraph.c should include mock_Header.h to override the WinDrawChars
function. You can add those changes to cvs.
Done! Now my only problem is my lack of 5-way/Jogdial
I ran across a site with lists in tables.
So I added support for lists (ordered and unordered) to tables.
Previously the text allran together and the 'bullets' were rendered
outside the table.
Christopher R. Hawks
HAWKSoft
I just commited the changes to PluckerDocs.py to fix this.
Can all who were able to reproduce this bug (I can't) test it and report back??
Christopher R. Hawks
HAWKSoft
-
Who is General Failure and why is he reading my
Mike:
I had some problems compiling the unit test. Mostly needed to add
#include VFSMgr.h to a few files. Also, paragraphtester fails after
printing to the screen.
What am I doing wrong??
Christopher R. Hawks
HAWKSoft
-
While (unsuccessfully) banging on other bugs, I took a look at bug 396.
Actually Agrajag has the right idea. We check to see if the current length
plus the new paragraph is greater than 3 (line 1452). If it is (and it
is) we force the addition of a continue paragraph. In the description we
---Reply to mail from Michael Nordstrom about Bug 942
It shouldn't be called before FrmDrawForm. Move the following code
to the line before the WinGetDrawWindow call and everything will be
OK,
backButtonIndex = FrmGetObjectIndex( fsForm, frmFullscreenBack );
FrmHideObject( fsForm,
Mike:
The failure (with Debug ROM) to return to the previous form (in your
bugnote) is caused by the System queuing a frmUpdateEvent (regular ROMs
don't) and then Plucker crashes trying to redisplay the fullscreen table
that was just released. (Happens with fullscreen images too.)
Is the
---Reply to mail from Chris Hawks about Leakage
Found and fixed dynamic memory leaks when tables are not in Cache.
Same for images (a bit more work!).
---End reply
Christopher R. Hawks
HAWKSoft
-
Windows '98 is so
---Reply to mail from Michael Nordstrom about NULL handles
On Fri, Jan 30, 2004, Chris Hawks wrote:
Should I also commit my 'Limit Cache size' changes??
Sure. Maybe you could also add a comment to the bug case(s) related
to the cache problem?
Done!
---End reply
Christopher R. Hawks
---Reply to mail from Michael Nordstrom about Displaying Plucker files from another
application?
On Tue, Jan 27, 2004, Chris Hawks wrote:
You could file a feature request and ask that an exit button be added.
There shouldn't be an exit button; that's not the Zen of Palm.
Then how
---Reply to mail from [EMAIL PROTECTED] about plucker-dev digest, Vol 1 #549 - 9 msgs
Put yourself in the mindset of a typical (not us, typical) web-browser
user. They long ago lost track of where they are exactly in any given
site-tree. But back takes them to previous options. Hit it too
---Reply to mail from Gerda Shank about Displaying Plucker files from another
application?
So it seems like the most attractive two options are
1) support for sysAppLaunchOpenDB plus some kind of exit (TBD)
2) stripped down reader
Plucker already supports sysAppLaunchOpenDB (it's used
---Reply to mail from Alexander R. Pruss about Displaying Plucker files from another
application?
On Tue, 27 Jan 2004, Chris Hawks wrote:
You could file a feature request and ask that an exit button be added.
I don't know where it would be put. Perhaps the best option would be to
make
---Reply to mail from Alexander R. Pruss about fs table
I've been looking at the fullscreen code.
1. Are anchors supposed to work in fullscreen tables?
Yes, the do for meG.
2. FsAdjust...() calls CopyTableAnchors() but MoveImage() does not. Is this
right?
Yes, CopyTableAnchors()
Plucker Plugin support is now in CVS (fingers crossed). Probably the first
major change will be splitting the language strings so they can be
translated.
Christopher R. Hawks
HAWKSoft
-
In any business, the customer is
With all the problems apparently from the cache getting too large, I took
a look at constraining the cache size. Attached is a diff for your perusal
and testing.
If the storage free space is too small, cache entries are removed until
the free space is available. If no more cache entries remain,
---Reply to mail from Alexander R. Pruss about Cache
On Fri, 23 Jan 2004, Chris Hawks wrote:
If the storage free space is too small, cache entries are removed until
the free space is available.
Oldest first, I assume? Or largest first? Or largest, but not very
recent, first?
Oldest
If we have to (and I'm not saying we do). Perhaps the best way to handle
exact Anchors is to add a new function code 0x0E (Anchor + 6).
The parser would (if turned on) output an exact anchor and a named anchor,
then the text (or image or table) and an end anchor. The old viewers would
ignore the
The Plucker Plugin Interface now allows you to select a 'custom' plugin.
You specify a program name, copy to clipboard, 'desk accessory', and the
launch command. This will allow use of plugins that are not (yet)
supported or custom plugins (like the one Alex mentioned).
I also changed the help
---Reply to mail from Alexander R. Pruss about PPI
Can you save custom plugin info? Or maybe have a pop-up list of the last
ten custom plugins, like on the search form?
So far it only saves one custom plugin.
---End reply
Christopher R. Hawks
HAWKSoft
---Reply to mail from David A. Desrosiers about Plucker Plugin Interface
Robert first:
If you'd like to try it, untar the attached (ppi.tgz) in the viewer
directory. You should be able to just install ppi_en.prc and then load a
modified viewer_XX.prc. The only modification is to change
---Reply to mail from Alexander R. Pruss about Plucker Plugin Interface
On the other hand, it'd be handy to be able to write one's own plugins for
private use and not have to have them be patches against the PPI code. (For
instance, I'd like to have a plugin that strips accents from an input
The Plucker Plugin Interface (PPI) is up and running! All that remains is
stiching it into configure.in and changing 'Look up in RoadLingua' to
'Look up in PPI' in the 'lang/*.rcp' files.
Christopher R. Hawks
HAWKSoft
-
No,
---Reply to mail from Matsumoto, Masakazu about Filtering in the parser
I saw the newest parser code. I wonder why filtering is limited to text
files. If it is applicable to any content types, people can prepare
pdf/word/excel/anything-to-html filter for their purposes.
That can be
The python parser not supports filtering documents through an external
program.
Use '--filter=foobar' on the command line or 'filter = foobar -x -y' in
.pluckerrrc/plucker.ini.
Christopher R. Hawks
HAWKSoft
-
Anybody that
---Reply to mail from Matsumoto, Masakazu about Japanese code converter kcc
Thank you. It works well. It is nice to have a user-defined filter setting.
For plucker-desktop users (who do not want to customize configuration
files by text editor or to install extra commands == most Windos
---Reply to mail from Michael Nordstrom about Japanese code converter kcc
On Mon, Jan 05, 2004, Chris Hawks wrote:
Cool! If that works, I can commit the changes to Spider.py
(after the next release Mike??)
I'm not the parser maintainer, i.e. I don't decide what and when
something
---Reply to mail from Matsumoto, Masakazu about Japanese code converter kcc
Alexander R. Pruss wrote:
I've been asking for a long time for a way to plug in a filter (e.g., a sed
filter or something). That would do the job.
Sounds like you both want something like this:
Right before
---Reply to mail from Jewett, Jim J about RoadLingua dictionaries
Michael Nordstrom
On Wed, Dec 31, 2003, Alexander R. Pruss wrote:
However, there ARE free dictionaries for RL and
there is a free tool for creating more.
... the dictionaries uses a different ZIP format (I can
only
---Reply to mail from Jewett, Jim J about PalmCUnit specifics
I've read the discussion so far, and I've looked in CVS.
The relevant stuff seems to be in viewer/test, but I
still have several questions.
Is it possible to run the tests on a device which
has a normal plucker installed? If
I hope I don't need to write test cases, but, Alex's screen redraw fix
broke anchors in tables. I had to set tContext.writeMode =
WRITEMODE_LAYOUT_ANCHORS in the 'fake' textContext I was passing to the
anchor functions.
I also added 'lookup.bmp' to viewer.rcp.in for those of us without hires
and
---Reply to mail from Michael Nordstrom about Tables + linked list (was: Test cases)
On Thu, Jan 01, 2004, Chris Hawks wrote:
About tables; why is there a linked list called tableIndexList in
uncompress.c that nothing is ever added to? Is it only a matter of
a missing ListAppend call
Well, after quite a bit of work, I've added some minor features,
cleaned-up the code and brought all 3 closer to the Plucker coding
guidelines.
The Posix Viewer and Explode now show tables and multiimages (Explode even
renders multiimages back into a single jpeg.
Enjoy!
Christopher R. Hawks
Explode and the posix viewer now handle tables and multiimages.
(there are a few changes to unpluck as well.)
Christopher R. Hawks
HAWKSoft
-
...[Linux's] capacity to talk via any medium except smoke signals.
-- Dr. Greg
---Reply to mail from hugh about building the viewer
i decided to try to co and build plucker. but i cant get the viewer to
build. it appears that config.h is not being generated. i'm getting just
the comment at the top only in that file. so i suppose there should be
other stuff (ie where
The GTK+2 viewer now does tables!!
Christopher R. Hawks
HAWKSoft
-
We are using Linux daily to UP our productivity - so UP yours!
-- Adapted from Pat Paulsen by Joe Sloan
---Reply to mail from Ruben Malchow about Subject: Bells and whistles?
a.) the table tr-gap thing (which is still not working, but i'll check
out from cvs again and have a look)
[...]
once the table thing is working, almost anything can be done,
even if one had to use images for some things
---Reply to mail from Ruben Malchow about plucker-dev digest, Vol 1 #490 - 17 msgs
i just checked out the cvs tree (from
pserver:[EMAIL PROTECTED]:/cvs/plucker) and built the viewer from
there - and it still produces that gap. am i using the wrong tree or
missing out on something?
Hmmm.
---Reply to mail from Ruben Malchow about tables / roadmap question
a.) say we have a table with two tr's and two td's each. if each cell
contains an image, the two adjacent columns are rendered with no space
inbetween, while between the rows, there is a large gap. why is this? can
this be
I just added support for large image links and multi-images to the
GTK+2-POSIX viewer. I also updated unpluck.h with missing document types
and function codes.
Christopher R. Hawks
-
Strange costumes, weird behavior - I
---Reply to mail from Alexander R. Pruss about rotation
I just committed a faster version of rotatebitmap.c. Under PACE with
armlets disabled, on my NX, it shows a 100%+ speedup for 16bit rotation.
The armlet version generally shows about a 20% speedup.
I am curious how it performs on real
---Reply to mail from Alexander R. Pruss about rotation
From: Chris Hawks [EMAIL PROTECTED]
Lots more accurate. 23.61 sec as opposed to 28.25 sec. About a 16%
increase in speed.
Oh well, not much then.
What bit depth was that at? (The 1-bit was not optimized as heavily as the
8
1 - 100 of 269 matches
Mail list logo
