headers
hbl | 1 Aug 2008 02:44
Picon
Gravatar

Re: enblend solution build fails


Thanks for the reply, Guido.  I am trying to sort through it but it is
tough going not being familiar with the code.  I put in a support
request on the project tracker.  Hopefully Andrew will step in with a
solution.

On Jul 31, 4:35 pm, Guido Kohlmeyer <d... <at> gekko-design.de> wrote:
> I can confirm this error on Win32, MSVC 9.0.
> It seems to be a problem of friendship specification between classes.
> The MSVC compiler needs to know the class. Unfortunately I'm not
> familiar with the code or the template class specifications.
> Maybe Andrew Mihal can clearify the case, because he introduced the
> changes in the latest CVS revision, which yields to the failed build on
> MSVC.
>
> Guido
>
> hbl schrieb:
>
>
>
> > bump
>
> > On Jul 27, 4:18 pm, hbl <hblar... <at> comcast.net> wrote:
> >> Building the latest enblend_msvc2008, fails in enfuse at the following
> >> line:
>
> >> class NotifyingDirectionSelector<UnstridedArrayTag, T, Notify>
> >> {
> >>         friend class Notify::self_type;  //error C2242
(Continue reading)

Permalink | Reply | 4 comments |
headers
Fahim Mannan | 2 Aug 2008 01:26
Picon

enblend solution build fails


Hi,

Using 'typename' instead of 'class' should fix the problem. Ie.

friend typename Notify::self_type;

-Fahim

hbl wrote:
> Thanks for the reply, Guido.  I am trying to sort through it but it is
> tough going not being familiar with the code.  I put in a support
> request on the project tracker.  Hopefully Andrew will step in with a
> solution.
>
> On Jul 31, 4:35 pm, Guido Kohlmeyer <d... <at> gekko-design.de> wrote:
>   
>> I can confirm this error on Win32, MSVC 9.0.
>> It seems to be a problem of friendship specification between classes.
>> The MSVC compiler needs to know the class. Unfortunately I'm not
>> familiar with the code or the template class specifications.
>> Maybe Andrew Mihal can clearify the case, because he introduced the
>> changes in the latest CVS revision, which yields to the failed build on
>> MSVC.
>>
>> Guido
>>
>> hbl schrieb:
>>
>>
(Continue reading)

Permalink | Reply | 3 comments |
headers
Marko Kuder | 2 Aug 2008 18:51
Picon

Re: GSoC - Batch Processing


This week's report (short and simple):
 - pause function finished, works on windows and linux
 - added some buttons (reset project, reset all, clear batch) and
icons (add/remove project)
 - output extension column fixed
 - put the button panel in a sizer, so it should look better
 - added "overwrite always" and "shutdown when done" checkboxes
 - added send to batch and open batch processor to menu in Hugin
 - removing/reseting was buggy while project were running, i'm making
that nice and robust now, it will be commited tommorow
After tommorow's commit I'll start working on the autosave/save to
file part. We should have quite a nice and working program by next
weekend.

--
Marko
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups "hugin and other free panoramic
software" group.
A list of frequently asked questions is available at: http://wiki.panotools.org/Hugin_FAQ
To post to this group, send email to hugin-ptx <at> googlegroups.com
To unsubscribe from this group, send email to hugin-ptx-unsubscribe <at> googlegroups.com
For more options, visit this group at http://groups.google.com/group/hugin-ptx
-~----------~----~----~----~------~----~------~--~---
Permalink | Reply | 4 comments |
headers
Zoran Mesec | 3 Aug 2008 00:56
Picon

Re: last chance!! panotools meeting, Monday - Friday!


Ohh my....why do I read this for the first time now! I would like to
join you guys, but timing is just not right for me, because next week
is an important week for GSOC Batch Processing and my presence is
required.

I often browse panotools wiki, but not the panotools site, so I must
have missed this. Sad thing is, that I am planning a trip to Prague in
late August, after GSOC pencil down time on August 18th but not now.

cheers,
Zoran Mesec

On Aug 1, 7:38 pm, Bruno Postle <br... <at> postle.net> wrote:
> ----- Forwarded message from Jeffrey Martin <360cit... <at> gmail.com> -----
>
> Hi Everyone,
> (sorry for the cross-post)
>
> It's time again for another Panotools Meeting! As you probably know it kicks
> off with an "opening dinner" this sunday August 3. The talks go from August
> 4 - 6.
>
> Anyone who still wants to come is welcome. There are certainly hotel rooms
> still available in Prague. [snip]
>
> here is the official website for the panotools meeting:www.panotools-meeting.com
>
> here is the schedule of talks:www.tinyurl.com/panotools2008talks
> here is the calendar of the other events:www.tinyurl.com/panotools2008calendar
(Continue reading)

Permalink | Reply | 0 comments |
headers
Guido Kohlmeyer | 3 Aug 2008 01:33
Picon

Re: enblend solution build fails


Dear Fahim,

Thank you for this hint. After small modification as suggested the build 
processed finished successfully.

Guido

Fahim Mannan schrieb:
> Hi,
> 
> Using 'typename' instead of 'class' should fix the problem. Ie.
> 
> friend typename Notify::self_type;
> 
> 
> -Fahim
> 
> hbl wrote:
>> Thanks for the reply, Guido.  I am trying to sort through it but it is
>> tough going not being familiar with the code.  I put in a support
>> request on the project tracker.  Hopefully Andrew will step in with a
>> solution.
>>
>> On Jul 31, 4:35 pm, Guido Kohlmeyer <d... <at> gekko-design.de> wrote:
>>   
>>> I can confirm this error on Win32, MSVC 9.0.
>>> It seems to be a problem of friendship specification between classes.
>>> The MSVC compiler needs to know the class. Unfortunately I'm not
>>> familiar with the code or the template class specifications.
(Continue reading)

Permalink | Reply | 2 comments |
headers
Dane | 3 Aug 2008 01:50
Picon

Re: Layered TIFF files and how do you save them?


Another way is to stitch the images with an alpha version with a
enblended output (enblend has significantly improved since the version
with the beta), then without changing anything, reopen the file in the
old beta version and use PTstitcher to produce a layered .psd file.
Open this file, add a layer on top, open the enblended tiff and drag
it to the new layer. Since the scaling and rotation have not changed,
moving it to the correct location at one point will insure correct
alignment. Then add a mask to the layer and brush on black to the mask
on anything you would rather have the layer below to show thru.

The PTsticher seems to work fine with a dozen layers on my XP machine,
but not on my vista laptop.

On Jul 6, 5:25 pm, Yuval Levy <goo... <at> levy.ch> wrote:
> Ken Warner wrote:
> > layered TIFF files are the
> > actual images that are stitched together for the final pano.
>
> kind of. hugin produces "Remapped images". These may (or may not,
> depending on your workflow) be blended automatically into a single
> panorama or manually.
>
> > If so, is there a way to force Hugin to just produce the layered
> > TIFF files and then stop?
>
> Hugin can produce the layers. Just tick the box "Remapeed images" under
> the header "Normal" (I would prefer "Single Exposure", but we are in a
> string freeze) and Hugin will produce individual TIFF files, one for
> each image.
(Continue reading)

Permalink | Reply | 0 comments |
headers
Guido Kohlmeyer | 3 Aug 2008 01:18
Picon

[Win32] hugin_0.7_svn3257_20080802 for download


For testing a current build of hugin can be found on
http://hugin.panotools.org/testing/hugin/hugin_0.7_svn3257_20080802_win32.exe 

The file is a self-extracting archive, no installer (see below)!

Autopano if needed:
http://hugin.panotools.org/testing/hugin/autopano-sift-C_svn3223_20080730_win32.exe 

Revision information
====================
Hugin ........ : SVN 3257
Enblend/Enfuse : CVS update on 2008-08-02 (patched for MSVC)
Autopano-sift-C: SVN 3223

Contents
====================
The hugin archive contains the hugin build folder INSTALL/FILES
/bin
/doc
/share

How to install
====================
1) Simply execute the exe file
2) Enter a destination path in dialog and extract
3) In destination path some additional files are stored to create
desktop icon and file association (see README-HuginTestInstall.txt,
thanks to Tom). Otherwise run only hugin.exe.
(Continue reading)

Permalink | Reply | 12 comments |
headers
Yuval Levy | 3 Aug 2008 01:40
Picon
Favicon

Re: enblend solution build fails

Confirm here on MSVC 8 EE - works like a charm. Patch attached. Thanks 
Fahim!

Guido Kohlmeyer wrote:
> Dear Fahim,
> 
> Thank you for this hint. After small modification as suggested the build 
> processed finished successfully.
> 
> Guido
> 
> Fahim Mannan schrieb:
>> Hi,
>>
>> Using 'typename' instead of 'class' should fix the problem. Ie.
>>
>> friend typename Notify::self_type;
>>
>>
>> -Fahim
>>
>> hbl wrote:
>>> Thanks for the reply, Guido.  I am trying to sort through it but it is
>>> tough going not being familiar with the code.  I put in a support
>>> request on the project tracker.  Hopefully Andrew will step in with a
>>> solution.
>>>
>>> On Jul 31, 4:35 pm, Guido Kohlmeyer <d... <at> gekko-design.de> wrote:
>>>   
>>>> I can confirm this error on Win32, MSVC 9.0.
(Continue reading)

Permalink | Reply | 1 comment |
headers
Tom Sharpless | 3 Aug 2008 04:41
Picon

Re: Parallax correction, image rotation and photo position


On the other hand, Aseem Agarwala does work for Adobe Systems and
wrote the new image blending codes for PS CS3.  So it is quite likely
that the only way to get your hands on this brilliant s/w will be to
wait until Adobe is ready to sell it to you for big $$$.  ==(sigh)==

-- Tom

On Jul 29, 5:44 pm, Errol <kiw... <at> thompsonz.net> wrote:
> > This work is done from a paper that was submitted at a
> > university.  I don't know the legality of the
> > copyright/patent of the process but the paper might be an
> > interesting read for those interested in this who have a
> > programming or math background.  The paper is on the website,
> > and gives a detailed explanation of what is going on.
>
> > Aseem Agarwala, Maneesh Agrawala, Michael Cohen, David
> > Salesin, Richard Szeliski. Photographing long scenes with
> > multi-viewpoint panoramas. ACM Transactions on Graphics
> > (Proceedings of SIGGRAPH 2006), 2006.
> >http://grail.cs.washington.edu/projects/multipano/agarwala_sig06.pdf
>
> This is a published academic work presented at a conference. In that sense
> you are free to access it an use the material contained in it. However, you
> should acknowledge the paper in any work that is derived from it. I would
> hope that you were doing this anyway. The value of an academic work is more
> from the number of citations rather than having it published.
>
> If they were endeavouring to patent some feature of the work then there
> should be some indication in the paper that this was occuring or they may
(Continue reading)

Permalink | Reply | 0 comments |
headers
Tom Sharpless | 3 Aug 2008 04:47
Picon

Re: Parallax correction, image rotation and photo position


PS -- from his CV

PATENTS
Aseem Agarwala, Maneesh Agrawala, Michael Cohen, David Salesin,
Richard Szeliski.
“Photographing big things,” Patent pending.
Aseem Agarwala. “Gradient-domain compositing,” Patent pending.

-- Tom

On Jul 29, 5:44 pm, Errol <kiw... <at> thompsonz.net> wrote:
> > This work is done from a paper that was submitted at a
> > university.  I don't know the legality of the
> > copyright/patent of the process but the paper might be an
> > interesting read for those interested in this who have a
> > programming or math background.  The paper is on the website,
> > and gives a detailed explanation of what is going on.
>
> > Aseem Agarwala, Maneesh Agrawala, Michael Cohen, David
> > Salesin, Richard Szeliski. Photographing long scenes with
> > multi-viewpoint panoramas. ACM Transactions on Graphics
> > (Proceedings of SIGGRAPH 2006), 2006.
> >http://grail.cs.washington.edu/projects/multipano/agarwala_sig06.pdf
>
> This is a published academic work presented at a conference. In that sense
> you are free to access it an use the material contained in it. However, you
> should acknowledge the paper in any work that is derived from it. I would
> hope that you were doing this anyway. The value of an academic work is more
> from the number of citations rather than having it published.
(Continue reading)

Permalink | Reply | 0 comments |

Gmane