From minc-development@bic.mni.mcgill.ca Sun Dec 1 21:18:58 2002 From: minc-development@bic.mni.mcgill.ca (Peter NEELIN) Date: Sun, 1 Dec 2002 16:18:58 -0500 Subject: [MINC-development] beta mincmorph In-Reply-To: Message-ID: On Thu, 28 Nov 2002, Andrew Janke wrote: > On another somewhat related note, this version appears to be able to generate > files that will crash register once it attempts to load the colourmap (after > loading the file). I assume this is some range problem but I'm currently > stumped as to where it is... Got a sample file? Peter ---- Peter Neelin (neelin@bic.mni.mcgill.ca) From minc-development@bic.mni.mcgill.ca Mon Dec 2 01:57:01 2002 From: minc-development@bic.mni.mcgill.ca (Andrew Janke) Date: Mon, 2 Dec 2002 11:57:01 +1000 Subject: [MINC-development] beta mincmorph In-Reply-To: Message-ID: On Sun, 1 Dec 2002, Peter NEELIN wrote: > > On another somewhat related note, this version appears to be able to generate > > files that will crash register once it attempts to load the colourmap (after > > loading the file). I assume this is some range problem but I'm currently > > stumped as to where it is... > > Got a sample file? http://www.cmr.uq.edu.au/~rotor/graagh/in.mnc.gz http://www.cmr.uq.edu.au/~rotor/graagh/out.mnc.gz It could however be just my version of register.... Created as such: $ mincmorph -clobber -byte -successive 'B[6000:25000]G' in.mnc out.mnc -- Andrew Janke ( rotor@cmr.uq.edu.au || www.cmr.uq.edu.au/~rotor ) Australia->University of Queensland->Centre for Magnetic Resonance Work: +61 7 3365 4100 || Home: +61 7 3800 4042 From minc-development@bic.mni.mcgill.ca Tue Dec 3 08:07:09 2002 From: minc-development@bic.mni.mcgill.ca (Andrew Janke) Date: Tue, 3 Dec 2002 18:07:09 +1000 Subject: [MINC-development] Re: reviews are a lottery In-Reply-To: <20021202182409.E1299124@shadow.bic.mni.mcgill.ca> Message-ID: On Mon, 2 Dec 2002, Steve ROBBINS wrote: > By the by, I really like mincpik -- it beats my scripts with > hardcoded ICBM image sizes. > > But I'm partial to the hotmetal colourmap. And sometimes other > colourmaps, so what do you think about this patch? It lets me > do things like > > mincpik -lookup -hotmetal foo.mnc bar.png > or mincpik -lookup '-discrete -lookup red-blue.lut' foo.mnc bar.png The patch looks fine, I have added major portions of it to the current version of mincpik in CVS. /s/s/minc_dev/mincpik I have been working on a much simplified version of minpik (I wrote most of that perl yonks ago when I knew not much of MINC, as such most of it was a perlerised version of mincview) I've added a few extra bits, WRT handling of the -image_range option when a lookup-table is specified, I assume this is what you'd expect such a combination of options to do.. ie: $ mincpik -lookup -hotmetal -image_range 0 100 foo.mnc | display - passed a -range option to minclookup. without a -lookup option, mincextract handles this option. Although now that I think about it a bit more it'd be far simpler to let mincreshape handle this option... Hrm.. next time perhaps.. it'll make for shorter code! Grrgh, just when I thought I was finished.. :) Let us know if this version behaves itself as you think it should, if not I'm happy to make changes to bring it inline with whatever you have already. -- Andrew Janke ( rotor@cmr.uq.edu.au || www.cmr.uq.edu.au/~rotor ) Australia->University of Queensland->Centre for Magnetic Resonance Work: +61 7 3365 4100 || Home: +61 7 3800 4042 From minc-development@bic.mni.mcgill.ca Tue Dec 3 08:15:33 2002 From: minc-development@bic.mni.mcgill.ca (Andrew Janke) Date: Tue, 3 Dec 2002 18:15:33 +1000 Subject: [MINC-development] Re: reviews are a lottery In-Reply-To: Message-ID: On Tue, 3 Dec 2002, Andrew Janke wrote: > I've added a few extra bits, WRT handling of the -image_range option when a > lookup-table is specified, I assume this is what you'd expect such a combination > of options to do.. ie: > > $ mincpik -lookup -hotmetal -image_range 0 100 foo.mnc | display - > > passed a -range option to minclookup. without a -lookup option, mincextract > handles this option. Although now that I think about it a bit more it'd be far > simpler to let mincreshape handle this option... Hrm.. next time perhaps.. > it'll make for shorter code! > > Grrgh, just when I thought I was finished.. :) Couldn't help myself and had to fix it.. I have a strange penchant for shorter less "iffy" code. The version in CVS has been updated (again). Same comments apply. -- Andrew Janke ( rotor@cmr.uq.edu.au || www.cmr.uq.edu.au/~rotor ) Australia->University of Queensland->Centre for Magnetic Resonance Work: +61 7 3365 4100 || Home: +61 7 3800 4042 From minc-development@bic.mni.mcgill.ca Wed Dec 4 04:32:18 2002 From: minc-development@bic.mni.mcgill.ca (Steve ROBBINS) Date: Tue, 3 Dec 2002 23:32:18 -0500 Subject: [MINC-development] beta mincmorph In-Reply-To: ; from rotor@cmr.uq.edu.au on Thu, Nov 28, 2002 at 12:39:05PM +1000 References: Message-ID: <20021203233218.A1489688@shadow.bic.mni.mcgill.ca> Howdy, On Thu, Nov 28, 2002 at 12:39:05PM +1000, Andrew Janke wrote: > If you look at the code of the Group option (Connected component labelling > -- in kernel_ops.c) you'll note I haven't used the classical approach to > resolving equivalences. (Building an array of equivalences as you go with > redundacies). Instead I've used an approach that I can't find anywhere > else... but I'm sure has been described before. The reason for using it is > that i can guarantee memory usage will only grow to a finite size this way, > something I can't ensure with the classical version especially in 256^3 > volumes. I'm not familiar with this stuff, so please forgive a naive question. The problem you're solving is the one described here: http://www.dai.ed.ac.uk/HIPR2/label.htm ? If so, why wouldn't you use parent pointer trees? They are trivial to implement, will use storage linear in the # of voxels. With union-by-rank and path compression, the running time is basically linear too. -S From minc-development@bic.mni.mcgill.ca Wed Dec 4 04:40:04 2002 From: minc-development@bic.mni.mcgill.ca (Steve ROBBINS) Date: Tue, 3 Dec 2002 23:40:04 -0500 Subject: [MINC-development] Re: reviews are a lottery In-Reply-To: ; from rotor@cmr.uq.edu.au on Tue, Dec 03, 2002 at 06:07:09PM +1000 References: <20021202182409.E1299124@shadow.bic.mni.mcgill.ca> Message-ID: <20021203234003.B1489688@shadow.bic.mni.mcgill.ca> On Tue, Dec 03, 2002 at 06:07:09PM +1000, Andrew Janke wrote: > Let us know if this version behaves itself as you think it should, if not I'm > happy to make changes to bring it inline with whatever you have already. Keen! The new version does what I expected for both my test cases. ;-) By the way I also snarfed your version of "minchistory" and munged it almost beyond recognition, introducing a dependency on Text::Format along the way ... -S #!/usr/bin/env perl # # Andrew Janke - rotor@cmr.uq.edu.au # Simple script to dump history information from a minc file # # Thu Feb 3 14:00:26 EST 2000 - initial version # Wed Oct 18 10:01:17 EST 2000 - rewrite and removed -clobber ommision # Tue Dec 3 23:38:15 EST 2002 - rewritten to pretty-print entries use Text::Format; chomp($me = `basename $0`); if ($#ARGV < 0) { die "Usage: $me \n"; } $text = new Text::Format( firstIndent => 0, bodyIndent => 5, columns => 78 ); foreach (@ARGV) { process_file($_); } sub process_file { my $fname = shift; if (!-e $fname){ warn "$me: Couldn't find $fname\n"; return; } print "--- History of $fname ---\n\n"; $counter = 1; foreach (`mincinfo -attvalue :history $fname`){ chomp; next if $_ eq ''; printf "[%02d] ", $counter ++; print $text->format($_); } print "\n"; } From minc-development@bic.mni.mcgill.ca Mon Dec 2 22:08:09 2002 From: minc-development@bic.mni.mcgill.ca (Andrew Janke) Date: Tue, 3 Dec 2002 08:08:09 +1000 Subject: [MINC-development] beta mincmorph In-Reply-To: <20021203233218.A1489688@shadow.bic.mni.mcgill.ca> Message-ID: On Tue, 3 Dec 2002, Steve ROBBINS wrote: > > If you look at the code of the Group option (Connected component labelling > > -- in kernel_ops.c) you'll note I haven't used the classical approach to > > resolving equivalences. (Building an array of equivalences as you go with > > redundacies). Instead I've used an approach that I can't find anywhere > > else... but I'm sure has been described before. The reason for using it is > > that i can guarantee memory usage will only grow to a finite size this way, > > something I can't ensure with the classical version especially in 256^3 > > volumes. > > I'm not familiar with this stuff, so please forgive a naive question. > The problem you're solving is the one described here: > http://www.dai.ed.ac.uk/HIPR2/label.htm albeit with an arbitrary "connection kernel" or structuring element. Thus allowing us to do 2d/3d 8-connected/4-connected CCL'ing. The std thing to do (in 2d) is either 4 or 8 connectivity, mincmorph is written in such a way with arbitrary kernels such that (i think) one cannot be sure that already visited neighbouring voxels will have only 2 different values. (real ugly to think about i know) ie: std 2D connectivity kernels where 0 is the current voxel: (use fixed width font M$laves!) 4 connectivity 8 connectivity X XXX X0X X0X X XXX To think of an "abberant case" try this structuring element on for size... ;) "wierd connectivity" X X X X 0 X X X X In this case I think you can get the case whereby there are more than 2 distinct neighbour values... > If so, why wouldn't you use parent pointer trees? They are trivial to > implement, will use storage linear in the # of voxels. With union-by-rank and > path compression, the running time is basically linear too. As in: http://home.swbell.net/mck9/cobol/tech/tree.html? I can't quite see how the storage would be linear and how redundant connections would be resolved? or can each node have more than 2 kiddies? At then end I assume this tree would be then traversed (depth first) and the top value assigned to each value in the tree? I think i am sort of doing this in my current implementation although doing it all in a single 1D array! :) a From minc-development@bic.mni.mcgill.ca Mon Dec 2 22:16:03 2002 From: minc-development@bic.mni.mcgill.ca (Andrew Janke) Date: Tue, 3 Dec 2002 08:16:03 +1000 Subject: [MINC-development] Re: reviews are a lottery In-Reply-To: <20021203234003.B1489688@shadow.bic.mni.mcgill.ca> Message-ID: On Tue, 3 Dec 2002, Steve ROBBINS wrote: > By the way I also snarfed your version of "minchistory" and munged it almost > beyond recognition, introducing a dependency on Text::Format along the way ... > Valid! Time to stuff your version in CVS I think... :) -- Andrew Janke ( rotor@cmr.uq.edu.au || www.cmr.uq.edu.au/~rotor ) Australia->University of Queensland->Centre for Magnetic Resonance Work: +61 7 3365 4100 || Home: +61 7 3800 4042 From minc-development@bic.mni.mcgill.ca Wed Dec 4 05:09:07 2002 From: minc-development@bic.mni.mcgill.ca (Steve ROBBINS) Date: Wed, 4 Dec 2002 00:09:07 -0500 Subject: [MINC-development] beta mincmorph In-Reply-To: ; from rotor@cmr.uq.edu.au on Tue, Dec 03, 2002 at 08:08:09AM +1000 References: <20021203233218.A1489688@shadow.bic.mni.mcgill.ca> Message-ID: <20021204000907.C1489688@shadow.bic.mni.mcgill.ca> On Tue, Dec 03, 2002 at 08:08:09AM +1000, Andrew Janke wrote: > On Tue, 3 Dec 2002, Steve ROBBINS wrote: > > If so, why wouldn't you use parent pointer trees? They are trivial to > > implement, will use storage linear in the # of voxels. With union-by-rank and > > path compression, the running time is basically linear too. > > As in: http://home.swbell.net/mck9/cobol/tech/tree.html? No, more like http://www.cs.mcgill.ca/~cs251/OldCourses/1997/topic24/ Cormen, Leiserson & Rivest have a fantastic discussion about this topic in the chapter "Data Structures for Disjoint Sets". > I can't quite see how the storage would be linear and how redundant > connections would be resolved? or can each node have more than 2 kiddies? Yeah -- you can have as many children as you want. The storage is linear because you store ONLY the parent pointer. > At then end I assume this tree would be then traversed (depth first) and the > top value assigned to each value in the tree? Each group is represented by a tree in the forest. You find out which group you are in by traversing parent pointers until you get to the root. The root's parent points to itself. When you use path compression and union-by-rank, the set operations ("union" and "find-set") become really really fast: amortized constant time, essentially. -S From minc-development@bic.mni.mcgill.ca Mon Dec 2 22:35:34 2002 From: minc-development@bic.mni.mcgill.ca (Andrew Janke) Date: Tue, 3 Dec 2002 08:35:34 +1000 Subject: [MINC-development] Re: reviews are a lottery In-Reply-To: Message-ID: On Tue, 3 Dec 2002, Andrew Janke wrote: > > By the way I also snarfed your version of "minchistory" and munged it almost > > beyond recognition, introducing a dependency on Text::Format along the way ... > > > > > Valid! Time to stuff your version in CVS I think... :) On this note, whilst I was at it I have also added "minccomplete". This a quick and dirty perl script that I use to check as to whether processing on a minc file is "complete". The most common use of this is in scripts that check the output of pipelines. ie: the file is often there and allocated but only half of the data has been written into it. a From minc-development@bic.mni.mcgill.ca Mon Dec 2 22:46:58 2002 From: minc-development@bic.mni.mcgill.ca (Andrew Janke) Date: Tue, 3 Dec 2002 08:46:58 +1000 Subject: [MINC-development] beta mincmorph In-Reply-To: <20021204000907.C1489688@shadow.bic.mni.mcgill.ca> Message-ID: On Wed, 4 Dec 2002, Steve ROBBINS wrote: > No, more like http://www.cs.mcgill.ca/~cs251/OldCourses/1997/topic24/ > > Cormen, Leiserson & Rivest have a fantastic discussion about this topic in the > chapter "Data Structures for Disjoint Sets". ah, I see. This would appear to be the bee's knees. > > At then end I assume this tree would be then traversed (depth first) and the > > top value assigned to each value in the tree? > > Each group is represented by a tree in the forest. You find out which group > you are in by traversing parent pointers until you get to the root. The > root's parent points to itself. > > When you use path compression and union-by-rank, the set operations ("union" > and "find-set") become really really fast: amortized constant time, > essentially. If you can manage to read the code this is essentially what I do (without knowwing it). I assume this approach also attempts to always assign the lowest label to a new element? ie: this sequence of equivalences is not only possible but common in CCL. 3 -> 1 4 -> 3 4 -> 1 3 -> 4 3 -> 1 In which case I think the rule should be to only point larger elments to smaller ones. Of course if you strike a new equivalence for a node in which a equivalence already exists you must traverse this equivalence to it's lowest common denomiator yes? This is essentially what I do. Perhaps the code would be more readable if I implemented such a library though.. :) However it get's a little more complex as during the re-labelling process counters of group sizes must also be aggregated and then groups re-labelled (again) WRT size. Hrm, as I said before I'm sure there is a more efficient (and memory stable) way of doing such things than my current implementation but my brain was beginning to bend so I got lazy at the expense of a few CPU cycles. :) -- Andrew Janke ( rotor@cmr.uq.edu.au || www.cmr.uq.edu.au/~rotor ) Australia->University of Queensland->Centre for Magnetic Resonance Work: +61 7 3365 4100 || Home: +61 7 3800 4042 From minc-development@bic.mni.mcgill.ca Wed Dec 4 05:35:14 2002 From: minc-development@bic.mni.mcgill.ca (Steve ROBBINS) Date: Wed, 4 Dec 2002 00:35:14 -0500 Subject: [MINC-development] beta mincmorph In-Reply-To: ; from rotor@cmr.uq.edu.au on Tue, Dec 03, 2002 at 08:46:58AM +1000 References: <20021204000907.C1489688@shadow.bic.mni.mcgill.ca> Message-ID: <20021204003514.E1489688@shadow.bic.mni.mcgill.ca> On Tue, Dec 03, 2002 at 08:46:58AM +1000, Andrew Janke wrote: > This is essentially what I do. Perhaps the code would be more readable if I > implemented such a library though.. :) I'd offer you my implementation of disjoint sets, but it's in C++ pthhhht! ;-) -S From minc-development@bic.mni.mcgill.ca Mon Dec 2 23:02:04 2002 From: minc-development@bic.mni.mcgill.ca (Andrew Janke) Date: Tue, 3 Dec 2002 09:02:04 +1000 Subject: [MINC-development] beta mincmorph In-Reply-To: <20021204003514.E1489688@shadow.bic.mni.mcgill.ca> Message-ID: On Wed, 4 Dec 2002, Steve ROBBINS wrote: > > This is essentially what I do. Perhaps the code would be more readable if I > > implemented such a library though.. :) > > I'd offer you my implementation of disjoint sets, but it's in C++ pthhhht! > ;-) Is it the array, list or tree version? If the latter send it on anyway... a From minc-development@bic.mni.mcgill.ca Mon Dec 9 13:53:26 2002 From: minc-development@bic.mni.mcgill.ca (Steve ROBBINS) Date: Mon, 9 Dec 2002 08:53:26 -0500 Subject: [MINC-development] [Fwd: data format] In-Reply-To: <20021122124045.A2052@sickkids.ca>; from jgsled@sickkids.ca on Fri, Nov 22, 2002 at 12:40:45PM -0500 References: <20021122124045.A2052@sickkids.ca> Message-ID: <20021209085326.A1819205@shadow.bic.mni.mcgill.ca> Howdy, Where'd all the discussion go? On Fri, Nov 22, 2002 at 12:40:45PM -0500, John G. Sled wrote: > > This is a message that Art Wetzel had sent describing the data layout > they are using at the Pittsburgh Supercomputing Centre to serve up the > visual human. From what I understand, some of the details are still > in flux, but it does give some insight into their approach. > > cheers, > > John > > > ############################ > The basic element of data going to the client is a cube. Since the > client is in control of the entire interaction every cube is delivered > by explicit client request. For the Vis Female we currently provide > 4 scales of cubes all composed of 8*8*8 = 512 voxels. The scales are > just multiples of 2 so we have full resolution, 1/2 res, 1/4 res and > 1/8 res cubes. The 1/8 res cubes for example enclose what was originally > a full scale 64*64*64 voxel region. Therefore, each of the 512 voxels > in a 1/8 scale cube corresponds to the low pass filtering of an 8*8*8 > voxel region from the full scale data. In going from a low res cube > to the next higher res (by a factor of 2) there are 8 possible octant > cubes to choose from. These subcubes nestle into the corners of the > larger cubes. Pretty cool. It strikes me, though, that they are paying a high price (a complicated on-disk format) for using mmap(). If I were to implement this, my first thought would be to put each resolution into a separate MINC file (e.g. foo.mnc, foo_scale2.mnc, etc). You'd be able to load the coarsest resolution or two into memory as a whole, and seek around in the other files as needed. Thoughts? -S From minc-development@bic.mni.mcgill.ca Mon Dec 9 21:48:17 2002 From: minc-development@bic.mni.mcgill.ca (Andrew Janke) Date: Tue, 10 Dec 2002 07:48:17 +1000 Subject: [MINC-development] [Fwd: data format] In-Reply-To: <20021209085326.A1819205@shadow.bic.mni.mcgill.ca> Message-ID: On Mon, 9 Dec 2002, Steve ROBBINS wrote: > Where'd all the discussion go? This method is in effect MIP-Mapping, something that OpenGL already does well for texture maps. I have fiddled a bit with it in viewnup with a fiar bit of success (zooming in and out of a volume). It appears that netCDF is pretty fast at this sort of thing (jumping through a volume with a step size). I also agree with your comments, a number of MINC files on disk would in effect do this a whole lot more transparently. (and be implementable with mincresample!) a From minc-development@bic.mni.mcgill.ca Wed Dec 11 15:21:54 2002 From: minc-development@bic.mni.mcgill.ca (Steve ROBBINS) Date: Wed, 11 Dec 2002 10:21:54 -0500 Subject: [MINC-development] time for 1.1? Message-ID: <20021211102153.C1431417@shadow.bic.mni.mcgill.ca> Howdy, While adding a function prototype to MINC, I had a look at the changelog. It strikes me that a new release is in order. Maybe we can make it an annual event! User-visible changes since 1.0 include: * Rawtominc has new "-skip" option, to allow skipping header information. * Mincstats option "-max_bins" renamed to "-int_max_bins", to avoid clash with "-max". * Minccalc has new functions: tan, asin, acos, and atan. In addition, there have been some important bug fixes: * Jason fixed some 64-bit issues * John fixed a coredump in mincstats * I fixed a bug in handing inverted concatenated transforms I've also written manpages for each program that lacked one. I'd like to change MINC to use automake, and to normalize the building process a bit, e.g. deprecate setting NETCDF_PREFIX. There are two immediate benefits to using automake. First, it comes with targets for building and checking a source distribution (make dist, make distcheck) which are really really handy. Second, it provides standard targets for regression / unit testing. I wrote a simple regression test when I fixed the transforms bug that I'd like to exercise with "make check". What say you? If there are no objections, I would volunteer to make the change to automake and drive this release forward. (I'd also be more than happy to let someone else be the release manager!) January 16 is the anniversary of MINC 1.0, so I propose that as a target release date. Thoughts? -Steve From minc-development@bic.mni.mcgill.ca Wed Dec 11 15:27:43 2002 From: minc-development@bic.mni.mcgill.ca (Jason Lerch) Date: Wed, 11 Dec 2002 10:27:43 -0500 Subject: [MINC-development] time for 1.1? In-Reply-To: <20021211102153.C1431417@shadow.bic.mni.mcgill.ca>; from stever@bic.mni.mcgill.ca on Wed, Dec 11, 2002 at 10:21:54AM -0500 References: <20021211102153.C1431417@shadow.bic.mni.mcgill.ca> Message-ID: <20021211102742.C2045322@bic.mni.mcgill.ca> > User-visible changes since 1.0 include: > > * Rawtominc has new "-skip" option, to allow skipping header information. > > * Mincstats option "-max_bins" renamed to "-int_max_bins", to avoid clash > with "-max". > > * Minccalc has new functions: tan, asin, acos, and atan. > And don't forget the rpm/tardist/deb packaging that is now in place as well. > If there are no objections, I would volunteer to make the change to automake > and drive this release forward. (I'd also be more than happy to let someone > else be the release manager!) Go Steve - I'm in entirely in favour, in so far as my vote counts for something. Cheers, Jason From minc-development@bic.mni.mcgill.ca Wed Dec 11 15:40:32 2002 From: minc-development@bic.mni.mcgill.ca (Robert VINCENT) Date: Wed, 11 Dec 2002 10:40:32 -0500 Subject: [MINC-development] time for 1.1? In-Reply-To: <20021211102153.C1431417@shadow.bic.mni.mcgill.ca> Message-ID: Steve, Funny you should mention automake. I spent the last day or two working on automake/autoconf scripts for MINC, blissfully unaware that anyone else was looking into it. I've got most of the code building now, but I haven't looked at the the autotest stuff. A new release sounds like a fine idea. I'm willing to take responsibility for it if you prefer! -bert On Wed, 11 Dec 2002, Steve ROBBINS wrote: > Howdy, > > While adding a function prototype to MINC, I had a look at the > changelog. It strikes me that a new release is in order. Maybe > we can make it an annual event! > > User-visible changes since 1.0 include: > > * Rawtominc has new "-skip" option, to allow skipping header information. > > * Mincstats option "-max_bins" renamed to "-int_max_bins", to avoid clash > with "-max". > > * Minccalc has new functions: tan, asin, acos, and atan. > > > In addition, there have been some important bug fixes: > > * Jason fixed some 64-bit issues > * John fixed a coredump in mincstats > * I fixed a bug in handing inverted concatenated transforms > > > I've also written manpages for each program that lacked one. > > > I'd like to change MINC to use automake, and to normalize the > building process a bit, e.g. deprecate setting NETCDF_PREFIX. > There are two immediate benefits to using automake. First, > it comes with targets for building and checking a source distribution > (make dist, make distcheck) which are really really handy. > Second, it provides standard targets for regression / unit testing. > I wrote a simple regression test when I fixed the transforms bug that > I'd like to exercise with "make check". > > What say you? > > If there are no objections, I would volunteer to make the change to automake > and drive this release forward. (I'd also be more than happy to let someone > else be the release manager!) > > January 16 is the anniversary of MINC 1.0, so I propose that as a target > release date. > > Thoughts? > > -Steve > > _______________________________________________ > MINC-development mailing list > MINC-development@bic.mni.mcgill.ca > http://www.bic.mni.mcgill.ca/mailman/listinfo/minc-development > From minc-development@bic.mni.mcgill.ca Thu Dec 12 00:44:11 2002 From: minc-development@bic.mni.mcgill.ca (Andrew Janke) Date: Thu, 12 Dec 2002 10:44:11 +1000 Subject: [MINC-development] time for 1.1? In-Reply-To: <20021211102153.C1431417@shadow.bic.mni.mcgill.ca> Message-ID: On Wed, 11 Dec 2002, Steve ROBBINS wrote: > User-visible changes since 1.0 include: > > * Rawtominc has new "-skip" option, to allow skipping header information. > * Mincstats option "-max_bins" renamed to "-int_max_bins", to avoid clash > with "-max". > * Minccalc has new functions: tan, asin, acos, and atan. > * Jason fixed some 64-bit issues > * John fixed a coredump in mincstats > * I fixed a bug in handing inverted concatenated transforms > * manpages > I'd like to change MINC to use automake, and to normalize the > building process a bit, e.g. deprecate setting NETCDF_PREFIX. I'll agree to a shift to automake on the provisor that the generation of the binary dists (from the work that jason did) is integrated into the new make structure for generation of a source distribution... :) ie: we should always endeavour to make the build process easier. Myself I still find configure a bit of a pandoras box at times and still often resort to hand munging of Makefiles. At least a mechanism should be developed such that the binary dists stay current with the source dist. > January 16 is the anniversary of MINC 1.0, so I propose that as a target > release date. -- Andrew Janke ( rotor@cmr.uq.edu.au || www.cmr.uq.edu.au/~rotor ) Australia->University of Queensland->Centre for Magnetic Resonance Work: +61 7 3365 4100 || Home: +61 7 3800 4042 From minc-development@bic.mni.mcgill.ca Thu Dec 12 03:57:25 2002 From: minc-development@bic.mni.mcgill.ca (Peter NEELIN) Date: Wed, 11 Dec 2002 22:57:25 -0500 Subject: [MINC-development] time for 1.1? In-Reply-To: <20021211102153.C1431417@shadow.bic.mni.mcgill.ca> Message-ID: On Wed, 11 Dec 2002, Steve ROBBINS wrote: > I'd like to change MINC to use automake Wonderful! > What say you? Go, Steve, go! Peter ---- Peter Neelin (neelin@bic.mni.mcgill.ca) From minc-development@bic.mni.mcgill.ca Thu Dec 12 03:58:16 2002 From: minc-development@bic.mni.mcgill.ca (Andrew Janke) Date: Thu, 12 Dec 2002 13:58:16 +1000 Subject: [MINC-development] mincinfo. Message-ID: Orright, I give up, what's the magic args to pass to mincinfo to get it to return the datatype of the file (and signedness). ta a From minc-development@bic.mni.mcgill.ca Thu Dec 12 04:01:50 2002 From: minc-development@bic.mni.mcgill.ca (Andrew Janke) Date: Thu, 12 Dec 2002 14:01:50 +1000 Subject: [MINC-development] mincinfo. In-Reply-To: Message-ID: On Thu, 12 Dec 2002, Andrew Janke wrote: > Orright, I give up, what's the magic args to pass to mincinfo to get it to > return the datatype of the file (and signedness). As usuall, you always find the answer right after you post something... $ mincinfo -vartype image foo.mnc does the trick.. -- Andrew Janke ( rotor@cmr.uq.edu.au || www.cmr.uq.edu.au/~rotor ) Australia->University of Queensland->Centre for Magnetic Resonance Work: +61 7 3365 4100 || Home: +61 7 3800 4042 From minc-development@bic.mni.mcgill.ca Thu Dec 12 04:12:28 2002 From: minc-development@bic.mni.mcgill.ca (Peter NEELIN) Date: Wed, 11 Dec 2002 23:12:28 -0500 Subject: [MINC-development] mincinfo. In-Reply-To: Message-ID: On Thu, 12 Dec 2002, Andrew Janke wrote: > On Thu, 12 Dec 2002, Andrew Janke wrote: > > > Orright, I give up, what's the magic args to pass to mincinfo to get it to > > return the datatype of the file (and signedness). > > $ mincinfo -vartype image foo.mnc mincinfo -vartype image -attvalue image:signtype foo.mnc will give the sign as well. Peter ---- Peter Neelin (neelin@bic.mni.mcgill.ca) From minc-development@bic.mni.mcgill.ca Thu Dec 12 04:21:11 2002 From: minc-development@bic.mni.mcgill.ca (Steve ROBBINS) Date: Wed, 11 Dec 2002 23:21:11 -0500 Subject: [MINC-development] time for 1.1? In-Reply-To: ; from neelin@bic.mni.mcgill.ca on Wed, Dec 11, 2002 at 10:57:25PM -0500 References: <20021211102153.C1431417@shadow.bic.mni.mcgill.ca> Message-ID: <20021211232111.J1431417@shadow.bic.mni.mcgill.ca> On Wed, Dec 11, 2002 at 10:57:25PM -0500, Peter NEELIN wrote: > On Wed, 11 Dec 2002, Steve ROBBINS wrote: > > > I'd like to change MINC to use automake > > Wonderful! As it turns out, Bert has already done the work! It looks pretty good, so I expect the new build system will appear in CVS sooner rather than later. -S From minc-development@bic.mni.mcgill.ca Thu Dec 12 15:56:01 2002 From: minc-development@bic.mni.mcgill.ca (Robert VINCENT) Date: Thu, 12 Dec 2002 10:56:01 -0500 Subject: [MINC-development] trivial bug in ParseArgv.h Message-ID: Hi all, Just wanted to confirm what I think is a small bug in ParseArgv.h. On line 72, ARGV_NO_PRINT is given as 0x16. Since it's a bit flag, I assume it should be 0x10. -bert From minc-development@bic.mni.mcgill.ca Thu Dec 12 16:06:04 2002 From: minc-development@bic.mni.mcgill.ca (Steve ROBBINS) Date: Thu, 12 Dec 2002 11:06:04 -0500 Subject: [MINC-development] trivial bug in ParseArgv.h In-Reply-To: ; from bert@bic.mni.mcgill.ca on Thu, Dec 12, 2002 at 10:56:01AM -0500 References: Message-ID: <20021212110604.A2242344@shadow.bic.mni.mcgill.ca> On Thu, Dec 12, 2002 at 10:56:01AM -0500, Robert VINCENT wrote: > Hi all, > > Just wanted to confirm what I think is a small bug in ParseArgv.h. > > On line 72, ARGV_NO_PRINT is given as 0x16. Since it's a bit flag, I > assume it should be 0x10. I agree. -S From minc-development@bic.mni.mcgill.ca Fri Dec 13 05:05:53 2002 From: minc-development@bic.mni.mcgill.ca (Peter NEELIN) Date: Fri, 13 Dec 2002 00:05:53 -0500 Subject: [MINC-development] trivial bug in ParseArgv.h In-Reply-To: <20021212110604.A2242344@shadow.bic.mni.mcgill.ca> Message-ID: On Thu, 12 Dec 2002, Steve ROBBINS wrote: > On Thu, Dec 12, 2002 at 10:56:01AM -0500, Robert VINCENT wrote: > > > > On line 72, ARGV_NO_PRINT is given as 0x16. Since it's a bit flag, I > > assume it should be 0x10. > > I agree. Yup. 'Wonder if/when that got fixed in tcl/tk (from which ParseArgv is lifted, for those who did not know). Peter ---- Peter Neelin (neelin@bic.mni.mcgill.ca) From minc-development@bic.mni.mcgill.ca Wed Dec 18 04:36:35 2002 From: minc-development@bic.mni.mcgill.ca (Peter NEELIN) Date: Tue, 17 Dec 2002 23:36:35 -0500 Subject: [MINC-development] time for 1.1? In-Reply-To: <20021211232111.J1431417@shadow.bic.mni.mcgill.ca> Message-ID: On Wed, 11 Dec 2002, Steve ROBBINS wrote: > > On Wed, 11 Dec 2002, Steve ROBBINS wrote: > > > > > I'd like to change MINC to use automake > > As it turns out, Bert has already done the work! > It looks pretty good, so I expect the new build system > will appear in CVS sooner rather than later. One thing to watch for: the conversion/* subdirectories (conversion is a sibling of minc in cvs) rely on the existence of ../../minc/Make_machine_specific and ../../minc/Make_configuration. If you change minc over to automake, then you should probably change these subdirectories as well. They are usually distributed separately, so each subdir should have its own configure script. Peter ---- Peter Neelin (neelin@bic.mni.mcgill.ca) From minc-development@bic.mni.mcgill.ca Wed Dec 18 19:13:28 2002 From: minc-development@bic.mni.mcgill.ca (Robert VINCENT) Date: Wed, 18 Dec 2002 14:13:28 -0500 Subject: [MINC-development] Meeting in January Message-ID: We'd like to schedule a large meeting in January to discuss plans for new MINC development. The MICe group has suggested something the week of January 20th. If you're interested in attending, please let me know your availability that week. Thanks, -bert From minc-development@bic.mni.mcgill.ca Mon Dec 23 19:55:09 2002 From: minc-development@bic.mni.mcgill.ca (John G. Sled) Date: Mon, 23 Dec 2002 14:55:09 -0500 Subject: [MINC-development] Draft proposal for MINC 2.0 In-Reply-To: <20021223144228.A30128@sickkids.ca>; from jgsled@sickkids.ca on Mon, Dec 23, 2002 at 02:42:28PM -0500 References: <20021223144228.A30128@sickkids.ca> Message-ID: <20021223145509.C30128@sickkids.ca> Hi everyone, Last time I was in Montreal, I had a chance to meet with Jason and Bert and put together an outline for MINC 2.0. Based on that discussion, Leila and I have written a proposal outlining the requirements and design of MINC 2.0. I've placed it on the web at http://www.bic.mni.mcgill.ca/users/jgsled/outgoing/minc2.0/ Please comment. I'm hoping that this will be the basis of a meeting in January in which we can all get together to hammer out the details. cheers, John P.S. I apologize if you receive a second copy of this message. My first attempt included an attachment which exceeds the 40kB limit for the list. From minc-development@bic.mni.mcgill.ca Mon Dec 23 19:42:28 2002 From: minc-development@bic.mni.mcgill.ca (John G. Sled) Date: Mon, 23 Dec 2002 14:42:28 -0500 Subject: [MINC-development] Draft proposal for MINC 2.0 Message-ID: <20021223144228.A30128@sickkids.ca> --AqsLC8rIMeq19msA Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Hi everyone, Last time I was in Montreal, I had a chance to meet with Jason and Bert and put together an outline for MINC 2.0. Based on that discussion, Leila and I have written a proposal outlining the requirements and design of MINC 2.0. I've attached it with this email. Please comment. I'm hoping that this will be the basis of a meeting in January in which we can all get together to hammer out the details. cheers, John --AqsLC8rIMeq19msA Content-Type: application/pdf Content-Disposition: attachment; filename="minc20_proposal_dec20.pdf" Content-Transfer-Encoding: base64 JVBERi0xLjIKJcfsj6IKNiAwIG9iago8PC9MZW5ndGggNyAwIFIvRmlsdGVyIC9GbGF0ZURl Y29kZT4+CnN0cmVhbQp4nJVVy27VMBDd5yu8o13Ezdie2LMtVAhUEJQsuyn39oV6W/rY8PeM H0mc2EFCke5VjsdnXmcmz6KTIDr/pP/doXluTi6MuH1tngWEo/FvdxCnAx+SAJDKEojhponX QGilpQEjdC/7XpEYDs2ROB5+NS0QSugMWw37hFWgyawFJQkdzUcnFwACnHQabXCZLHunpHZ6 NlQgkZwSw7kH3gVDpZgyHsDKW3akFiQWMhJ+NzZ7D9EowyAqHw3HC1IDM/WSyLrAdBay7KXR rk9ZPgZIe4gSdJcgxAm6ChDnhbR9bxcgkJ0mSNB1gDppaYIOIZuOO8U1WloRTFaPZQxvHnIS kLTOWzRm8i1cMZytcwl6KYmfUiZKj7y/S1el0atHyOev9GZNHkI4LAgyuIhQApB2ObQWD7fL St0R90txtXJRrOUz/6w1UWjAkGRFURCBxSSCL4FUcyHdWJRPMTtjSSXka0jOSgQ3pvI+Qb2e IFXUSU75mjHfrjDKKoCy6yBMbGs5G/SDxmUGZ4Ll5ZE3RS8pHPk+lJp6iV4te110h+PHufA3 HjJeDBPXW0l/eRwx1hmuAuYNw2WjPsbLjeotl7DlJJR2NtsjfBs8UHQszv9ofz69l+axPPNC aw2yWHrF7rT3S3MtljNxXaa0KxO/npO0+WCSb9Mk8Z+xrgwZt03/UkKhDDpXCbJuCHWuiSV0 kl+JwpfOku3/i6UCZRcXU+kraiFUdL3aJyB1pEUfCq5MK9X5U7qvpBLdO48prLhvLXGL9eqj c14qutKM+1BGH5QdZ/khQjyTFvPpQJ7uf8Z5Wnq8KhO8LaG7cgvsy2JVuCpW92X0lVArXI8l V4W+wvW55HrazJG3v7HbHiv0H+PMMpcareSGRpbQj5K+0trKZO+ndiDk9KuNBryIesuiM37t u0yPmjqplQ2rLq6qrVU3fn+z3bbYdWeD+N745y9OGgoAZW5kc3RyZWFtCmVuZG9iago3IDAg b2JqCjcyMQplbmRvYmoKMTUgMCBvYmoKPDwvTGVuZ3RoIDE2IDAgUi9GaWx0ZXIgL0ZsYXRl RGVjb2RlPj4Kc3RyZWFtCnic7V3bjtw2En3vr9BbbCAjN0nd+OhkHSCAk42NWexj4NvYm8y0 7dhxMvv1IUVdiqpDtbqnZ0aSiTwY4VCHxVudKlaR/THZpiLZ2v+af19dbT5uPiaiLmv/eXWV fHe+efRcJ0KkstQiOb/YuPoiEbJKlUhUkRaF1Mn51eZB8vD8t43QMpUyM3XOX28eSFv05Dx5 tnlmGlCyEGmZJVlRpplMrmyBTHVXcNkVKKXTrYG/7L7pS95t/pvsDhPWgKTCggyl7UQLwQnT 0yqtVF6CzpfbIt1WJem8VGmukzMpzD+VrIfg3P3BFOhmUF7YEp2qXEvVFL2sK23TPNeiKbq0 RbntimnbFb2xRZUtUrIpqpvN0qLURfvheweflbqtdGFLzAgInVXh776vZTAdyCsdhNo1gmaq E/SzQ5eqRweC7ngPwYefbFFpxlYrT9J6XlUqrLR2HppxFnmq86qg9eyE9RXPlN6mSpZ11ay0 c2WqPu4nQNIJKG23lCdLnoq8n4DPbtSMeEW7wv9wtSozTXR6CzvfWVv0ilX63I9/uwZSW6Ss VKX0i8z4l2JvraOLIvyq4fWipY/wc4c/an3dvvTF7a77CH9H8FrsgS+a/xGOd0tj5bSUWtsE ng0gVWMDOHvN0LuwBU83D358+M2A7XfcGGlo25gJRStpYwAUSudts+97SVq+f82x/uRFQTuB Whz/c+NgRqbQfoPU5Ng5E6zMtZqFlojwq4WfKQFE+JXAlwx+QfQV4W8CfxC3Gyqv+j89eG4I 3Xfy3zB6/RikZXoS0DCuab2TrmF9Q695GP2qtjG2qRadEfOGf7fjDQLW/0SHY827PcLfH7xe tPQRfk7wgCeqU7W4MpZbNnxLSjLI0WWZCqkRRz81HF1ZypQtCjg9B1z7tm+rrQWYNeGd+Jct chENEW4R8G9Ty3yYtfZoEyBgERpnA0gfSxh3vaAfljYUouLhf4S/Tfh4+L9W+FlQSHWsRzIL 6b8++JqaCi87w7GoF22vverMlJ1JmRalMdvsn36yf/ISGRpX2VBbi3jp2qWeMjgy7/3pljfP OEnywDp3sBsG3iqdtx0AR+FAKHD2TjhfBeUk6G2Xd7bIHbQvWZ1E+NnDx3P2+4PPjoRfEFMM vfNlSb8m+JuwdJ3F5zu4IBzdnkyXfXj9A6fEAyiYxs0BK4MTdCDVrpFq2+cbrlmnRPjIpRF+ NvCRSyP8kVTdwh9G1d/ZP+09/eY5Yb87pjYsmbXL7C/H+jZbTvtQlF85nzdpacSDB2fmYZOC +sEg2A1Mihf8w/ABganV7fmXHIu755f8O14JOPrXbiaMEaNytniWrdUi/LzgT6WJlsWHEf6O 4IfZ0vfCh/le1zXbqrTICp8PXUj43w+/sTQi+q6ED2spZ/S6v+sd8Dp3HGta4BicK1/0RDJg T7o5f3EtGlpUgztjtEXiDbek9BvnKSAqMQ8kHZwi1Xkfql6zRo3w9wdfLlr6CD9zNpsQXp6x 9DOEb8dTcbrZd1dK2YvapVfbb/OAWmf1Fe0KpYSdspm9bmzwgJve5gKONDA/wlYElZRkkRV+ LbrJQfoZgB+/nz4yNiBvrY/kK248DdBpkB7ku4H+XPAPgb0DrCIygtIfQXqAALz+SyY8GAcw Wu/cOEwzD2kXQa2/uaTgXGPaLPLJAIbsW94gGEDQIJgMMGVArPCxExXiD75ImgciKvJAxAq0 a4SfM7OVWVpste9yPnHr0Gxb2W4GoO75Eey7Tl+OnJt+YJWAQ8ojqry1cW4ZcSHBLga6azyL KvOxKOcdpBEUcUdDGsEvejRFr3PN+JJzGWjvBR8tQEDXQZVKJxpE08EtMjDMwI4A8KBD01YE mAwwzHxMAaECsV5zscASBOcT4DwHjGAvaTZyeENGkOcodB+SMMdIh17yWtcBbQINCbrN3nXz P2LBNYopl2lpBtpTTOC9AK4pvnBVBexXkP4IDrNArIY3GDbRgH1pFzpocMQ0AXsSbJppllzT Ir2vAVTfK96h8ZudI6ICIaYdLYKrJTfdzAGd2c7jGdc7B+1m2iGg6UZ38/De654cXPAh2Btg aj/wDgHbFKin37kaIB+2G/wtLxpXyjmbx+5eMvA6iKVQhceeaxnQ4DSno9FFWZZuxUAXcT0A wtQAEjxUAg7przlWeI+MB8YnWU5Ao4RlHzcrmt5UZhsNLoBR/QvWJsAatzRGXPCwywq8cmCe 0o0LmDnc7fEoez0XzrJuoVZ8gBrh7xF+RV5shB+B93OtRhz9Mt02jn6paWzZX42N729UtRwh AuCigUNc8HzotHg0aHHS2QIw18J3jwPCh7mOu4SASMNmGBV0mjkdfodlPOFr2kHCKe3RkOm0 95R6x5cImMQwT4PEfJBAQFt8PYnggQUcNmKmuS22Fhci4+gD41GS1BFgtQKTe9fvFc+xd/Cr 1oYLgvfvsxDFPSFJFihu8IYDUE8gO+eU+UDAIQG1psED1x7oFNKhcFgI2OZhbwCEhWhQEGg/ oHimqRTARODDcKQoF1PesVy+LRvhox8R4VcJPyE1FzNnfjLm/NFhUuN12k1QEEMIP3eIzeww IxJuGLy2BK6JHPc+A3RTjg6e0g9/7o2RYwiezvgPjuq2qn8WKpwms8e/mTaAQIj/sP6Mnux2 Swu4N429vieu0bjh9hZUu0buJHOGel1v+TCDoXnMhuaXfsYGW8wUeYsms699nez9uxlrua8P 3vduipvq6KLfNe0KynrMQQSPrnQSvmZrf/gaD7izANIUx+8OTsmKCSYR7kmdDHsQtMcHeBAw BkSLFmCfRvgFwu9XO3OW/pTwX8ETOncPX93uo2OLHRxHAS38JFIeixURK3/kyPEvXgtgPea1 wB1B4Kp9y4l6GneHk4L2pPgBsj4oQ55igfxRcOsAeGvTbAHgIADnCcCDMMaULFOQiRFOFgSe JkitBZ4Z/fALhwdJblz2oxO0fuWShkeZ1nrEZPjCBQXfAb8PJHOCpUskLdg4dAlUZLSy8BKp Tz9qn25AWEtWh8uGP8LDyvOwMgfaFjxGBkId4WfB6YdBPwwkwlHXbJq3du3YQwgtjzWo5mEO Rvi5w8dfA4nwt0gO8ddAIvwU3m/hqym8P3ayCng/fNBJeR8k/AUvTYEQGYWalhX4/6AM4Bb6 kV6QH3JZseqJ8PcIH39/ZK3ws2CV+Psjy4L3PfmjGJ1itj9JknV6Bhyb/myL/PPW5ikYRX4b dNpvhQFP/qDESXq8Nu2RU3B6dzPe1/ZQUR61b9aoxiL8rcDHsGdbFMOetwAfw54j/Jof5DGP hT3rokfPhUiEsAxnKl5szkRln4bLzdfbVOtK9sTrh/euOMt+cBIZBsoGqaVC9Ful8ZvpFe2r nroGLwNQqt9x+M8OnnrXL9yH9HoTqNVQKg2tvucfNi0a6bv4lBvH5n+eOvq2P/VZhZsnKF7y pF/rT1dLkYd0ABb4EIznJzddRq7CU1DSLqDStw7u+e9RpDtrspifSPfe5GlEyu7y7yeBHFDs HESafZPHiGT1eZeFUTnC1ZRv/auyNbs8OU+ebex//wBMAi5YZW5kc3RyZWFtCmVuZG9iagox NiAwIG9iagoyNjY2CmVuZG9iagoxOSAwIG9iago8PC9MZW5ndGggMjAgMCBSL0ZpbHRlciAv RmxhdGVEZWNvZGU+PgpzdHJlYW0KeJztXUuPHLcRRnKcXzE3S4CXmm72izrFjhPDgR3EySK5 GDCklfVwVrJlPRJf8ieSHxz2TPdMVddXZLFn1ggCQQdhuE12sVjvB/v1dueq7W78N/1/83Lz evN6W+3H5v9uXm4/vd48+HPYVpWr+1Btr59uDs9X26oenK+2vnNdV4ft9cvNve396+83Vahd XTfxmesnm3t+HPrd9fbrzdfxBb7uKtc326brXVNvX44DtQvHgdvjgPfB7eLyt8c5p5Hnm79t X5UBGxdx1bjIEtojaNpyVdzp4Abf9nzznevbbb/r3G7o98t9st+8d207hGnzj/dDuzgUqmno zTjUR8QFP0xDb8ehCF4VmnniT+NQRG8Ifhp5NI4E1/Shnpe6OTwUt+orutQB5hMOOlf70Mxv 2x4eYVi66sZ3NX57VTcutH04PVnF0xg3dP3l5t71/Y/iYsP+ycNiz8dHWtf50M5D3x2GdnGI vZJPvJFPvZNr7dHgXV+FLuhD4I2v5FpvDxMjKnrP4YprNW3R8reHp+KJ9Q19KiKvOj31SE58 c3iq8aGri/D1w+Gptjpt6OkR1F7uZ7HrEdI2e2ZVG89eX+uriZp3kQanoS8EDH88EGVcys/z fnsg+bhpn9jhrYT0xenI5hc+ThFJryNePvRCvg8cBTgwgJlH8vABBT6RQ2At8MYf5MTp8CMC u14/6n/IwzDhJk0PndxPyyGlZHpCaid32Op4eKWuTocIVy/WasMJ0h/lRMLog778e7mfF/KN ABFy1zpz0qWeq0uJw7jyYed8PWyvjjKaSYWjAAOCFVAnkMiEohq+ezoEqBMsD+gOSFGwPJDb P59OpaEYaVwVTjQFoAeiD8AFJk7LR8nXdZIOZj6dCDvq9s5LuDpOeZE++0o/4tcS1HcqMVKe V7mZblETRXyLQMiAw5YvBJvWGQkIYIqHSYRF4D3gwBnzgOCeyTeCLT6SbyyQ3HwtgGepo+UI wAPgRZu60u0eOvFWIvBnFVv0FMHp/yiHAFxA2gKdCeQh2XZCMehnrUDf6muBiW8lEJqc5hsC a93KtQANAnyBPSpKhr+Q0G4W8wCq5eKT1um51iGSr9GpAIhyMHErnwJEBgx1QJ2AcYkUnd8I 2K1AZnJNAQxUTV1xpiTIqSWR5bwrvhaw1IFlBuyCAv+Ev1E6Ai8PgrtvQ51wiABVA0lXIN45 BsEbC8T76gMCUiB7QDdy4uQ4te3JcbIZW2AIqFbgcR2GdHx+KiFS3C++E9uZkhdKujoKwiRh dVm6oqsXqBIuHMHJSGEMVgcaDoh61WWilqUbh+oRDX2vA3o9DnVRUIcGEG1Cy76RmiThddgM a1XYI+oDBKvxae2GPvRAWyfsfs2r4hgF/ogttKTbQNHuR6GlwHFEJTCw4ArUE0cO0KS6+UFd FACEzYDXnKKEfQqEHNgK0Npg4iQt4lCd0O0MIqAXAdoARCDcAnBk0yOAGAFHMbgBkIyjbL4O 2NdZfiBkTrB5bNjinT4xTQE+p5SsxBdKhJlAAAmoobUqgOBpEVQB8d54CgC/87xkjPaoYPQg LdVotYTByTfu5FMfy6cSvgPw8AE/EO2diFgRWmokkkDoBqi7NkFxiqWXEN4ASOC6AD0DHAmg LgAngA3b9J8WmOKmHCGBQN+4zJTVQ+V8lF1XdTzzMHhEBnUVsTDUJzLlibFXEzFHfDI1xBNa U66MpsH23MOTZU+mtUgq7p1cfkqpkbwbSc4xY7YdX9j28oUU9mUmjr/sgIxqV8d5VeSJJh7j MFCM8lRmPR7NPhH31f2P4tsiFVQzAF9ImrIx+VZyzDf35GJArthsEEDZN5L0iMswvxFEhwms Xux74XdG+qxZwCPrdyrRFP4QQGo60FHrqP9MCr7fS/L+5j7Wq1mnWUtd8YlAzAD7BywPgojA lANvtLmwYGJRhEQY9nyt5PmnYgW6YU/No3S2I4EckEqSzq09WpezIICpA/xRk8cNgJpAiNK0 nZ9K8g8wEmeo0uwzS9QT+8zzgMGiSxGCmLdU+S8dXx/F4GS82JwQQEfAF9bNEvpUQRCVpyeI SdkIguiCPrHIoRY2O18LSHWCr3mIOFLTyJUQgihYYLP+MzEwAHVJ5QMV55q+VXHyi1Q+6Lme Cyb/QKo0E/3Twxw4M5IIN+q5bIpRAFaBbuJbLHL9M9U4YHlAW6rYzMQpP5OQSrE5eXAiuHiA qqGYR/kn+pAt3Aj08WnisTrrJAU6oFZ1zBTLAnAUDUdpNKsMvq2txCHJ2kfJrPM2ld96somu NZv4dOZ7OROwA9gRsBeJxZCoKfpWTrTxJLCJScyhE/teFBNQ2tdDDSAhT4HXGJe9MJm4TIhV W7mEFpSwlbAxGM6rE2RLAdjTdNSa6GjIwkBVFUBNutJnYW2ntwjmAaWqlUZwXfUxtWPkq9IZ zEQFGuF63WjXi+wUxytfp0ZWB64EgOoCCuI48Vs59AsoCL4fzejDmbP/HQ3Bhed6DfFPk1Qi 4QMvDy+hbcAbgQcBeK9EaUw/zkx8gOQNJScAo5rCycmaM4smTMIGTrlASVhC1F3cas9mo/Tc Iz1NsB8AvZ6gUNFpU54lLDb9cELNcFQQqQX0cSI4ZKpfsOXBgHwFpC4VJo/v6NV6GfEO1jJF zBiebflEkvOaY1fnladg9wMUg9soF4R8bO05uv0polOc/pKRi9TygDuBxWvL6YMhoFuk5Lpk hUa6gDAx8cz2nMj53KGtms7tlubKlJiq/XZwzRieGzN9XQStXzz4q8ODLIPFq0A+kaCka8IS og4IRFtRzitp0hDqARZZqx9TrnwJ1NPZah+A7WCrgFAsEw7DxUNX2RJM8BSoErLl1cBaydqJ boGJjJdvMy2SlSaFxfkg9FucUUgch27SUuhNhLM6/HCyixKeItCHhOcu0x/AlzdV9ABPG8Ag dyh9b92bpEMADVo6jscVnpUgK67V1hlkXfl255pQcTmvB1ooGwF7AqT8pY260DQRhLhsX5Wq mr+MT/C0+iUMgoQdpPWQAYEE2hcWlgatyNFMJ1Ryrkunkg5ZyvIgfmWLheV8RqWZmW8UKL00 rSU6XtKFwyuielmppxdOcRqvusr5oS+l8dXNAqTxYGFmZ4IVMkeeMJrVAxZBkuw5/VuiG9gw pkANCK4AgfYfxYThwOulABlvXoKl11Flmh5tMW7di6JoBiQM2KGgXp0fLNBiZg4ZBtfumv8n LZAPL1qUgOrbJNxGkyoA3g4gSHDwFy5R8PrWbNmjdFWtbNXgKF+jwGSDDkMgB3Vl1WJC+adL MhcmZSe7JDLdAsDpWB2g1ZqckhZh3VZuVwW7LLjajaCF0O6fLAhCcLYBE/8k5Qsoy13dsaTK dnTs4GFA+gDja3k4fX1Eq78Q4ANQoy0zYQuU6X5ruRkJpCCwKWzbtt/xkziskpZUjo+Mf8Bo TQvqcHG3Nn1mE4aS+K0akb5da6AAKE5QGegPkqEDgzCLhk3w3V6YDb5YmIEIMNCKIKeqX/RC w6eyBjOdrpRlYqn8wpnEQoUFO0DAe3dj9SUEBzFseTNtGPtWKhRjv/Ih4qJbuIH7E+adHzZL 9JLtrunYRQIIUBVmow1AjLpJkZH+BRPzYtZiiWWlElDMtpIw26ECzax3gmaQ91gF4iKqwHZT VU5kq/ebXP5sAUskourqhQu5aP/FL0FS+5ZV5QVOPidmQbgZaFPTZQcPj/O4ISQUaOid921p ZAAYTUAKgYwi0BEP5cQH6hBdCxwyCGuRodk1BdoFnBYgGHDlCFgLWGNAWDBvllk6wAsHIsl2 EYzOQRl1C7YKJPEDiSNAjLb8IshVghO2Xb4AZOwDuW1wUHpZGIUeHMcJEwPLtAqG28X9jIao YLj9xbrRFo0G8efT/88KLxhug6uafn/xbxuNpPGS3Q+c+oFTP3BqBhNgol5MkSmAAMYWQCHY IyATW2Nxsqygk4QDmH5eipQFJWrqQNE4sFMALWlcCcyKxMKA625PAnQUnqWCs+mjwKwrJjhx Or1uXBeQ9P5gLn0QwmuE8L/GoUSgFUjR1b1OQOi8l0/dQe8RxYHew5bx6XTobZ1TdPk09HpE Ekg+4LSB+mUgflVhp/MFOYdFp0yUS957EBZrB9f1Lb9BZR5jT08X32U+VgACXlq7HJcrgBzB Udm+crA6+aall7OggnBG6dVFiduL9SCR5V44DnxBeAYEwrh6WNuMaQsEAwTqt4CICtZ+PAx/ biwcbPqCVxfUEgSmuXby7wAgW/at5O62jKy6/N1tomSTd3KY2iy1jk1O/rJABFwCDwxfW+uO FtXEnYpAVoKAvi0lBE7cdlGjTQreQVYBVLnTBAjZdsO2rUO0tgIClrETZrZpPv1aZ7r6+T0l WenBDU9NhnCwSiQ55ZTvpfQFeLCVMgCw1E4KXfhr98/whW1VS2DiuW39dC3lwxJZMyKdgk5k Bk4wJJrvlFsxYeM73qDMkyU+GJDr9QO5J9C1YLt2S71jK/OBhILvHBixKzr48WVY514U1Yfj TkGiG7DlHVyQrjbQgob1CeyuD92s++0XpPNX2dJ54IUXvADxvB7yw/FV+sms7oiyf5+Ko4bc Y9nrxyMRCGDXLmHkm7ZVX4OJRWJaYZx5Lf3bDxkGKMF8pn1Yv0kpnk+buKBUWuE2S1aV1LlG 7DQbSD8+8d0g26bBbcxAOINPpAEFaUuwAwxqaficL3CsSVOviOUiG5wwqIglzJWqki36VFPm dpPzNNWKvmvwDcRcTYkq2TN1sUWS3dL4p14UsTpateDkzI3ToGo+lydjU4BDZIqmrBabwPPI FShpUcfsBwYAvlQ3I8tbyrUNiYkAR6t9jnSjC+jFB6eSk3C5rmNTNZK+eCZzCgrjQHTK9j0w mz7QA6/0yy7aLbmJ7pVJOfTBVd3d3RGR/pzBwpDIhC9l+XK6edSr8wBUAAQpvxkCHwpssh/L VG3rI420i8q2XwME8k0Dg9d0T4Rmt3Lu/6uURzYnwvbFtbS0TaKrG6c1xehCdg2wFbRLouYf WgakrIIUxY4ytaP6rTMZrQ6WX91LamtY1O+5yXRe2Rqw9TwOqMjNtLjbFKZeY0ws4/nhpGU8 E4B+9f9BXWSl7lW9q9xQFxd42MSuZgnyXnLFpjU2BcxPpb/ndnYFdFG4IvNNApPTTMpeGJUt D3DYxdctM+F5OfaHwzlkPiIOzDJb8+PnUtupn6ObfoDmayBzCuTX/OMgss+vNq2jho0vrfoq zq73NVMgXQDICZjuRTujuP7NxHjkKzFgeaYYQKnO3+XCYAjwD4BI//KgpV5ofyzV9sUmYrd1 EbtdqF3E7hCxXEV+b10II9//9N3m6bpaN9/sXFyUnZsmN+VNG5Gpms7OXLzd7ssDSVO/Epwy SJvaeuQA3wFrDtiw6atyvX7WAAi1ffxOWbGLpNTHP7W9G+288UyBSWWLfRRtPfO1SWDcruVY /q4ijs2UxRGOXZRrWbrC9qcXDyC4/QGMHOLjoVUR3nq7jzS33Tns2nWu4SercWs0XmrZW4AZ 2cftDqG4bzbap3Zzp+D7HvPfwbcWSr9TmIrq2GqVzi2soG+0mi19Exf3i6CxXbICvNkq9QhV z8dm+8q13r5mu4gvE+Upjl2X+T+695Z5o1qZvDYLCGuhM98At92hcEm4bMurVxshVp9EVh2Z oKtshaugITxhQF9KuUYuc75p4rOdq3y/F8HgnGzoBh9kuUB9H8ojJq6S0WVepnZblXkL3U6+ GalehKSrVb1HgIIHYAFHYFvr3H6BDCeWWhOR4rzbU1wT9XQfnYAh+gLdJYz/rj+YJydSpslE pqL2oIz//gvCDf0fZW5kc3RyZWFtCmVuZG9iagoyMCAwIG9iago0MzY1CmVuZG9iagoyNiAw IG9iago8PC9MZW5ndGggMjcgMCBSL0ZpbHRlciAvRmxhdGVEZWNvZGU+PgpzdHJlYW0KeJzl XUuP3DYSBnLsX9G3OIBbFkm9eEx2g8UCTrD2DpBLLvZM/IjHXjtx1sj++qVa0jSp+opVVLdn EyxySMKRqGKxWI+vitUf9nVl9vX4z/zv67e7D7sPe3McW/51/Xb/zdXu0VO/N6ayvTf7qxe7 6XmzN3aonNm7ruo66/dXb3cP9l9d/bwz3lbWNuGZq5vdg2Yc+vZq/2T3JHzA2c5UfbNvur5q 7P7tOGArfzdwezfgnK/qMP3t3TunkVe7H/bvyogNk1RmnGRN7R1pRdONJDXe73vjqtZ2p+kO xrnKDH5/MGEZbfiPkQlX45+6anC+cTNfXo1DbdU53y5DP01DdRga5qHjnE3V9b4z89Av45Cr euM7z7/4dhwaxuntMvRsesq0vu3nodfTXMb4ftmvd5Sum2moNb5t+C8CugD1/6LTv5heDMzp +vTFtvWdnYc+nkhteRa+pk/9Os3VuNNcYPpojY4n9Zqy8Df61Mz7vj3xHvDrHf3ivEbrfJ8h FRABSOWnj5nzjNL1mhIB5pq56sJuO363n01D3mQl+kZ1FMAXr+n0JykEMr6s+j39IGA9eJHn czx0rtjHopoX+zZlVywSQHBm3rTe94Z/8QOd/jc6FDHHEdZ3VOw74Wyku/iOfvAj/WB0slvC +tYTPoPT0vK053UQUOMtv57JNDhfV872qWkAsgg0KZAf8CJ4Kjo1Db/n4AQeSbbjRvU95bfl vxjJtad8W4j4NMlBIMJldBEvscRI2mrofd9RIjI68pYeLjAEXpypD0S4jMbiuRqT+pw+9TvV t2A7gD0HT/HKNX5xs2kA/NrMCcBoYF0BqYAInRXjhSlmIS/kkZkB69G5LJzuTmnIqokVofEK f6Z8AIu+pkOA+GiJDRGSLuVyrBBv6BCY6sTSbiVvsb4F01Mabul7YIg/ZTFZkWMgG+CIBjDV e/oeIEvFUp5/KjsXK9TYNg2pbQLhQWS/M6J5S18EYg5iFN5PZpw14PhnIpmXpy1vKCfBuW3o lifndjLznp8LnKxXVO6AKs9TL/vJsZV+SPUJIFXnhv9bRcQtVTs6F+bc2InRV13KaEG+QDB4 nCsT/ABDBVan89moYgeLq5g9TQg+2CZoBJue6hmkCeJWN+Evrqr7+C+PnhqzN0NQJEEWr17s DneP2rqy/eCOjz49TuKCmA0JCjFUzvuFhmMoYepRFk3MpzAU3Opl6PXEAdufhP+Xaaow5Aw7 +1FUTBv4NPTpQ+N7idVPP3jkZjdys0kOzQrr6ccNdt3+UFfG1rWN2GoCP8enDqYJh3YYIr4u jH48Dny5s10Qo37Uq11QFUN3fNLQPf/xqxPvg3zW5kgCCA/TuODxOOTDJ71J3MJ+XK9L3Ohu 1PrNIj4v6VM/TU+F6Zs2/WJMxF9nfvbeG/6LH+k5A0/9Oj0VDmhrKRGOnwttVzrPtE9DUEG9 M0cpdybep0dPrdsPVROoGV9+8EXYK7L/Ma3fToTFEB6IdF7QLQKeKDAKr+OzC2wlawXR2acK pEjvAfUUq0JgtpOvA9AHLBgM6YwZhzwhG87r2ITiW0I+B20u/HypspW8kRWM/0eBIMAW8Hlg NeYXjffdQlFFxGh1NoIBCOwemtSKfAE1Znxivp7kZAzxnSAHqbQA8QebzslBxiEAwgxCLYAd sIGvKLksecDRUbkCHLq0JfTktQLr+liWhfHngacKWAieAhTp8BGA3+rUCaD+xweE63+jsvwN tWeTEZdgAYArS1wHwgRm+c+daHSEBclpP8r7kAaA1oawsDFn5K0IgpEuCNgbsEOfKKPBXGDx 4DzlA9MMXMUHIcAkl2dYwFN8rkFgIWAOEHKwaTqwDfCLw/JEIu49RAJkAVt7HzaJlRKwxAz1 FOKJtGoGUqJqgEudpN9DflHMAZA1yWIUHf8ezyYBedPDZQcXVNwQpk9iXsuI8/I/s+aPoX6g +UGgpMvJghcdfZElMnkRWEqgV7iDks7F1w8AJz6WCYBu8QavPOVLvVVqbtOVfE83EGwN5z8A Tyudnve0VD6+tA+66TkLnPDmLyqx5XMp8RAQtRO03A38qsHR53D4RNHlITeQ2M8M6fK/QP2w Pn1MaUdnbxiNlLz3nD4E8h+Aqj2lnTUUNCXCKum1zsznrMGR0h17neNO7b0OBihRwqnmEwIz 4N+IeIyQxtZ5HTwLhYwoe8xYFgMFqPOfNyMWEvDOm0YhIH2uskq6HIIukQG2DohBUXQQH3rg 5ADXC6QJgecHJAOcLqB5VOU9nBFPyQI4DE1nrIEpOwzVEaEPymlwCmBqhPLNkf3jk/8Yn+zG FTVUz+RKb6KQIVOCyWOX8ZBOS3J7m1G/bImCDusStEkB4pWeHjBXCeIq5B3pe58BHBPKWIpc NgEuBkSABGy+sjfdRf4V4NNJICegmE/qg9S/Gv3MOYs0JDhYZ6vGrsCsH8Y/mXosDjT8tqug XhQVlJhREDeVVD7w3P+zAWMpWy5ZhLvZP2T0ZmYl3OnOSMkciwVD4FZnKdbOoIqiKBbLJwDp Q5KLmAf1VhFXPMu5QVgctgCfhg+52KUURVvg89iIitEWheVQobD1pnKzX/P/EXSd69gLGhyp TaGUli6eL+WiXxf4LGbHwYsAPmdsjeiWcfVgmTXwj6H8siAgQCJ5SE6IxorK4IU6Ap1y1ZlK IGVcTUJRfW1kEzKIP0geXZOppMIDyYjlnbhMyak0se6eBV80SXTGwZmu6vtVCQNA1XV5Xl7n gWwN50joql/np8KnbMb/0ZtyraYWYhzuYtcmfR6fkaIcBrkod1GtrgOLsG7v6KnbptuFYyNV geoS+MBigkKhAt2f0WgARgJLBXoEeJrAidua/QQfjA4oX09x8UJ86PsB712XOmcjCF48Jatw 8YisLLLSIaY6KIXP4At1C7pbjOdCT+wW8LpOAGdAdY2uNKTI59HVOwo7WJBDQCACEjCQYCnw IVg7Jd37XGZ5w7yS6BMQKeuCUP6uVEmmS4Xo60J3DqmDuD8ZCh5bU9l25bF9N/7J1FUd3KES bwjIKnhPd7WFgZh1TnFGzUjQHlDVqpBQxQ1dVlF3vnWlUm/YL+ouDekux5YXvBelLOI1Arr4 EmTdgsD2C10f8mDvypUSatL4DGZ8zCVHUVWylT15GVcL5ScuVza+WgK2YStMA4SedV8NXg49 lVVnBDLOhPyqW8q6G3dby9WBFXgoKDpwHoAbq7tlorsGns+cZ3oKqBTw5mvgXIyQksXqd3AQ 4lMKtK2qeYouoi2KEoTKg4oueo6V6iAAjUPXG5cKz8djDPzl9t5LInAJmP1pOqlBNF3i02Qi KF5vC8dA1/ppw4Uy3ku/yDZL1xvOrViPieSqikQgme2wIHiNgCu6y9W6sBU8BRhxYnomOazr kwIUEcAlgX1WRRg6tBcEHVK1WNYippXnB9uHxbS9CLsU3GRMjdN9VAOJ9kqqCtFZjoTJl6/n yeO1bF4lOfIcqMK2mcrwkQOu5ZwBSO/HE/NYE87b5ADq+0TSdHBfQQkql1uAAgfs4uZaAgCa gLm4CyUb9Iy2t0xJLQlor0PDE2TbhWPOtolKFOeYWavXOA3oxsJdvVzhAQL4wnbuQ1MWXLQ8 IyS699TcZYooPlfSDXBFF6rp8lK6Jl66MgQdGq60meUOWrYkPnNVjl8c07Rs5ewRpbTMUlBG leoxvpgqVm38QQWZNl5mwGUzoYoL6G82J0i4GhRcXbVupeB0eTNdKiG/rkz0w+PKQI0L8lnQ LWH5H3B8VPoPIL+6w6m7scE5R2f05hN4t7m5L1iQvuHDyvz/ITJsG1obFJXBXDoLlTRuO9gm nH8AJh3c2GGp69MmTgDC3djEacpqNWNzrSSzkzbAuj0d7UT6AeA5PbV88UCfmhtCxV2W5sZL Y4+oxGVL2yodmZ4WQ0V0dSn1znvrVaR2/PTvJpYm13nSDl/THpm6rVx71+opaWiW3Bpa/l7Y zSYR2q8nOlmED6A9fLm7cCZ1t3ZUmlfnm/HHG7S91XltBTGsiH5dsmmvrh8M0Ec0s8KiLTwL gLa8ZGr2f9aHC2Azwtbokq7AIQFegr7dBD0omcblfM0YQOhj7xPUrfFIROwv80iE4I7z2VRd LheUROvoioc43PDgbJCQfmVns/dw0t+TKAO5CsAT0Xvi/866FG2IHUw7qJ0KEMJvdCqON4xC 6BJWnhRVplZt6qFpq+aO7vfU89jsLYChyxr9IThNzn02mw86ZhTXXPDWuKheQUgS6NCih1TN ngmLpEZE6AMDvrWh8hhoXf73PHQ9tgQkG0T2CQKkYysQE4D2F/WVALjd5jLUrT1DAcEAoc+m hNZWl0lTyzgS31hrc5P3ElsCLL0OVges4CwKaYHg+8qatlSznfeDVNq+4YJq07WE1CXpzm16 xzhx2lQz8LX5OEUocOTKR7SlKJp2iiICea6dE3rdbM6GF6RFYDY8X8sxEw8aocWr5rJrieLZ UGeo01WglQ1bKxlvMjiTdCGnTViZTSrUyxCXzjrYvq+6dUmF7ncaIoHM5MOoydraPFJy+QHv gEbQ6uw+rN+78xCof04KfOxhv3wr318T/AhJEYt02b/N9UT0ooZYrn+uPrpHx1tXnVOC4WF9 lIMWS/qmb+iNs+H6wob201IgDnJLurvZYC7g1/C9d/L9R4Gyp++9pO+piuALyk8yxifboijz dT6siRcCbq3rynGOyYuDq301mNWvlYD94X+waMS85XydENhsPv86GJv38spbieq7RWQCmq1X CviCPE04tf0HE6T+QqBTlHG+ahs9QNdQvm4E6I41guJPt1xPdj7+UZY3k0THPv3RzQ+OQ+1y P8oCcLybaS6CrNHGselPvsyYXRfeSwwTghLjNCD/UzGkeVkz3vK7i5qf07kAlHhLXwRPgS/+ znpfibgYbyrTTHqocRnA8eBcX5neFrcm9C09ht9RSfk7lafvqVgA/BIUhIHrFbVghvNK/xws q1xB6H4y4symLRsSWCp1C1QZMGuPKA0gsFfd3kFhTXlp0GasiE+ZlV/YoKaIb12p6est/hyl ruJkK2DAXY6ef6D+ye6/fDpjFWVuZHN0cmVhbQplbmRvYmoKMjcgMCBvYmoKMzgwMgplbmRv YmoKMzAgMCBvYmoKPDwvTGVuZ3RoIDMxIDAgUi9GaWx0ZXIgL0ZsYXRlRGVjb2RlPj4Kc3Ry ZWFtCnic7V1Ljxy3EUaQ2/yKuVk+TGvY7Bdzi2IFSKA4lrFILgYCadfyrrxaSWutH/n1Iaen Z4pdX5HFnlkbCAwBksBpVheLZLEeH6s/rreVWW/Dn/2/l+9WH1cf12bXNv1z+W797GL19Gu3 Nqaqe2fWF29W4/NmbeqhsmZtu6rrare+eLd6sv784u3KuLqq68Y/c3G1etKGpucX65erl/4F tu5M1Tfrpuurpl6/Cw115Q4Nt4cGa1219eRvD32OLderf6/vypj1RCoTiMy5PbAmkTN+pG21 3Zp48K6p+qFZ91tXtV1/pDd2GKrBtn3osOm6qumbdr2pt1XfDnYnlX/uBNVUnR2GvaAedk3b qrHO7Js+haauqq1rpqduQ1NbmdZ58mPTzdjkR3xoutvTatsDrW9D0xCeslPTjl/PQ++6qel9 aHKeYVdPTW9GJow5MgE6frV7o6la4/Yt98cX1jF127ra7pveHofYxpxa5+qJ1iWn9WkUdjTZ sah2XG5MY6uhG9b+P15qXtqHn4xfZ+3gl8KL1ZOLzz/zvQffe2LsepRqZ11rKWN+Kfim6B2W SsoGKp2jI46p3I5PeXn2jdwEOv40CsFPvbV06uOOd7zjd2OTn5q2QXwDKj+MM+zXYlfTLnU1 9K7v9k2vMgJ5zXm5H9/V+3fJUgV0v+ekrniTLDVP6yA1IKKIbUCFTG0vy+PTUZBtdhnRLQxo fdDIDnD6dnzIb4S+liV8yZsA8zmpPPAmQOWWN93wJjApuf2mowI4AkOJCO9XvzWum1HJiBVM GqBVHWcb7seLUScORCeSZdSy13fyq/hD0ZtesZ+Pa6wbdDTeH7ULX6gHIleaAaheDiihyQuq a/r9gXe5ZHSv+TjGE8S6bWXrcILYatuHAyNeVQ2fIy5ftM2iyW88bdd1fBB1LGdK/t2o23p/ nia0KSAva7ggexd3NO4oSZ0SuE/S6jj3g0x+T6t1rjfxGJkCjcl/SK3LhLiA6MHGuuH6IH3y Wnk2wFNAEK/4qIFmlQ7u+I1A/YOlKp2TsZxfc+51ixC88ZKPEWxLoPfBG38U+aLkgREhj5Ed Tk3lzPGNwEiQF/R8EcakAA9EcVhZNvLeCKJv2PwDjTDJFOzPD/ypW84WGA9gnuwgoF3Agd7K 4wFn0HEyEjpClik7qMD87E+G3p8MwZVwkgkA1KFsuNGDodhwmzoSC6OPOQ7j94p0c3B+pp9M 5wc4dDtvyHhvKGbum8/DM3NnOFrFaz7Kv+wIe7fXHdzQ+9G28eo89ve60PHQtFMNfXBfbTTK eODfclqACUALPPWJ7xtCPnKL2iDXtpb5OpLv5X4P4+6iXu6NsEDjpQfY4i98w4fznsvhXkMK yO+WMwqaOCkgq2suBtUAyaKZ1ujPnNEbwbKCU8hiKnHHO87pd6mllZh7MBVvhH0dz/3OmTUm PGbk6SHT2lFxsWhJeCIKBm1q5xWJ7Xa2rjVUPzz9urbroWr8e3ZxsP3vQZEMdv/gH8BbJrNp sw0aKuie8OSfR2HSWIZkbcCQA52rh6Pyb6josgEVYD4Tm8RxWglr40qw9KIwRB8m0zbyUH4R qMQySFrvZeEJ2f4Fzi/w4Bfav5AvSksKm4BTuA4T2PexIKgNBhyUO84XMD3lYWeCCYC8znoH 5IFwilYJsInYoR29qnwGEmb/z5wKGaMu3gKEA3TG9JS85OgyAbMNjE8paJGQARgKsCjBu0CM 4pbLQFIzcTRJcsmgDWu7qjZ1bBECrwg4FsTDSlisuuA2MJtlb40OrOKTCzYDaEpH2RJ+/g+S qRP/Tl1wQFiOpNDRAR0hLwEqA9kNybinwG0GWkleIGK09pWov+msAKcTLF1dsALsRWWcP32O ypEFyhD384Ec+fuAcpfj3F5MB6NSUnzgEE9oJh4RBv2AyjnNk5bOcHCEJwIDPKohhx1oky7s wCUDQh+AOuAhfYADLcmXSzfT4o2r+rlfDw45XbAWNIEFLuZaxN2fTZaAmK1sJVGdfKnSEaca dBrlEquKBccHeBcw5aRsXTZBC+Sp8wtUwfO0fZJgSydO2VPQ2DoxrX+EJrMNPv5E629siF+O xxH1U3dBrT4c7zZBHSxuOd6dSXSDjkTZd2w2gNHD7bPEYQYUDzBFk6cgyGNN/dIGWyM3gbkA nEYqSCUTXWAdOHBc23lreqi6uTVNVm3Cu5JRGGFTALmAQ242zZRZonaAa5k3NTRuvc4JylvA AC5Dt7zOAtZ5ozrdB2SQC/jIsIPF0RCNjktMARCuHG9gAkm8Sx9p+DU1sA73AXgHKy7nSZDF 0HHRAscQSBvYY0NMi2pfkpFMRBAyylAVtByxU0ZmRac/I8K7mXchZ2YGNvOHYSenHqS6Ih05 i1dvbDtU+3j1QTGn4tUjKRIkyKtuqkh1qlvOw1PVTcKJCVqv+VNAkwHVrQvlysYt3WHAfZc8 +mxkB5AHMQSdAgdWr+SEJayugmMgO886L6M48j5TBuVmsS4Yr4IiggMHLL/FlkEOsCijMUsO UgB5ABoaHCMkny9rYDHrQLVbtLDmeq12XdXbtlSxpY1pIMKETvkTfwosrGf81AZ6AIQeuJkd E/4XH4rOjAW78tQwh+w4lDu7II4cvRicFHxbnjF7mV0t/O3AmT3Nw8gmQkHqTakUyqP+aHPG AJlNbbfV0IW43ASvOT4a7c6QcDduZ6D0vnH/ZM1ZOxsSZ4+8oBdCAC4BNBFAA11RMYJnMRDn 6shWAi5xeXxjdELE+Iyd6eH1YusS2BKA6wA4FR0iCSB4APN3fIgigoeZDFkEzzRo2yYEr0It gX5AWGDCOKkr3m8xNgeMGcj99ZEttkAsB291S2aQktqPmSQt+ESQ1dfJ47tmnAMOroAO2TFk ZoC/1i9/a3aAnmbC6QD0DdCuuutM4MBO42pyQPg4inOaFZ6IhoBTUD7WSzKFAGZQbvf/yJtK TJPHTXH85mE+we1NSEo2rzJr4pYOGNirIh43a2wB5zoDZFKZXDrvU7ppA1NoAiYKWHRgMeVx OTR0w5OsgFRutwPYziNE4+hTYBGK0bgzYpq8dndVZ5ZkG8hsO3nl6K5YynlmGozSXVbSqbSi 3C7VKQUeQ7wd5QtZZwE36gJb0ipPBLPB3gfaV47RPfLclAMpIxno0v7JaOaQ2MO6iBlwjlXY HyAV3S1TecWmVSnAq+igL7l839nzBDBBS2ktCOSJd54oER3iU8o5bGzTV8bOFPLz8FMXkh1N zeYmBQUgg4T3hAHG4CZje0ghL3ASJJTKUrR5Eao7EwpTmeSAB7BbwWzIt04zVuVTQWvHT+lA TXL5BMqXzH0aVRhNMjgAwDGRi0jK5rkGKAq1Kq3bIOd1qNRuMxwB67HIpll8bgL4FlVp+vQs BHtTlfULb1qMe1FBaACnaby4TF13SIGsTDqZMINLtv6ds5yJTscUo9nKgDtE2ZaljXWQTZ1t Jm9QFplJnLdFGzhTL2QzNpHgWoFnEvNYtItBFl7MAy1Iq0ggwZjjIpAg5RiYbTkjCUxaGqrG sdfIfEnsHkBLlx6Tl7cAe0tgcoHS0WG0uZENlhy3PGV8Hh20XHYCxB4yaM3cxMunSMYxLgD+ 81PEG8ymatwMLwm2Bdi1wLIg4wYCa/gMTaqSZ/JCGm7ohxbclN9Y5zVR18dJPcvZUaTqdr5B IHcEQvH0mi4dtU94hJwEFUcM+bzhTfv30ask14yUnKjL3NzW8Q6yLuBCOSAPmODJJ5AwAtfH P46kaB5Ld0dfJk9dvmQ2qCwXCdJkj5q8+8D7gTFLSaq448PxhXQ7xlMBZocswMQ6Ak8debA1 FzzPPDaJmQb3/e/YeFB2fxTn3Abtq60DN+21lxnP8tQ3T/jOfsaHqbNx07i+hEGbDv3B4Idc gIt2KbAu8nHVDHYIQO72lqXveIpleVZYIji7F1+8z6QKZOFTJ0sXoS4yjxZH/B8NXFRig+vq Aoh35hHJUePUnX/eDsCkaf0a7dspogRCZLnLvcBvADyLzh4iqSunp7uLu8B91YV2RBTBWYtl UL6jCG+ukunf+SLR6Q+wA88SZ1x8K0aHpJD1bSbaCMYL7qIqlzydZ6mmRj4ud2K6Q5fJAElw 4GKClQmyRKO7g28nm8rYOq6gidDMtbfKTD8UVxXyrlJ4UndbQ7ciFtdxAx3JU4kCywsCSFIZ BnGSaz7JiSRfTr3otoBK65/pvoRuqKo8xmzyzpj/KscqyWCgzMU1IPviyuKLYNrlZqbqfrnO dopPrEyBLVUZYqAnya2HWNvVvffobA/MLE0NtT8CbTf5YrG2e8G1na7+ec7aXFxMFdDSQTrB fgdx1MVR0/OYIOcsPgA29I1uVjLB7RxsVef7yM5c5lbs95kDi4QbZhd8QZ5IvBW1vwVKAxRf cBX213F/0E9CEBNlIkywHFEEkd2zqsNXTWbQjtRmHWnpoNa6wPWviAcoQp/St+vwUb/N5aEl lnsOvxIRTl5rXnShPYMIFq0tWZMvKIp1qs7ImF45nSHnJIVbzHI9KB3eDNgNAByoK1oIbIlk ihccNjPRidr+JDzHpnZNtXUdrCAFsrqZ8KiuRNNjFZYAhrBkGcUlWWVbshTOkgI4FAHCaJPq ezrytVCqp3U1RMFxDWSr47Sk2gfd6+Cwl47xaCTzA9wMXRWyVb8f4L8f4P/PB7gOwiN/mCwT PlhcoloC48VnL4i6qj5SBM7LxaAeYEroIPfJnEGiwikAOIIhSszvSkA0bgZVQYldYHyRwDf/ OMjZnPEMqFK4xXaqGZAubZdwtWWMoCdfJ8Ik/1FpPDAeABvXcQ9KmumMEMCqzmzShbyTnvYQ HTam9ZxtcbxsVk7h+MEVgrxqJAMhXrBf7O0PUgIB1C0A1+4BZoqgPhK0AGhKVwMBFAgAkBjw MQiO3UmPpwxsxanrvodyz3l/4CMEOCBAC0CydPAuzjwC6mxMX1ddV/gBjIK6O0UpKvAVJ6pJ 5U8m0o5SADimdeqNFO6ALLNfm7A3DlydmGDiXldMXoczkVI0sQCBVagDB8jBTDpsEMxcLK8C wzgmn05tJkSoS8vJiF5NNcqY1ZLqn/Le0RVFWPwlM8mUj4cC4jpSwiLuCJhQXcEBtqfuexm6 4v2AU17xHaAQdF+am8CVW28Sm+wNH+Ck/5evHEkD9OET2BbEcAD8cOYjUEMyyvyALJN4a0fe 4kXZ8nRNq3GQ07ycBrHIhla+4mqOO+CArRPvL/KDIiZfcIsx7kimLqHldHypTtpcPRLAo07v 6T5Zk67vCCScuBwO3gg6qq65gwsxgDq4xg5eyHWV7L6zwBocNEB6qe7OHHQcQVWlXfxJHsVr MKHQQL2RdLHOBDZ7Aea2JDgAUAj6VQ9GkrvXnr1TXoIMkAFUp2KI0uY8UK25AymaIF3IV/fV OV2NIl3hKH1Nr+nnL8cdTBNGIlIBmfg0PCzjFXzHw9mnKP/aes/XLv4O4zkRk+BzYMkSpZ3j KyTxhTAQqZQ+dq0Nu2c8xsVQnrPvq1z6EChB4KudWiIps18BX6d+Ci3jChZViCtf0Kpp1IEb ASndV9+B6MWP+iCdBeZJch7BMpnphOLvoZQcuboE0yxJUe6QAiSHrghbbsCRF1fwJfONbbZV Pb8DnTydOlioR3cjLlfuRw4q0gUHUsILSuxHIkt/bj6RowBq+owVHWLCcu3GzJFQXF0xkYsS kNIQ9H/+75PJX+pMR3zBnBQVRSsvYS5/GiSklUjQv279eBy6i3vYki9Cw2fjX88v1i9X4c// ADRovOJlbmRzdHJlYW0KZW5kb2JqCjMxIDAgb2JqCjQzMjEKZW5kb2JqCjM0IDAgb2JqCjw8 L0xlbmd0aCAzNSAwIFIvRmlsdGVyIC9GbGF0ZURlY29kZT4+CnN0cmVhbQp4nN1dS48ctxFG ktv+ijnagLc9/Z72LQ4cOIc8bCzgHHyRtFYkayVbyjp+/I3kB4e93T1TNfV9ZLF7VjGChSAs l6wuFslivfl2ty/K3X78mf9/9vrq7dXbXfnQtvz37PXu05urj78cdmVZVP1Q7m6eX039y11Z HYq63NVd0XXVsLt5ffXB7sObb6/KoSqqqgl9bm6vPujGps9udl9cfRE+UFddWfTNrun6oql2 r8eGqhiODXfHhroein0Af3ccc2p5cfXV7k0esgFIUY5AzrE9osbAlWVk8n1ZF2XXn8B9/GVV 7w5F0w712D18tS37w+66DJNqD8MDSX4zddSAmwBu6OqZap+OXfqiDWAOc9Pd2BS+Vg79Qtvv xqa26OqhXQY+m5rKdmj7uenV2NSNA5thbno5warqoV8GvrG9/mHBP8xx+eXrDyYwfUBpbqrs kL+MTYcRo6Xl2gwryZeq4tAPfTc33U4oBqybg57ISJR2bno9Ng1jr2qB9c0Efh/AH/R0Jax/ jk1N0dRDV3HwgOYcllzSrz800553jNqwYaeU484ad0w7DIwWgvbdkNw0cqHnCbTl0DZ600j6 vLLTFMRo5Mx1L7ZpNPrvDC2+mSg2hPnMLc9xH725LSBAeoDAPV3ZkTBwu4MhL+xkT/M40lJB +d5+5Yklv50W+Lo4xVWyl2cPw/WW4Pl6G6ppcgPa2uUF454kqAm+Dub7L9uLsYQIu5sZWRhS l5qTHQLLWCZxGz04ikkdir4dqtZg1FmElk6CR4H1re36thzWE/PBOzldNffb2K7VZ0r/WZ5V cDLBib43SExbouHzBZO7p1QxO/W6rquiDntcXdLgNhCL1hsitBGWq+gKdt582gO7P/JyRX1A OHA4f7FnBXwLHDIw1R/sLn56bFrG2Xv8qR0mxJYIo/XdQD7WDo5bMfUKK9f3nDSg10cWPPgi EF862/T3kyBUv5+BnBHLXqfd15d84MGF1+peTDrSywHwEk3q7FxXh33RlL0+1UhGD2rF0HZn HX87dVSiWT8ekSBaXe/HzTVSa+x5M/bswsihWU4PEwvASaQc5sAXzCcKgEsPcCsfLJ/AsLPH 87m9IPkOPF97TZss3gJUIr90C+7MlLSTVBsAjwXkurekB3vJt4wALzAhsC9d4ihYjZ8pbYJG 0kW2JViznyimUlS4s+QSVwQgRERmAjIEGAjEihd2IEAVLPatHQiEJvBFK3lxycqIZw/UajlW AAVwQVjogFhMwFSTAagD8XUR1PqiPApqAxO3gKIKDjzY+cDE8vNpYKNn2gdprbSLFGFEOyug gWPEbwr5RX6U5cq5jjKA7uP3fJtLWAAHAMt3zfkuVjpFSXkfv+drneDRvlUE1yO4fa36oKdS yL+Dr3xmJRSfRmwJyXm8RCjO4xdy/2SbOI+XLANM8cdJcJJCJvjiD7YJgAd8GcwRsH3AOjmd 8alpAUep9RyDNJrJKGeRFqAHjBWg6bQNjuqp2CxLE2DkPoV/7hUk6m65Jz4CO7rrPfzJR2kp kGwUURPXwAvXQQOwntqBlxR3AVF9jBrsj0emRNbVkFh/uoyPtCO2ckKX7CRlrtO6Hg8m4I1A /gXSISBpluBMNkTt2RAdB/XUfhDQJmpkWqC/sp3sZYq5qU/tB6Io0wTb0YYRURfBufbJHcCo wA7ehEPki+AOYVJTp5fLc+47/mGwnODDqw39QOryDQScCGilTGyKmCzBqn1vPw++xXdAGHi+ AxILqWzxADBwC8z7aBhOfmRwYClBHhzQVv7s7M7tLFXa5Lh8d4TPt+Dji8x6pLxdKReQGiGc 6zGeNbCrGwQCAC+H7yhs858nLZVgqyMXVb6hDPSK2lPOYRmMgAsTgjdXQMQGAwDHYy1qJ404 6wnfAs7Jx9wBZu20ypXaAS5rh20B1AbsqbALsLNz+3xs6kes60iQDpDILF/zqY6A1QmfVWMx jehmPiZmpSgitUH5cp0AqMdlCYBx1F/ZTkyoxmIgi6RCcQRKg763A+PinPYMK+3EhlL4mA66 3hM3JON5GiGf78llXuJ2unBDHUVVdrNqtLgLqW6jzq4MnXpDNMhqhksOcR7DBdi71kwc2V6D ivNYIDt+azcXuAbAYsSpFZHbTpM+87dIPZiZ2dQ4X8wN5W8CkksKBPoQ1+DjGIArxCKVofin yMmQAmFemOcK+8xg0VvuPODMEZ9ZBlpLYxIw87Um7fgsfEUP5M4RstgRPsycNmsOFLfvyitl LZsHyIMpMqeagg4iUC0KABKIeQIaN0c031GVdUX5zL4m8NsXQQX5PnHD5Gly0Zs54uQGXEIc qdYzH3AJR0BFzWSPGJIZNz6qiXw+oSgjULnFIm75BGyA6Q3QlpAtvsLwXb7bScR6nkBGd4eE oomQYeA8m4THwKkPkAgoXnqdQiqPdzHLjdCwgDT1UiL5P9PpV1p1kjaUDNdUxFabE8CP6J2w WHBrk8cQmYQFdGO+pR1WRahme9Rzfc2CQCE53SyDbW6QExX7OT7M3x8xN3ABJh6nFRFqfdY8 f+qUXoZfueFvrem3aPrTHLeZ/+QwZv6r+cfBWvn81vyqkOvHg47y43s5+NWcUcICMjM1GiTC QHxOKsDyfETlSUkHa8yQA7kBEpgNXB5oIJv6fEmAh0tMfzyJGKUGLzEF1lMWdgUJ4Qm6eTgu R7UaCNuWWwLGnGcoAEIqiGXyCReAxYGdznkQjp51uJYN16PREMtAHqT3WHGMv7fc729jk06K /ZNhAwCULxnJFwyxOuzTd10CKzew+gBYYEL81gMx/f5Yy6WJx9MnvsjzFuSiRaXOmEk2HhKd sJpqHHzC51sLHUSFuoRPYLSw43zBpBkiJj3EWTbfc6ZZF/t+Zppgc/rOZMYW1tqd0INr3is7 sdbnHHhEEdbaz/jsfMnWQMsHNyI47Cfv9UHl+IXrcryY00ly11UXAD5UsggUGTyVLHSa3J/H ruV+jEws5R7SVwQw2t/bGxs4Xa1i9M4uO7dqy2UHawz2MUD+3uCwun7EtDjtSPPQLb06y6LI qhGxFMa9WpN9UPjMHX2kNlBt/jA26UoTc9xXPQxthGeCS9pniQcqIDP3QCQSvAuIcqDJFQXM NC6NFmCWoBdQyDlt5Bd9aZ3A2hjz717o0CRV4m0JqJqCrvlwlTUhtfvixLgrWt7bcfMhEDAj 7mkmaWnagIEn/bfn47JCeYwiqGHxJMiMwhKrcWSfskmSWMGMpx2l2Zm85wQllPulGzdh0xiM Y0IW4JY8DM3IXfqLQMCw8Ua+RDagJoAb0ufSYmy2G+0MTcSTCZC4pAMcUDAjO1Rjz8Oo4jkt PrHDJ+gDtHxB575YfK5yGq+dJqqfeU3IN4ZasdhwwJZ+sPPxhEIyBXCaTSQGh+qg0tMMbIzA cgc03K1XC6qiAxgkD8omOe7gNhzs9MG2GzgssKFQyqzMhZuze8JWaiKia0aMTzJ72+ct8KXH AfDgnPksDL76E75pA1SB+ZKb1S7H0PRq8yz8hOsBMDR7NwLkge3VR8Gs1ERT0UfnBcbNl5HD A9hJPGIILOPS9IvhclQmu5wYrOnAS4WBgMuEzwdQi4XVaCSABdDHtSPMV2iPEbVTsIEzubNu jSCVby3mko60d/hOIBiYk5Cocst8cRjbEoH0F30mPjBHMDArOzAhXPPswPzcx21VlPSW8+Xn AVjZNuCUVv643Ci5L8FALlUGJNJRNQkCAj4NZE9ACPBFl4q9uoZlhPu9/xRqmtanIkV8tQt8 UnJWFEzCMwZggWn7TLM+zgDMj+DM+7LQ82vk6KmCpr9Os5fGHZ8ICGC5gmLAVekzQvvi8Jkz TmO6qfiFBuVTHXzcjUuTko9k5FzreJRHCECX4FcPfB91jZKxTWn3cTzmxsdj/22bRLx7Vn4J 252SqDpJN6t0U+JQZdQ0jvi2wbd8ZTF8oiHnIokqw5esMP4fOyHmqDYOz7Yp2nPH9e+mnmce z8lLOXujg3AkdpY8V2Bn+cvn6eUEpxaA2rrBrAaCNju/UeUG9RVnYprCwxsa5e7lVXkY9kXV 7upDH+A3u6FrRlLtptpm3e7dN1fP897tqA5t0QRa14ehCFKyfGhj2A1F3ZXhyhsf2oCvI8iN ++DG7kZ0GnUItCnoflrNQIiDXt96OLbcTsNkyahZ4pDBqzOkQCxdc1r3emeX9/XEnsKKHZ3r Ly2sNxQJU2NU97Lz+XGijCll5J2iINYLS1HfpClWcr0AGe5cTRb6U9cE7TiBwsDX6/sjGY4G njs7cP7gPhCr4bAAWm8s+HsKXg68tQPtFL+zpBELdnaLhbXvDxotY/HVsACmS/FvA//8bZRq aIvwT3P9r8aeOrJo43WeushdEZIBo9CWlZzAdIkAKlAwAgr0Arzap1NmXQUyMsWX/+RzC2eo L3raPl+rL1oFgN+kHWlQWclScjXW+gozkqUuqGfpY2BdtEB3cRWd8RXh8lmReMmXKuB6HoDn K/Mn1pLHIrm4ArdaJSLPs86VLTicchbyikjyePg8qTy1SMLyxegC1kRt45zB0XI5yIAS5+Eg Hhesw1aTfqKwQFayU3yLAugglChL4xmPScR0YVkXr+Ui+eLaigG+YqieYi4gZd9axG0L8+Aq 2FE1M6+MGKN5Mh4xwicBenHWaZl9MrknmfgCjhCvoibJ7cul9IWJ7VwSCA9olXOMvsSVWeYQ gM9IJ0tWZPBdNmAgCLUCfI+Ks3xnZfgik67OjGQ5Dd4XkS7k94hMB8C70v2ZLRjuNWAClyzg FPFy5LtWzKNlp+KviqysP+ijVPRRws7B4cAG51XNZIk5VtUMZjG7avRaqUTfxoCbv5/oABDj JC0S1i+n8ebiYyIJCsC6tNFXQ99q9E1UW+LXsQk7SdYfzboUjTM/KRO4sicFBblyCLgNwMoX mwA+mBFjkBLh+LmPVwTxcLOMUoZujdVltQA36Cle79wHlRBlfBKJL2GUnhjOoFbnhfriiWck 5VuFvsg0VkE9+VqxT/xg0drJRyHZwAhre4SCnjnl6cG8cksoR6yiqpgWcKDRctWIePwylhqn ZQRAJowTKsIuWFA8eHBTc5iqCxi0Z0/az9UeSvF6PKi04Yv/znrTQB5oEEP4CeEnupfv2U1f AjowqvCqxivf1k6GwIJePtuZDwnfCgFCP3KxCvDFjGfe9MAslotpH1G2wDgrCQDxJG4kjGQF ufwBPKqUkLTmaFlRB2wkYkFLfo/dwEH8qQJzOhN/fCZZLhT7KkckinD77iJhU9evGTdBqA9E UvP6dPyLzsL3iQA8bQUkcCbSyjnTBeHwkiGtdaUxz2PSAObjDjwFEtwYIDE/YbQCOptAYsH+ /zIxX1OQvkrGCbW6eEQ8V7ix6xCRgi2bdvlj6Zu9+c8Sc6lAzsZnuwNIgIp0PDlDDgQXEdF3 WbG+s/BSxMt5jh94xFgeDVAVHBwgxn/Pww+bsWBS16wNPwSFV3xFAnLreuRJM9wMLW92ILKz 93s0cQG35TqSLxNgW/rGFLc4xyxWbX8o9k1YvLCmdbkhZrHejwC7AHEomoOIWQTO8utyfwgH q9Q7aYlflMqTsIQ2ydmDxA5pf6a3vd55RsCHHnoJmIcoX8Q85JPmuYtXorqxtg4/fSBkJ8tr L1Mo1r7TfMmXuHhaE3g1xxdfkAgoz3o9QSJxyTfgAKopj2Mco1UutETynkv2yHoKR2afuFx0 vCpFIh+F5jfboxUBzMO54m4mbCn/48RkpZ6Sk/00tNHHk0Av3+53FVDmR0R+MKuMIvD6JUyy viSSrU8AghAs8B5aohpRymXAVyuhlGRolTCKalM5KWB0qvmEuB0yEdtBdmTyMQ2xPyK2MBqR lUjQoxFZcs6A7wAetsV+pb8XvRkjDGRTbVd66KFDI0MNxHxzdXXsjOLtWnrOCFRafgFPJXxs D05vey35Cb1xtwYJf2+k91+PM2SB8om5/pCbZ0Xcbgar08vH47nkvMCKbrP4RHw54JhmRPfp 2W2r75SklM8u7XMt0xRPvpypvBlbAknRnqdo0JObkaaCVFS7WdcY+KLRbNw1my2I25uHPcZw XQVe3HUlzMCWxHJlHvhUP8BqfEzgkgW21bJGw/jOMgBBgAYIYfB5fR8rKzpH7EWXOXfGJcq+ sPOzJlfdta/BgQRbPc5yI/UgV7+gDcBnyJ+Ugo1e8wRe4Iu+ahBgQkCYY8Ee6AHAqmuKbjgL +bCveTDHz5pYJl5a0eTnninJvpLNCUkVLCRQgsBUPiJkTd9kPrGKzp4lJ8BJgNT/BMcFGL21 A4EklrqUs4paJHK6XHk0/uhUfRR8iYrAbOdLRvYdZDtD/haV5Bwp1chXA/HifDFwlrqYXpWw iUMgmeyS9Vk80QJJzuWr0pJ1xCTL23bEvC+MJDSEnDyuJJtZHYmaMHf97Lp7gGiljgGPpUqc S5Z4FDlMLNI8KdYAFgMG0jQgsgKRKIiN0V3GsHddNW0xOtjVqffVVgNNQCD1RWgDay5TRJO2 J5/RymU1BWi5QvMBCj4JOMtQ4ImkhGcalzCIuCVcjs6tloW1hfn4q27Z726nLmJ/5rFXzRDj RCBkyjeoD4+vPB7gCi4bNlPKrqs2sIz6jHX8imLj8tX9rLyFx/X6p0wscojLxJLxxiSUXyQs XxVkHvmVoCZ/IW0I3HLf6EfytobeSvabCsbhtd2QWd9EGA2jJ6hcmuaXOqVs7Xv7CjgifKn7 vgCMi0en0OeAIgZkV3wHfKHrkqkj4kjx1Hvf5DyRYRqWzxgRZevHyAnyJt/6Z07WBpRcsNb0 6terEqZQH3AmzeiBAHcwQ5pCWvVlcTi39m2VznNcO+zJAM3cfK/48XIZ9JT7RPmcSkocytYL EjgKE+oM0Ob9grZLm0k+qsDFLok7FbtcN0bqHEd0deDKy/CfuOpQeC2L4MGzBCuj77bzU79W v3Cl6/PXcR3+zYf47PHnv/KSEZdlbmRzdHJlYW0KZW5kb2JqCjM1IDAgb2JqCjUyMTQKZW5k b2JqCjM4IDAgb2JqCjw8L0xlbmd0aCAzOSAwIFIvRmlsdGVyIC9GbGF0ZURlY29kZT4+CnN0 cmVhbQp4nMVdS48bxxFGkht/BZFLbMA74rw5vgTOywjgCLGyiC++yJIlbbxyLMUPOYf89kxz Zsgq1vd1Vw+5MnRYqDnTXV1dXe+qebPdFeV2F/7Nf5+93rzZvNmWh7Hlz7PX2z/cbh49GbZl WVT9UG5vX2ym58ttWe2LutzWXdF11bC9fb35YPvh7b825VAVVdWMz9w+33zQh6E/324/33w+ LlBXXVn0zbbp+qKptq/DQFUMx4H740BdD8VunP7++M5p5NXmi+23ecCOkxRlmOQc2iNoWdNV TRdA6su6aKvuMNtdmK0uynLo23nvr8PQvujbodrPQ0/DUDvuY2iXoZfTUDcO1fPQ1/apA7B9 MZRDvaAWrPifMNQUTT10lX6xH04v/jTBVbZDvaz4dpqrH+fis38/DVX10Fd6SD4FgP/WbhHs 59/2qR/sEFixCENVse+HvuPT34ahrtjXQ7PM9Wot6j+xGPz7NH3dDs2Cwr8apIrD6O2RtRas 1uKmtbhZhu5PuFnAem6fAvv5bnpqpNS2MhTR7TUQ8qkfp6G2PGHwzgXEjOcRg23P4ZqRU49A 1BwT8z0bz79S0zdhLn2MN1VfFvvxPG7Kkc20++Hw0wtzUIDwABE/s8QCyBPcJEDqYHrALO5d 04OL+rM9qK1dcUbEeEe6noNqmYVvKgA8wDPgYeCegl1/a7f40rXrp5YWAahgCMwFzp9zGUP9 SfJ6ZTd0opLOklKrEQFuoHgP3LatfQ8wLHAnwYsCNYABt3Q74D0w+1N74+9P1yDCIVdDmuAx e81j7KUBogbQPrh/vnt0TU7EKEMLXcCvyK4B1iIi/zs79Dx2D2LMGNw8MNesUYySrh8kkEOY 66jbHoYePSlLqSLeVHUTFNibUTesy/nwJ2R140O93nRV1GeUuC+GISwRiKcd5vcfT7+Mezvq G+DgvmckJnf7MWbeZ4puE3SUGB4zmNphrq60ukBE5wIXARAvvVMJGC4Sc3oqALvVJsB6d5Lm AaO/syKcK9aSdMFZbe2Vc3EjcKPA6YHZH7lg+MkSNTVD8m0OsCAQEQDzp6HOzB6RlIxF6vWA ZGHCU78IbEmfASB0QSsGjqr9zQnx9iyOT1FEJJQBABYAHuDmzgIPpge3YBLEdVUV3TgWePGu n3kpIGGf6ABPgUMHpAjUcU4t5chczgSq5GZAFn5ntw+ErEs4M/+BhhTgBrBP8OJXBFJ4c+WL Pm0fXGYAF7AWAVIBr+AKlTzZr+xc95LX+zYMXDTczkvIZUBYgCIzfCEaVEZ+ik8AZvLCDoG7 DV7kbjV5UYAR9MyeqYU0l2wjRpd96KQIADcbuF5p4ZJwjcwLtsPQl/JFwRqDjTIwgQwQC1w+ gFCAjeLjjED2+VTGrNsviQfcfvAiOAKf52W10cUVV8Gxc9wzEjVxx13kyHycad6OdBVyXwxQ nq2qnPLFAM2f+zMSROIjcWofALELDBL5lIt5AE1O3NjSbifCQ7+xTxVSSgEYhf884qgGGlv0 GDp7yJEbB9RNoEhSx4xmesYx0w38sN+clmnsypGbsNbvs7VTgbsBYLALcmm/jsKTF++dnR2o n4AVA30K6HlUXeez+DguCyTpuT6xYhLMxf0zD3QOgNOAGwSEnU+WMr8CxLPcD6BaAf2ZXB7h qhbUc/tWchduNY5D+wWuk33bWaSe27cJFHIDV8KVZeAm/AFAPAAUAvngChcw/zpzGiS1Usxn V4fUfKqky74FYg0wC5/eAAwmcOeBTg3m4ha1UZc09ACIuC105kDPNlOTOAAiEGh+PmvTpcmC nfhyMcBt56EPaT2vli88RgWiimSPkXgXuMhcKMgNAUz4jFAhMBsD/TL7X8JIF6Zq8gx9O5VV it+5MBP3+UTclECV/Gjac4gNKaqEJrXA2YIgQfp5mS9CyFSWLiJxOe47llcSTO8LIUZlayQw ExetEbAA8JcGOw2bPYNI/g4uNrjFLOEiosRy3+6Fa2WIl8hlB44+wF7oSfBXlOAB9Hsph497 6eKOnTytDMz1+9N+lWk3sou+2A1nStKUQbkMfBYGfrd59KSqt/uiacenb1+EOHJb9nsdYf5V ePE8Fq0h+eOEIOm5ARldBzQO4b5V6uAnBr5YyG/tKfrMWsAQM8TpAYhWezLbZnx1PAGLxzO8 7cLLwQ86hdanB389PaiC332AuF428SlFW0L1Y8xZJ2Ne0225lu36fJsZXsu+6PpTiuNlYawV aj5Pg5VgZYSx9JFlhLH0iz73xqUUsZZBqyEaxnLF3XWKa9xlHAl7APMVTA80R1dOHFCqeXKp xDOAIZpa0NERV0AdKOwnvbtTevdNXfdFORKXinjPnttRJ23yon3MNTptdBlam00DFvzICn8Q Bbk0einncnFRsJ3VQSNfsukM6ciujpBmOCqScWservelaknoObuS9/a9CDCQt+CjG19GAnAb AQVUujDBYQNnCA1Ck8OIeHy5C8DOTrYDZHvLwcqKX0uuxnPpElF85OqXa/n8ugDIn8mGBVMV GiYQaj4mxHSmiLGWwd0088yJpnJcgYkzao8i7NGn8gGEgad8rhAgp33OUedWHioBm5eCSG62 wjXwS5kfUpthlANJzvqn0LYs0wN81z6UwQbPHPyED/BoPptXe1NBOJ2Fa5IvMm0hshPAS30M Hwz5UYAZrpDekXs4R65G8/6ossbdR2elZoEIl6FvLMnwCLFMsVqtcM9zNf1prmsmNb61oGZ7 bWpLYJHgLxV2CggQJLu6S1yvCMxb33Fk+TIS1rPyi/LClMRWMtT75ItcHTb6gJ4LcIVrq8N6 wRx1mLL70/kemYUvk5RV1WoYwYvAwcDD7+SwUmIGZj4hjuqr4eFVIOHQGnNorfIRg6yV+T/r dVMJCmXImhUnqr4z1Fm9li8b0qeqgek5m3Lp51nCwefei0ZZYjr19+AgPcwuiZCUhrs67+RS PVhStthwRB0DZrJPI2W5tWnfalxJRfqhTDb1dSoA2VGJpTh3N3lLnG9Hi3NSyU+YYwqlJ0JL Ys5IhH9m8iO/7CKsFjyVukmrU9wtw3ZFW+gFVFBzLTzh0fzGhTrArllVD0h1hEUhQOWRKzLe rOe6onOcw5DQ13h+voSB8hC14hULR7ktJRcE2bYgT8dXhyjCCw1f0ReI4kRiq3sk5/PF7Vws E2jYLiPb5SR+gGpMxFF93T6yrEFJ59yEqlvDdSQQvjYEa/kqT6aWtibgMDyr3FMEqHcN0MVD WMAbIIe4DJArAr6Q431VyMmCPhErWl1OxOs25Laf2bmu72Nd68HhAPGsc7k5kHVumZCNfAPN EawHtN839imQD+g7Ppa+qnd4auN1dIJym136EqL93xZE8Jsjp/rK7pqo0O5oVZRRHRcGuizj LlYGHQnynX0FKOy8ICihuvC8Go+C0wYCaCPMjfdhShjmQkBHjGXGRIAkjjB4X6aPmvh9uF01 cld32uKVFZKk1e58Rf0gwhHtmBHRmIDOdkLBMXMS1AWxUnUqMtK6pHwq6vVP5x0lvBEZ5Uod fwU0A4hwM+HQAsViQB1s9NHKIRAK4v6beAuYXyx7PeE0Yx1PYJqQBPXSjifAn4rF715Bf5a3 Ox59oF5NAum83Sfhib5o29MQb9YUzjpS/Xu6QrGnfG5F8OLHJ3TtY4hoxmdCeFJh4jcQE3Lb QoFZoAcCRDCTheoP9rJmrkEriTQ4FY3T1CTAzNe9Ui8zDTRYcyvWkZKbiEfCatLA08SBlwv+ wy7IEHEGtnzlt9MNkRlRX4ShchdOs+Q7+ZEqIzZfJHln5f0HlT3AJl3NJXynWdun/mQVF/Bi PDHxzImSKA1b2wwrHjhPe1GsDFW3BqxITa2EeeRyJgksL8LStrThxX2oDc34anvWoQuIZO5l l754VwkpT9XOz5LnjUKMX1Crab72Kz6/1Oo2wf+1mFhdL5BRWqcvG3iRdQxIotB1/szg1mAB DLpkCM9nyKeIrNavpoYLNMS2d1nDwLuqenKMtegC0oxVzj560qiyNQ35oWxNKzTcFJLBScDG fU0ZSPsy3Olg4V9CIwMVY2CCp/Z0gcnKkxWkoAaizteyd67NlplqzC+k5wJ8SBxBnTjx5fd/ TsvL7Awf0YPYUWlf3Ft6ZgqdHnpMac4kA9Z6ebAW3b2vCA5s9aQSLSOifydHR7sWHX8LQ0PQ PKoI9b610srXwXwL8Civ6GBfGehTcoj5dw8f/dgVu/HBT+e/L+kHQDRvOn79ZBh52z58UGS8 if30CRBwV4HDAegtgBBsm2bQVBcM/WQplw9JcxmQma/hJThjQByAOYAVaRtzLuDfuZbn3gd5 O31pSQBIcBD/m1ib7HoB6MOXnASmByzZ51rm4foEx/fdZXCo/FpIMgIEcn+8q+X2bjMisi8O N65si5HxVdWuDd8MmlqVd9u3X29ebHbhQmdf5macsR7NHnmZJ/le7ftiV3bp4nJZlF8OnqL8 m6kefWgPj97aS8qy72BxvnQkAMcnOOJTt/hzn1Yit5CntEhCAB0CAPRRuM4ysvrxspb18GCF /uCjPlvLJwFHj+dgRdLElUsmlb7Ma60leGvzjXzfqmIRJQ3Dlx+Y6YHFyPmvJ7+3D7eojoC6 tRdqVmNGZbe0rtEFhseWCoAFAmC4tHNp4mCv3hFaTw+MddaY/vBipM7IFfbOsJPVetEPf0Ry l8Dt+vLDEwe0iD/7WJo0LMGe1U1mH5nRJMRC+u70fZCMmsN2EAuT6fsXZnqnSJw5fzWJX9Lw eB/4fB1xnPm0IzCXypxhMTj9iq+Htd3u+0iA0CgHWOEfQ8jvMujyovmIA6TuAp7l6zUHMs59 qfc+Bxl1vwUasYiPVO8CS4alYCW/IMLK8hVYOb1U1id9xvsDA2a2cC4QYV+d7wzgEvqH+hLU TTNCVp5HpWHUdjQNunpFv67PrMwA6vJBjGj9P96OX2e6zf+Z/YDAESeLvK6h0ysktn1A4lm7 mqRK3/RplR5YV6sLcnUfFt+LvhT4nK7yEgZftVI8aKZuXrK5Ce+MCpquSFB9gmJ+sW1Ryo1D 2ZAb8od8NKRgKt4nB3TmSaDLVwDDvGnJXB1fuyNfzS+3/MCXOgAKV5YUH+Y66p/xgNgyPej/ CqZnGnlyxXjCIvDi8Xgb74EjYbDS1pV6ED/qlEWnQXB97uYhiuR8HciADuBKis8yS3L6ogLu AgKwvmIOX+Ceu9wTSQBAWWK8UXtmfH2kFaaQc0uC56ttBpvw+fIVLDitsN8V9b7LzSvM0CV0 nzzgSo3rEpYnrNMlTNpNxIHE+1EmTgXY1+Au8RhF/rdleaYrSKsxMnbNt91W92ryqyTJzF1f SyVwZvTbmom2n5dVj+r9+BTljKixnt6XtBPt/QKmAgs2dqqIyQ5etGLelzwIAi3gqFHDbese TRQ4ZLQZ0HuzyPSl3LEzHrWMXbEbzrQMoaxFiuxB4VFcyDScurLVwwhrZJ7e5XdQmuP7Eq7v ooADAU/5PIGub/ZmcYxESN7ncnaxXObdMapCO0rRff9LqgrxiqCEJpw86muHxmAiSUKUAt7m 0/7AirzeUU5/YQ25FgKJ7++qV4Av1tdAXnF2X8Qxw8jxnlr+iy66zAg4ayaRLfe5lc57S1BD xxc6ABVuYIikntJ2toC/RhJbweZ86opN8gdg8axcWVQwi/SyL/q6XhPQBCcrREkkoJlRGZck Qp8AXd0jeC0bX92dMKfnucnySRkVPnci2CFL/tBDvMWbp613ssmsry9DqoeyrxTcp9r4SDS7 QWdr0dLwF1nFLoA7dSp5gpJ/FhI0oMU2DvospKXovEqvuPYS8RT7rLG8RjtNsT+3l0Chq68D BCi2FgSo+0SgI3yALgM+ZyXwhSZajbFwj57elw0BmgeniN1nzmX1qpRw+8yyFd8KY+Z90uL3 0UGqmSQrI78p93URPlb9cCH6a35SS5BoRGYBh4Y6i8f2tv0wrSUzBGaIZEGHL4/grR1ixHez L4s2JEDZ4P4hpzv8+z+6XEjJZW5kc3RyZWFtCmVuZG9iagozOSAwIG9iago0NDk1CmVuZG9i ago0NCAwIG9iago8PC9MZW5ndGggNDUgMCBSL0ZpbHRlciAvRmxhdGVEZWNvZGU+PgpzdHJl YW0KeJzVXd2TFMcNL+dx/4p9C67yDjvf03kjiZOiKqEMPldeXEXhw3AEDg44sPnv0/O50uin bvXuYkj5wXXNtFotqSW1WtK+3e6zfLvv/5v+f3m9ebt5u82Hsfl/l9fbv15s7j5y2zzPitbl 24tnm/H7fJsXXVbm27LJmqZw24vrzZ3ttxf/3eSuyIqi8t9cPN3c6fqh7y+2DzcP/QJl0eRZ W22rps2qYnvdDxSZWwZeLQNl6bK9B/9qmXMYudr8Z/s6DVkPJMt7IGtsF9Q0cEW57bKqdiXb e9NlRdNu27zM6qIZwP2pB7da3m+qdGU30eJ+/0WZtblr3DR0Ow4VpWvLaWhAq8r2rWuaaeiq H6qzpnT1/NWTcSivXd1OQ+/HiaUHX3FYuXNNPQ39ImH9Og7t/VDHhyj413IiAT+jeiO/erds ewb1Rn4E5oGvyBZLHdOnKqaUqIT0RYTOVebyA22mibnfT60T4lISFWzouh/qsq51Ra1T4pVc EfDsdxODABs/qHi1tSs6Ljh1LgSH0l4yG/AMoKWLJYUO2PhCshFsEXCDEBUQopayVOt4ga+2 I3hXu6alssQnAjq/kbBey6HnkhuAqrdyj1fqhjxxGhODan0/gNmEQTXlBj9Uw4q7smy9mi62 u9wbgbpzmo78dNj9rOrAsQFnZMLPa+WmoCs3PY2qTu41cEaAAgGwgHxqCqTpv1pPpF/ZsAfi DzSPDftLyeLx3FASc04CqZBKASz123gU/LwSyGqApzZFDiaqaFE2AGZdqZowDF1TaFGB0VTo gGlItj9ITJ8JtKaRrlxGXlAWH1QnULm6CnktlwbaiGwjqnAJcKA2n2gYRLS5xEDSA6jkA3eb TuflraQU0L+M3ISLDEWsGw+4Li4l0FQ2g0jYU3FOU+EHqgSAfyLBAycGgAfH7ZMEn/VDRc// lp2kovekWmbFdnneeCnkZBsvKvPAv/qBP2+8L79vyd8r37/owbSdB+TRcxOgb3pAdx/lOfX6 55O685erYr8vhi9/lKrlVZDgFecU1RpE6zq615k8Px4k0elQnki62jB6LbfyXDlYuzrPhiUl +Vfk3fdLuHZkVNWOH4buVLtlRv9lwt2KHxewP83M8olhM1vp8kxgtZw78rjo5gWcMMWe8aVO NbM2Q2iw/hyto82sTk2KA1CKEi3VYtP1ANtvKasAPkCePsm9ga8AUQBG4A4JkHijugL0Kx17 6nwAcT1a4SfcbZhNlk7Bx5C8NJzHBBBghVwNWGVwf9SuOgwD5baFlgNxDjLtRlIu6Hs3jFdq TCNwksGWNeqt3BWniTjQPcDDAMcXSJcW95n/0CIY4CbKHYmgu1kHLmyqsmBfHQ0LKIOnCj/4 hsDJB1SfRKcejKwKS9fcEVQBLNseJxPmh8pz7ZHofI3Z0SBgdsCBeWVr9zCv91m7bwdHssy7 jn7JPJ1d6fZZWfRfet+z7prhy1pu4udvD+t4Uu3zYToyudQ/uj8C8vjWdOv+vJaLW3kzEqN0 jp07HlIeWNT0lK5KFdL0kRemikk0A36rGJbWb/8QR7+VmunFiBUNDL6REw8LFk7QpdXnPVPd DjLvwbBlz1rncrnnisqu57GHlaug/tGPuF64lncUgNU7nezeAkXJF97NTxLP9+M8fzDqgpNd xJc5jZ9LHOSC34979rvJWUCbzwMLvl/EuHJih40F0bKwILqYz3uS0T9Ihg2ni1n41ekegIr4 Z5k1rphVg35bmv+dXau+AatwJ+JCioyuqxUvwvGvgBh5eW+Y/GnXlERYNo1ucvUn6P6euZg2 gCnwV7QwM7cO4AYCDIZ+z6R4gYnAlz+Qa5kICAGwB04T8CIBEjYXRfflKapHh3ABLHA5Oe32 Bx5BlahmrWOlX4bopsHDA3DCTFd/wB6wH3LRCCAPwpggJHqQ3eXEgmsTeGMDFxkwZAvCajFm uCHbu5iI13NYQCTkXQ1EXQC5tEg8QxRcwpQr5a70nmbl8mjIOOgXL1+ROMO8DJBEFgzJ6B/b A6FmI/nPEST15MGpCF7sQ4GZrZyoaxpLnIl/Zcp1sIXQ1XmRd0Pt+s2HwIJa7CcaITrH80LA hHyx54UICYHcAFTBRHDyTfx/qShYCL0/QLkUHLDrgLoD4IGusYWYgNlXlJQmcHENGH5/1N7R +Ho/35G4s+c4cqmv6Frc3wPvUd8J1YfVMTD5ZH8BkQFH7KX8Snu+iXppY0yBiaTcEgxXR7y4 hLwGfnk5VRlJf/ZLqyFKKU0NcbIAbQLAv5LgAV4AfFIWFp0IrBWIo6mpbfpaAIpu+yMKGxxc PVWwrg+7A5IHqK6mvlEdpPmnfMGTkhn4guB1CGSFAEUPMAWmZSf1JzAHJvUMHHxbjp52mIHy DLw6BXQ0sHlAG5DQO8u/5WHp5LzQeGYtvY3Yct7CaXoBBwccNLBHm8+T9KgTWVFTXOBpJ6A3 U1SW8VXWT1wrrogY2R59bWTRYzQyIKOLBmAUoDcQM8CCtyY5SKGU3P0qdUV6wqpW0h5Lbc5D mr4BhAbO5mFbHXcmC9dmlQfGFNUU33W1q2JxNJA+pL6fBeIEemgt8mz4xcKtIouS++2m4Jrt NfMvUv5G3rX7rGmrfAjzjyln8z+twvw7Gufvyi6UFTW+WrCcs0f9h/w9Q8O86Z/qqoB3k1Kp ENWKNncuohUB3uQ6B3IW0gKv/+6HXH+Wlgep+wL6A2np/iaJDu5SAHnbbQL49uBYgidyW6GE Fq+HSdFHxuv5toELDibaYoA6cfqJgcQZkynQQgose9iW8qsEQzhlnkkUTMHf3+V6wEqBgg5b tBl8BTat3Wf5FsElQ3tZ2JWFt0YeGLN4wJ22VcvZZF8PBHvFuzY2RweVKax7UqP8MNKNJpZL TYRC7RSu7ek3ofig7p/O65XLTxVOkstPV9Qu5XxDtkJGMBEIg65nI0EhsG3dWY9sCLg/Nv89 nCEOKDE7l8Cmg9MHBACcdwAewJLxXTBPD1dHMo0BrGO1o77pSEAkIZ8VazQgBrb3G+AigYu4 7t9GwghbKVImx9gWEdXjFqBwtHWuyXXa6JXKFHlwRGyR2qTiMnEL4isCtwZsCFT22gj92KRv wYq2Cm0bO5TsGz7PRgiAvK14WaoBQhpwLwiow+EVBFzSOWPBwxxwA23aVq3PpQvekwtOfgNx TkmVTMyp5ewJvzsEnEUwlKb7wlVplS4Y7Cp6njxwvcIyclBs3Q3kQdFjgaD6nO5Hz0Xm8QXX V6gXq6LCUHxhhJUQWIglNZz1mQhEW21xBVEGzdD+yrV0JCdF7ufY1HgJ6bTkf47Cacn/HJbN 6dCiPvm+zySfCZ8S9aHtXmRZDwg8pDyWRdqeAFBAcIHaTirBAsl2x8U6YpvW/XS6nu1ykrTp WOTBUXtru6ev7ztrNVzmXnK8Mft8ahg2TfnMb1/ni/LyPiofTdYUoJrqKy6oAg1tSxoBtD9J +cIWRkcWzXJYSdoX5LFSBn0NMXeA1onql7rIpnjxqdpXKCK+nZO078j8M0VJIgkunyNK8jWo 358kDkkFFpHsWNsjY0LsOXq/ObfzE8XUFnkA6sT2lmvzTmNJ7OzumBA+C+SeAJmwpUWCHZlS zpM6Wh2ZepL29mizXeF6Y/CQURs5ELQ0iwZPMDU8InQjtyJz+mRvAJD0LacB60C4xFRn4VxW uRokGuyaJqvaqh5qiz0NxtriRspEuLaYtmABJcaNJFZ1GJo3tJMy9stIeK93eV47LzwOVrI2 gRLYqXTWf7UUI1+OQ7Q++WpEgtQngzJSgAMo+72UQ2Dih2XBpXb2nZz46wF5ZodXtmqs7G5X fXJG5udFz2TeyugsZaZ8saA9WfjzQOoL1Z+TzlaghJTkmlWUwtEnz6S8SmCDIr0jgImzGQQ1 D1kolFBflvPkvoczxk7uKgE0USX1Hsg+BpZOtT7HG3hb2kFsYSBUtlQ2W9adTWa1Xmrzv4cT ByLXDG4VU0ol6UMGw0daTfCKAnyUmEcASGN7sQEZ6aBCSU/YpZQHt0gAXs9er7zWz1fZ6wnn efVCEvHLwdMMceEqBLiUUAoTYcCxDKfgw33pWiU9v9ykVaI5r9J7B81xU1SL9N6PyZU1Vcwn 6Zc+g5md551HIas71BpwV/aNQJqWO6etxOm4xjfhTi0sd0Lr1LLAmpuBeBLHmoGwVjvfSVDy dQV4rlP7HRq5Ix15ZICd8ZizfXJvex+74GSgriV4XAHtcD5IWKARjO4Fn6WXj04uShsAHrji nyQsSdQJLRr3MLUK+ijngR0C3Ccye2VVsKp81vIHMPGxXBBcZQAOd+VXH+VXYCLo3UPOWEB0 H5sIcaB8oEHSzagc2EVlkJicKQdvNb0iKsd0/Ko0XYdSWsB6vP2xXxQPTMeiTZWuR/3hGVpE H1CnP0DD15txiGZtk4avq+ay4vYK8up29T4bPktr5erZPXwY/3mMcxav6K8Z/b6Yq9n2sEro rfzhTVu0pOloGspnaKOoOlHAqQCpDyemvIiEDOAOxp15SiJbXz9b6XRCvl/A5T71dzdSltfb v659peidXr+QRJKS9aehs1TPgQtnSh21cuhWScngtS18ndVyN+DPX+TeL8mraE7c/0XT1OcS MAg2ndbbJ3C4wNnW3/pBSyzpTkspDgiC7YkPkM2mbfSabhVvwDVbDabeYMLSRgdWv6u/J2LD yPZQZOuVYXs/BeBBYdmxL/LhkCjIcU0r69KrmsP5CwC6Sb9pZcxrJ3lX7ousyWvQ7nL92wyh utj1wwT2xEc8/i5lApwuUPoMfixh0K/RX3S4PdB1hnVXfmU7duBmANx8XYMIn55LtJ5TQYO/ 9yWqo/XqvATnZfx6QHkXvh0cSpoPb0oASb3eg7IelH0B22d7iNF7mdn6Z0TSD2zNeoKVAlGr c0yFEaBq2D9c2W/gH9o6xKVTVS8eAiT8o30ra9MlkLIXaY5g+q3FcF5aAHmp/W09iBKqfGHz ItB4nx7jl3I/atYbha4XMoP2TCL7L/qrZ0AEw02+9CJJvffpOrUNZ8sFJQW195jtEnDipYgB DQYqS22ZtvbfjwwcfvWtits+GlAAekzfFt2Drf0ae6YHVxOt1mL6I8mVtaRUzX+cmHAryx10 VLSm1bBjka1CSOlgFvB6SJ0Q+z0Wzg0QJNAqnrkUgesNOLpgQ0m/A6pykwm5dPFBti5IODsi RWD4Y/ol5Yeb/wEJITaAZW5kc3RyZWFtCmVuZG9iago0NSAwIG9iagozOTAwCmVuZG9iago0 OCAwIG9iago8PC9MZW5ndGggNDkgMCBSL0ZpbHRlciAvRmxhdGVEZWNvZGU+PgpzdHJlYW0K eJzVXdmuHLcRRfJ4v6IfZSDTmt6Hj1mMREA2CRfIi4FAurIsRZItKbJj+evTPT1LVdc5ZLFn rmzDD4J5yepikaw6tZDzvtiWVbGd/jv8e/f25v3N+6Latx3/uXtb/OH25uGTUFRVWQ+hKm5f 3Mz9q6Kqd2VTFU1f9n0ditu3Nw+KL27/c1OFuqzrduxz+/zmQZiavrwtHt88Hj/Q1H1VDm3R 9kPZ1sXbqaEuw6nhzamhaUK5Hcm/OY05t7y8+VfxbR6zI5GymogsuT2xljf30JbDri2Gqim7 uj+Te/ikbopd2XahmfpvRp5D1w/FZuzYVLvdXia/mXtWlSTdjn8PfXMQ25+nLrty6EI9HJq+ npq6supCd2z6ODU15TAODIemf8+9+iZ0R1ovp6Zx+iPTx16fbK93c6+6Ce1Of3E79joO/GCZ +O/UpLl/M/M10hrqQ9NT+8VnJ75OTQcZKplvpuWfmNpUddmFIDsqYc8iDvt+odvNHX8LKA5l Nw45zvL3s6jHKTVHNr61zBa2lxDPkdaL04IMumXXhP7Y9GrfdPifO0tk/vskl6HjnwI8frQD D2yPA09sv52bdkOoI+QBrZf2i9/ZpudzU7dcVS07sGl85L+3TW8sq89dC/jM9vp65itUZ0mA ge/swA9m5QHzP1jZgMUGzIMF8k0RbD/A14H58XANFaf1jeX+zFdv981iW07aoLOzBmeu4x8E bD01PABSz22Tb1fmzFA2vbKTzpjhpqnHRRvblC57anfBedudLIDY1a1lprYsd4bWYEQLPjhw 4uBEfiIzbcsqhL7XC9CO0j7SAmoBsPUutjeTZ0Yz8XerMw+9tsO5F6AFJHFQtmOvGljViC4H kgCqApxlQAvIC6wQkARQkGDgd1b2dxYmvKZzlMcNCBqoMJeyBaIXU4woJ7HjlIEEZkGLCygZ q5w+Suv/lMIpSZcjP6lxAENczSbO4dO18wWCEjs5ov6FImv4tBm4EPqyGTfPQV+CIwqOgtgB dWL69YSaBmAIIloOYAWgAUpLXnwxqLm2/YiH++pgGxQe1i7FZnR6+mbYae8D+ikSO48nbd8x 6qZsttPEJ7wwdf0TNcsAn0pjBXpxScuB2R6MWsgI+gJKCTR97/qW2PqDWsJ+klW7MO8pn6YL bp/mdmZvXMk2Aq/BAZulM22rE8c+AwzMDjh+wHoAowloMaCkWQVWAFiPwjaBLwJHw+c5MYSk TaPdkmomWbhdfOrA4hQ32PHFAiqP2WLNlg+fXd1t00w8c2E9MNCKC5wAQOpH+0Fr18Hm8Jlf 62EAK1fKLQSm9uisIcGJ7PhAACQY/lE8AlLv7LgPli1LCgAE2wnMhmEg7DyBrQuOItOSGgz4 FEExK+HRq26Pjtj/5oHStfDpZRaR6st2CO2xFzhlPj8iQ+PqCYGIxuH0jyqoi/iDXL1K3QUk AQKcPscIyAvowZcu5UI1SWL9Od6QCwQ45Y4e4NQY9JlWaxcbrEanV0Pu8We2lz2o4MgDHoDb FcfOCU00i9m6jEB+Ed3EGV2nHxeaKKTt+Anfip1yZAWcSB94yvD/u7IJoYu43mAHA1rgNPxu agIejt6uwCEGEwIi9MWrXUf5wNaIrLs0HJOrQUlJma5mFLDlW34w0KdqQS8XCGWxPi0IbjoS KJErqxBMumx5JGc5HI/URbFjPR1LCnzPp01A7BgoK4ASxRlrORNgpQFS5awmLA5LJmBYJk5W sEve6o0BEDgJigEIvgLrdJMoTygGRHLAhvWdSZaq0190nTYwjs9aThHwAJZTyKaSvTQitmzt 9+JCmhIT+rAXd16B6CR5HrVIBFuBOBnc1yLwBUUuC3dovsARBTiHy34dqNnzcKokAMrYlUUD WsL6uPpToOmRpWKxHxjnU25AKwJfFZDnA018BWtF4JnyDAeB6BHvKBqFSGbL9O4FLq0v3QhO qIrNguOq/n5Hj6Ac4vPxLzMTMOVnsJVmFWTIsgKtiRoC4FUCvgD3fKESu4UBz6Q9Zvo8uc2A rgPaSKAJEJjuLKuRow7W0Wo2cCoB9Ti2XsA+qd940D6hblwxNV+yDgQC3sijmXaHRZZOrH9r JbGwB9O0jzTB0QaJO99B4/5WIm7uUyYsyIOccq7LwEn5CMQOwLiFiwlElZXINHsx5WH7UJ0v Mrp2hj6AyN1pSR7sniwbmB1tBJGTjNxXqpzJ5+dZUmvDbjwaYQGjxXXxuRzshsxMA2dSpN9b LuJ/nG1QZfdHxwcyKIPBH4gGgk0hamyTzhJfRF4RJXUVD8EBf1JueWAwWRxQf9FXNWEPhu9U g2oxXovUddHUCYMvmhbfhtfLkmpWedA3kfYFKwvIZ9TO6oE+4wtYBYsGFBwVoeVIB+W4xgKT AAjS5yzyStH8WlgAdkHkMF5pHZkjy/boFfapXKvOXWg0L24IUia+1AQ7+7AIaW3AShKOdGaI H5yrhSsBakyBDXJdIUgKIVH+qdjL0BT1hAGGvOwGj2LGLYWrhp9WXvIvsRJFteawQjKx/3LE wdn7ZqVhjfOYzKdIcVD3m6sUoLGAomZOb+Ts+NAvq+K1tSd6B9txFsXysyAFC5QjEADgSiBb FUc1F9XqUFZNowpAY3WGuv5TlB8FvjcBMhZKFax12jsHoT7ZBNCA3dquYg5f/gSkWq+R5owE jXIqZPjneXJeTmW1A8tT14mIK/C1f67UdTzmAFjwG5hkvXxGOZzWPIkrAwB6Aq3i83HvDI+u 5DrIT7sUMDhWNF/lKzD5VV/C4ter5NbKuhAhuWIBjLYcwnQVwSjzmvPlc+0BeeBX0Oir2ZSx 7A/4VkYMl/pDeeU9TPslyQOPwYf1wYRWRKCzrtmB3UKqNjs+kFTa6E688hg7Rb2SALpYJZch q4RSGqFo7XqqhFLzAPxzrgXIMU1deUuWNPLiat+d8odWZUSvGSerQDQtYPOzIouJQwUU3upL msB5YZfCNPkD9k5cZwEDL6tETGI3MBBocEDrc1Tqg5j0Kkl5ciQ5FWnHpo2x8NE7B4vc8xQK Ojb5UjCfbC9xnLKuc/wCroG4/HUQK80Av0nU/MLOOaJAQRI8nv7UCoEbXhBuYhdeI/ckwcH1 FYKzOnPdxEKEgCNvTDafFi92ByG3RMiXTjuxUKuB+HVvJtOaq9V5dGqu78EW8VtQiUv1JM9u vXetxfTpem9VAYCG4OsAAACkxu9pJsp/GLZeU/VN9avkVMnuFVhs2tmGErJKvSRdx4MBQgtn pCpAEi7iV2XkCfRJ9ilenxrkCWcb1XSeUVe0wgd2gcqkegRtKh/K5u8syLh9YWSwqZpt2dYh fWN/+eSb4y2F+Tt/mZc68eQbeDBhv5P0+25MuQF/ru5HaQy73KcI3BmC0O17XvoigdwX4MW2 f55lcNx2wC8CtO4B2chjmIFskjcxfrS0fPEY39sMWRXgCe+IFXAmUScXdMKtB/XB4IuuVA9g y/fYU1bJOFB0+WvGy89lNYArEA/cIeDJUaubgwwuSzt4Pa9lpWByHJjuNXMm/lgjBiegvAVs ElAXk/NwiAwvcNcRxDMSx5PlWLTdvfCpL1Nyr6lf5mNo0OAL/wAJ+hSvnXX0gaidvmhZNaOp 3uZilSp4sIq26n+01hMAFBa90+8nsf2te4n6FVWpu2mqcjuMI6+OY7ZXwC9hGlhFdLuNdLFY aZiYrhZ3ZRMRaIBCrpxn1BP8HNWu+otg0kApgIgHubeS+dRLRnWI5ptfYpPkr3GJbVFIasxq mJa5AjYjYVbn+VznNoGmFS+MPbSIQPnCGsggeHaFyIpQ+SXF/GJjtubLEW+QX/NKFDjw0GIi ALk6v1VY255RA6QH0quhqhdwCcBAptE0razAZyJFWJiVBZ2AUEETeFSViyvh0ICKmpya4ZRn 50vrAu5zH/UDQnVVq/WU0upq/9d2oKX+q0qq0UzSdXw7rEP5DpBq1ZlPMIsfCe7mpGn434H7 8NUDowOAMgSYP+NClW4ChSe+wArAHq+ZytSbDrxELL/1k50KUAigQIQEHLqFgU5U3AL0CRTZ V1+YhTqvym4RIu7GOe+GewwR/3Xqov0S/rsaibdpwSs7ahF/oAZoEiy4YNNKwc6fPx7NPdYK U1Od1P1KpF0oJ6in7ounfLp2uNfYdOJk89i0XI9HZlcBUp/nsdXr39vJCV1z2V4atE4k1jjk SgxkTCRfIrHXECOx69XXx8F68F/jkWLzvarru0Lqugqz4mZQItTsep+SL4MP5gGP/ZowiF4Z kjgPUKfQjxQULoNn23LbOYJnF0cM/zZ1Hb/WT72MwYgYJ/AGvDgkxxnaurlz3jMGM4Cuzbil o5n/aHjIcJo1qXl1+knmbf3LDG3200RPAy8yf5rUz/TWuGaC528TggB7jOdvzZMJ95y/1axm 5G/1QGrkUolynr814W7Iav6DzYXd4+ytoOTzQS4vm1nFmYeIm83qtPQLxxfV8GuZ+sweKGe1 AWHXM39r4w8XB1nFKkUemvLdxLy08p+ibp7IABl98k5n5Bz4rppn1SPQqTCO6FWaiIZd+4bU pdlpKRXgH9jfe1wbhUouUypWdT8bJ6LhDmByBPmL9HMdqrIfRybhSj3qtOr880TuMAjIOoMf HQBBh8MvBMlKOfCDpj54KJYocsOTw7xKPKhbAIy26dqRUtftxTgcpfNk6qnh26UVTImCErE9 YJwoGpxdRDekQwgMK6/skFk6fq8aWxnHIq4u1QXFK9m/jJihpfMCBsrT5W7viqLne76FClY4 46Wc5AN+KvJU2oVUU721gOl+YoNHQYJf8hGkFj8kQyos0jXyElixqAsM4BCwD9BThC+xfysk 9IwqArBLN03Tll2z8JdXF+oDcBm/cgMerYeaU2RWIqV6viePXGVyrGT5/srkLnsRIPkLjPmI jD+sLYn4SodcxcRx5BDR3qthKt/o+DVg98tV51Pru6KYel8jedOF/tZw4sbSZb+9rBVtRhVB MjFiHWbAaeIhZRbC/fK2eHwz/fd/fF7UF2VuZHN0cmVhbQplbmRvYmoKNDkgMCBvYmoKMzc1 MQplbmRvYmoKNTIgMCBvYmoKPDwvTGVuZ3RoIDUzIDAgUi9GaWx0ZXIgL0ZsYXRlRGVjb2Rl Pj4Kc3RyZWFtCniczV3bjhy3EYXzuF8xb5IQT2ua7BsfYzkIBMiJpazhPBgIdNcGq7UUS47l rw/Z7J5hdZ0iq3tnkcCAjOWyi8UiWZdTRe7H3aGqd4fw3/T/l+8vPl583NVj2/y/l+9331xe PHzmdnVdmd7Vu8s3F7F/vavNUNl6Z7uq64zbXb6/uL97cPmvi3pwVdM1vs/lq4v7dWhqq866 1k5Nh9D058vd04unfkxr3KHyv2u6vmrM7n1oqKv+2HB9bLDWVQc/4vXxm1PLu4sfdzfr+PdE qkMgspzAkTWJnLG7oWpaZ4k42qYazLDra1uZoR/JfRXIPXxW1+n4jZea6+xu70VvDgczyuRJ 6NlXrSc6TGJ6ziX379h08E1zr7exqa1dO4v8Ne81zstUQ+/6bmp6E5psNVjXuanpKjbVtetn Wte8CZD/JTQ1VeNptemI+8ZUxvq2fW0q392dfrWQ4iEM4fp67Nm62PEPsSNZvT6McpTRp8ie sa6fZfSOi+11ZM95GTlZINecFhDIC2FVbOC6r+napTKaevWe1tTyhZMCbP0nNA1V7QU19/qZ f/iZNyVLNy/KK3HEvnV9T5n3I7a9TAs0Ab4m7r0gLBlx/uHDaftaUVCA7ksu4Py+NHLTFZ/J DR/xLecUrNbPvNcbNp/8aexltoDM5VlHLVPY7vKS82UAEgCj8+/Ign9KfwBn9XRyOroWTPB8 K3R8v7SUq/SMvmbfJVJqCLujxAHTLZtzN8g8fOZNQKSAPBBgiaNE5REbsLfePDV1QzXyS778 z7nOBGcOLCDQmfJpKui+91F7+F5mccpTVmW+UrkClQz4ujoZgYXG8OapW/DlleaRr4rrUbJs l+GHLhzwJrP7wZmXTrMtyiidqm7tVsmowCTgi/ANLMMqu5Yu0SuVbtbZNTAVwJdkNumRS1gF KhyQb6lwUqMPRjwZjYXN4AOKOi9vLIAbKCtqoAWzJiPznVLjgfOdrq5kOfbWmKrzbUQJGr7K f+HL8E1ooo462GqJX2roKhf26AuVjQUfgoMClhbQks5OZvNLcqUcWf6hKYwFnGugbIgLIZ/z vFOyypil60hGB2YW2NJPfA5guTiPOt+uOkpgoLu87ruqbkLwZatDnwm+9saHwLYfxp61m3rq g9df+WTArvmNN4EJgijsn6INSJ0MMCJo+hDNsCdPnKtFkLk3zi+g31AxGs2JbpZY6DnYoRy4 7o+hbuj5J1HRpzsaBC0gpv3EhSJHKKlQtkYoYBMD1QBOGjgQ4KTJmgdoudQ903EPyIMoAdD6 IG5kED8FWnYLE8lqAHfzdjErsKwzFWBZdQYI7FHgEYGp6MhLZxmymo9jgX2QN2QB30iE3nLm gW5arJZgQznWNsvh140e1ZfCVgB0Ade/8SY+OnAWyVBX/PelGF/nAspBrz1U1prJ33Op3wPC H8F0N2kvqnkl/VzcZCfb3RPltbQyjfc46476q2V0dIo5PZ+NoVJbj44C7ZIcNgJW9oGWbdA6 gngLKP7zY6wFVxWsHxDUbWFXcXdvBUHBmQVaC2x0oCZ1UgHkZeOULilCGxWo4Z27ICmPqzR+ AVoSN7qsC5Kt2qFdogP+AMgoZWcomAD2BFk0oLWnpfdLZIDVVAkMBMxlUgUEVYZxU/Kr7Iex le0XeEFeI2ZiuEQdAD+95SyXIYR0IcGxBXCSGFeTJl2UvBnHAHMEFkQnVR0oIXu8gkmEiFTe WdoU8KYcyQFvITzSOck6yARsbzl9mNLSCQdMG+hyXbiQyCujPKTTAwDrjo4lOIQZLQxc51e8 SYe9ZlOKGDEFvwcCAX66kBPb27qvamuJAw285Zenc5NJGuVhrUYWTinHEn84mKqrUcnB3jof Bpje/8o7xfXQjb8a+K776UFoi9CX5/NQj741yGmOEp5n+n1ockGchiwdBYtHR6cLjmxD9C6V x3P+4Yu4Pa1zRIVH5IOA/bRp6nXoT/jIJ75yX+KMF8FEGJ9uhM5Vtm3MiBTaOgOMHaX/JCCJ 95a048LOPOqSYzqgAAQPQLEkLmoGU8xb6oyf/IV/CHoBYwTmCHTg/08IU5R9yZSqUG9doKOL wkvBGACtdAUVsj9YUH5ygKbhu+jflOweAFDBngCgJ2hSwZnA6wNuLDAcgDoIQRK2Mt5AqfwF 2HnZp9AAexs+oX7AC/4hWFo50mEY5T6UTzatoabyhyjRNFb4K296zO3fP2IvsdRrlY/JVO78 +x/DD/Wh6oIhWfLSywx/y7n7G+8FqP892iYv0GYTPF041reEp9NDtLloLW/ArNxLMmAQ80/P KNA5VbrKUwLTeT5NAxKYIWPZ1mPGMnHlHBdA3pVLcpg6hw74ZcChA72AYwaaXnNaskeXNv0e P/Tc55gAjuYVXwjgtN5EIRIPbnRFjaOZ4Lg4LgC1xtB0MnISc54hoQvOpc6Du2a7av4hQSga vncd+gSUtiXO2Xx+wbnXVVHrqt3AKeclCUUHA2gBOaUo2rJsVWWuClfnOsjZ3kJNC3AnNlfk 6LiXC0HYaaMiFgt/ZSorfE/KEdhNYGtIuhmkqWLVCUmVs9IS21c1rY74CmgTeOrPGRYmIUSm YkSG7ISqwZnW6hqVDBMSZGflLfM/q8TYmgYBJVVSMgTKQKhqXIle6gI7HVSuiz6vwULmS+Vv G8sBeQNEQTRaqw4ZoCLfLCjkcnVXTnSAOiiL/Z2rRzErdZYgMNGhJLDZm8FVxnUIsIRernSj 7hwY5nejn+U3UDvMUnkcaXu+58mN0ZV3xnp3dMYe8SZA/duplxc68eIoOvnLaUCiIMAmiq5y Q2mlm38CTb2VsTOtz7EpOq+EPPCLWRW4xi9OR5sCGeOqdiwStF5B5FHTaCbzqOlUQ+hXJBAC cSxQfxIQn/F3QSyeB+cyGTtdxaYuZgZQ2Nc5DyVXlr0iDUxpAXMiXx5J9xFgAmhEYHR0yPY5 LfoZy0fPk3j9btIgXl/MMxHRnxTqeRTPblochewmuPQB0m6p4wWmBZLtqoIXsNEfijawULey toIqUzgIbOfWWyQACPrMpSwfGVZRWARIM9fggtFdVHQAF+5ukPAuU7ItB2mFNNRP9zl9sDVA LAekBHpNKlbMvuoub+jihq2XpZ/xc/59dBRS/HbUIi4ksuu5SUJ5N4HXhYtiwADkRTDT2vPT KVuhQmwCFBcXuq42c5URWl9dd0YjBOSgy4gK94cgvJK/ULyqNIMpXZiXBIV96YgySMfu+xLm QcGKXE2bv5YoX8hmJSx7a03VOFO8oCwnAwtgrO6Vi3wWpqG0UhA6j2POH8bwjK6KhLLtu4Pv 17jbRIigaVuEyMPBUeqeM3sMIj7EdUhKVeRcRwrucUogIXLDiOvqZ0AMCYJPKcijAeNJLjOp J9ykAB4mUmmA+PbUi+gfOiCg9ZY3JdIiVSeLKJLG1lOE2nmP9dCV63qOHchmzEG6ic2VPPfj 7gKJW+C6g5u6JHt4EP0l0XnJV8gCPZHBpFaUw2ZCHVCHIls0kA/hxQxolK3vleiKXUqvv0gF 5uUwhufL83e+APHNL6eAvaIaEDhFugownVOk+1B370EVnYJQVHfdZkOmce1NxYwHdlIomStK oleku4IAbmGAbap6HUEqegVQ8tirI4p9qb1t4x0zV6/V3o+4sfljNDbeVjc9bTpPECVWAE0B nd8KjePbDtz4yOQOdUlnsIXBvW0QaHJsqvSaC9CYunegJDu0iM7XWCAAs+pqM3W64pzwb0mG uqd2dA8Jgd0APgTmG8zujC9DrIK7Mea6UJr5mA6oODnlfFeVlfknfTLkQZlk3nZtuI2x5SUx GbBsfMel4tZdQSt5Yckeb7jE5ibw0KR0FczypdDljzKT4IVc8PXGAuK0xqE8Bx6ZYLKwuufO wNjzpFt09V35grdWlqqMoaUrr3vgTCyLQ6EZ1mzrHuXYcCFw9Z23zHlYdect7fWQ97rt2w8F 9QtY1b2CBppAcb8IpZFzGXGuwa+zZ3X0hE3jFard7c2hsuHppJOuPYJqpvYrPsRitNGX64Pn dnSBI17VhheCCN5YH4KnXKdzbYNjTa+vRhyoXtDy42V63UzkffAyN427nLqdV3xE8OGo0MN1 rtM+28UD2baLPAEl/47Tmlj1O4hiZJRWrBLxKxXsl0jreeSrOQFFN1yob1mnRFok+c4utaUr OK137Sc41is3XqcMw+lX6S22ywf3thcjwneF06Y1hcR3pytTjsA7J3kEfN27TTvVh0A777gW XVMWALDFNKyVsEVqfkHBFZiP7to9CI3EYA6l4gqOE/gQJA11a5bf3ZnUIjgpqhdOgHnMP5mb WY0VLlFRgrrLafmVlTEm+aZXIYMm3mNb5SFIb1rRcAl8yINE8B1Ivj/mM8kf1rkJJMSzT4uk n+3t0FW9GejtooS7jN4AmQviDoLXjUHxh26/AyuQaMBZDcu4UarSgf4uxXArMASkGGQroSuf INHS5veOgFhXPQacFgyUEE3pdusmJJuroTUWD6TOZA2eMsRJgZJxgHAB3SyvvKfVGvlDMl1g qFf7NYtiSwWaNhMBUTxQvmDJADK0+fnjgu4lnwCFCpQ6QKRAwK97ygSggXcA/YkTvmXJyfgu yECxNd2tfOl9QPHeV6bkBCibJLY9mg2pTIS66re7OwWKwptQ3A8yeOAm5DpgSPfkxebs5ubr 1CvCmCib+cSscfupWJjRoITFMj55qvrXF4pD5QuuMuskiptSScuC5X2ZMqnzKsCHt7PBTbiG fKSl+8stQPeA6j2pzrn4hq5OZerehpBSlFSEeveffmf4dwdVk+VNgLz8VwCAAbNu8dQiANS9 cfDhwdDDesSz/DkN3V/wAi9kgXxDVgN0pKouc+8HDA8uxn7I8p15/0tX9LKqllpzR/LWTxWt epEG2MGUI6DwRc0pqyPdm0CrXn8sBHorSneKJdvgQ91dfpCpzVY0Z6yh+McTTNOFqhlQc+v/ vRf/mf5Q4NOL/wIrD0NbZW5kc3RyZWFtCmVuZG9iago1MyAwIG9iagozNzMxCmVuZG9iago1 IDAgb2JqCjw8L1R5cGUvUGFnZS9NZWRpYUJveCBbMCAwIDYxMiA3OTJdCi9Sb3RhdGUgMC9Q YXJlbnQgMyAwIFIKL1Jlc291cmNlczw8L1Byb2NTZXRbL1BERiAvVGV4dF0KL0V4dEdTdGF0 ZSAxMiAwIFIKL0ZvbnQgMTMgMCBSCj4+Ci9Db250ZW50cyA2IDAgUgo+PgplbmRvYmoKMTQg MCBvYmoKPDwvVHlwZS9QYWdlL01lZGlhQm94IFswIDAgNjEyIDc5Ml0KL1JvdGF0ZSAwL1Bh cmVudCAzIDAgUgovUmVzb3VyY2VzPDwvUHJvY1NldFsvUERGIC9UZXh0XQovRm9udCAxNyAw IFIKPj4KL0NvbnRlbnRzIDE1IDAgUgo+PgplbmRvYmoKMTggMCBvYmoKPDwvVHlwZS9QYWdl L01lZGlhQm94IFswIDAgNjEyIDc5Ml0KL1JvdGF0ZSAwL1BhcmVudCAzIDAgUgovUmVzb3Vy Y2VzPDwvUHJvY1NldFsvUERGIC9UZXh0XQovRm9udCAyNCAwIFIKPj4KL0NvbnRlbnRzIDE5 IDAgUgo+PgplbmRvYmoKMjUgMCBvYmoKPDwvVHlwZS9QYWdlL01lZGlhQm94IFswIDAgNjEy IDc5Ml0KL1JvdGF0ZSAwL1BhcmVudCAzIDAgUgovUmVzb3VyY2VzPDwvUHJvY1NldFsvUERG IC9UZXh0XQovRm9udCAyOCAwIFIKPj4KL0NvbnRlbnRzIDI2IDAgUgo+PgplbmRvYmoKMjkg MCBvYmoKPDwvVHlwZS9QYWdlL01lZGlhQm94IFswIDAgNjEyIDc5Ml0KL1JvdGF0ZSAwL1Bh cmVudCAzIDAgUgovUmVzb3VyY2VzPDwvUHJvY1NldFsvUERGIC9UZXh0XQovRm9udCAzMiAw IFIKPj4KL0NvbnRlbnRzIDMwIDAgUgo+PgplbmRvYmoKMzMgMCBvYmoKPDwvVHlwZS9QYWdl L01lZGlhQm94IFswIDAgNjEyIDc5Ml0KL1JvdGF0ZSAwL1BhcmVudCAzIDAgUgovUmVzb3Vy Y2VzPDwvUHJvY1NldFsvUERGIC9UZXh0XQovRm9udCAzNiAwIFIKPj4KL0NvbnRlbnRzIDM0 IDAgUgo+PgplbmRvYmoKMzcgMCBvYmoKPDwvVHlwZS9QYWdlL01lZGlhQm94IFswIDAgNjEy IDc5Ml0KL1JvdGF0ZSAwL1BhcmVudCAzIDAgUgovUmVzb3VyY2VzPDwvUHJvY1NldFsvUERG IC9UZXh0XQovRm9udCA0MiAwIFIKPj4KL0NvbnRlbnRzIDM4IDAgUgo+PgplbmRvYmoKNDMg MCBvYmoKPDwvVHlwZS9QYWdlL01lZGlhQm94IFswIDAgNjEyIDc5Ml0KL1JvdGF0ZSAwL1Bh cmVudCAzIDAgUgovUmVzb3VyY2VzPDwvUHJvY1NldFsvUERGIC9UZXh0XQovRm9udCA0NiAw IFIKPj4KL0NvbnRlbnRzIDQ0IDAgUgo+PgplbmRvYmoKNDcgMCBvYmoKPDwvVHlwZS9QYWdl L01lZGlhQm94IFswIDAgNjEyIDc5Ml0KL1JvdGF0ZSAwL1BhcmVudCAzIDAgUgovUmVzb3Vy Y2VzPDwvUHJvY1NldFsvUERGIC9UZXh0XQovRm9udCA1MCAwIFIKPj4KL0NvbnRlbnRzIDQ4 IDAgUgo+PgplbmRvYmoKNTEgMCBvYmoKPDwvVHlwZS9QYWdlL01lZGlhQm94IFswIDAgNjEy IDc5Ml0KL1JvdGF0ZSAwL1BhcmVudCAzIDAgUgovUmVzb3VyY2VzPDwvUHJvY1NldFsvUERG IC9UZXh0XQovRm9udCA1NCAwIFIKPj4KL0NvbnRlbnRzIDUyIDAgUgo+PgplbmRvYmoKMyAw IG9iago8PCAvVHlwZSAvUGFnZXMgL0tpZHMgWwo1IDAgUgoxNCAwIFIKMTggMCBSCjI1IDAg UgoyOSAwIFIKMzMgMCBSCjM3IDAgUgo0MyAwIFIKNDcgMCBSCjUxIDAgUgpdIC9Db3VudCAx MAovUm90YXRlIDA+PgplbmRvYmoKMSAwIG9iago8PC9UeXBlIC9DYXRhbG9nIC9QYWdlcyAz IDAgUgo+PgplbmRvYmoKNCAwIG9iago8PC9UeXBlL0V4dEdTdGF0ZS9OYW1lL1I0L1RSL0lk ZW50aXR5Pj4KZW5kb2JqCjEyIDAgb2JqCjw8L1I0CjQgMCBSPj4KZW5kb2JqCjEzIDAgb2Jq Cjw8L1IxMQoxMSAwIFIvUjkKOSAwIFI+PgplbmRvYmoKMTcgMCBvYmoKPDwvUjExCjExIDAg Ui9SOQo5IDAgUj4+CmVuZG9iagoyNCAwIG9iago8PC9SMTEKMTEgMCBSL1IyMwoyMyAwIFIv UjkKOSAwIFI+PgplbmRvYmoKMjggMCBvYmoKPDwvUjExCjExIDAgUi9SMjMKMjMgMCBSL1I5 CjkgMCBSPj4KZW5kb2JqCjMyIDAgb2JqCjw8L1IxMQoxMSAwIFIvUjIzCjIzIDAgUi9SOQo5 IDAgUj4+CmVuZG9iagozNiAwIG9iago8PC9SMTEKMTEgMCBSL1IyMwoyMyAwIFIvUjkKOSAw IFI+PgplbmRvYmoKNDIgMCBvYmoKPDwvUjExCjExIDAgUi9SNDEKNDEgMCBSL1IyMwoyMyAw IFIvUjkKOSAwIFI+PgplbmRvYmoKNDYgMCBvYmoKPDwvUjExCjExIDAgUi9SMjMKMjMgMCBS L1I5CjkgMCBSPj4KZW5kb2JqCjUwIDAgb2JqCjw8L1IxMQoxMSAwIFIvUjIzCjIzIDAgUi9S OQo5IDAgUj4+CmVuZG9iago1NCAwIG9iago8PC9SMTEKMTEgMCBSL1IyMwoyMyAwIFIvUjkK OSAwIFI+PgplbmRvYmoKMjIgMCBvYmoKPDwvVHlwZS9Gb250RGVzY3JpcHRvci9Gb250TmFt ZS9QRU1CV1MrVFRFMUM5Njg0OHQwMC9Gb250QkJveFs2MiAwIDgwNyA3MjJdL0ZsYWdzIDQK L0FzY2VudCA3MjIKL0NhcEhlaWdodCA3MjIKL0Rlc2NlbnQgMAovSXRhbGljQW5nbGUgMAov U3RlbVYgMTIxCi9NaXNzaW5nV2lkdGggNTAwCi9Gb250RmlsZTIgMjEgMCBSPj4KZW5kb2Jq CjIxIDAgb2JqCjw8L0xlbmd0aDEgNTk1Ni9GaWx0ZXIvRmxhdGVEZWNvZGUvTGVuZ3RoIDU1 IDAgUj4+c3RyZWFtCnic5Vh9VFVVFt/n3M/3eMjjU8aXep9XyAREQBM/wifwkI/VCIL5nn34 +DJMTFIzTXPhqKM+0PUaqZm+FCtLTeo+sEJiTY6FMctajWusqdGpmYas1UirVprzITD73AME 9DWrNX/Mau7md/Y5e++7z977nPvuuQABgFCoBQEKFxQnp4J5xTRgc1P5ytIaPo5OBCAZ5evW as8mHz6MgvMA0qxlNbevfPXSbANAjgUQl95evWEZtw//Fdq/XlVZWvHbHctKAEbbUHh9FQoi Iq17AKzrcTyxauXa9XmZ/l04fhRA2Fm9qryU3x9ZgfefWlm6vka4SZ4MMGoiCrU7S1dW9sdX jU18zao1a/vHF5m+ZnVlzYLXrl2K9tnorwT+164z36pJQSonHrqFLsHeI1CG7cOICsRD0AAN tIXbQBrCwF4+fCR1QiqsNuVpsAnbbPg7OQQ/NyVzoAz1ZWjdgTwDdeXIiemjgdSb/D7Yhr4/ py30JD1pauei33xmwYm2SJ0oZ/62wrPwHjmBNhthL+qOwxl2F3pugCa4QiYh1ZEPSTctRClh 86OfFWjdgPH+Gt6FL0gUySB+0o42EXSLGQufrRZtOpDOmF4Y3UiqySqymuxCn11UoNPR6yq6 kzZSg54UvGKG1ClHyDOUavRCgOLuDccMmbefQjHOXAZ3DXrl9DtCSREpIVXkQdKIMXSQbqRL NInOxaozekDwiTbxY2mF9DhSp7xIeUyV0bcEMowBDeJgGmblxjmKMOYKuAPuNWkj0ias5c9g PzTCATgMQWiD37A54Ry8B1ewOmFILK8ZZCZZjORFWk02k21Yj7ohtJs8SlpIG8Z3mrxFx2PW nKoxex7lVvowPUZP09fp+7SLfkI/F0CwCEuFMmGNcFA4IrwpvCnmio3iAfG8eF4ikmFWKkKO km+V65DqFYuyQtmm3K88prxgnQKjMa9EzCsfFmNWGzCTTbAT/OaqBZGOwfNInfAJywOprz8T RjNJNskhi5C8ZAnxkZVkDVk/mNGT5ClyiBzDXN5CeoecI38hfyOfmnSFyjSGJgzmV0iL6WK6 gj5IH6KP0mdwR7bQdvoOfQ9z7KKXMccQIUKIFsYJbiEHqUS4WVgvbBWahJPCOaEb180m3iBm iIvEWzH3U2KX+DGuJJUEKU6aLs1CqpLulDZLddI+3NHdUrdsM6sSIUfKs+Ud8n65RX5X7lGi lRhlAtIUJUUpVqqVdcoRpUv5SD1qmWdZblltTYQjMBVeHPH0Po+7+xV6q5wMY8g53A13CWFo pbFnj9qUasty2sKiU4rJJFypP8EVwQIF4ilYLNwM1VKZEKJchENkjbiFPCPkwFE4qKwj7YJP 6BYOSnHybF5P+rBwRNmg+JSPMNJLwl6pSplC5kl15BCdi0/0alIEX5LLcBvOvJZOhlOwC3aS daBCg3qUhOKz1kHHkzrpcaFZbBTc0mZyHa6gQ+oUtsN0iAYbTIIJuNcliEKAa0b6jGlpqSlT k6ckJSZMvm7StfFxE/UJTm38uLHXOMb8JHZ0THRUZES4PWxUqC3EalEVWRIFSiDRref4NCPe Z4jxem5uEhvrpSgoHSLwGRqKcobbGJrPNNOGW7rQctkISxe3dA1aErs2B+YkJWpuXTPeyNa1 VrKkyIP93dm6VzO6zf6NZl+MNwehOHA68Q7NHVuVrRnEp7mNnHVVfrcvG/0FQ6xZelalNSkR gtYQ7IZgz8jRa4IkJ4OYHZrjnhWkoIZiVEa+nu028vRsFoIhxLlLK4zCIo872+F0epMSDZJV rpcZoGcaYQmmCWSZ0xhylqGY02jLWTpQpwUTT/jrW+1Q5kuwVegVpbd4DKHUy+YITzDm69nG /Hu7YpMSW8lTJR7DktVKoMRzHPL7aoN5tdnZXjZbRJZnh2k+Gs1H39vlEPzu2OUaG/r9OzSj scgzVOtkrdeLTpMSCxZ6nBi17q7XWBoLPWYG6JTEJmOQTMbS5AlX6m4m8d2hGRY9U6/y3+HD xRrjN2DhBmfzmHzX8b4/Q75b85d4dKcx16F7S7OvCUaBf+GGljyXljdck5QYtIfzSgdHhfV3 bKFDO5WDOrNnmrMeRj1QasIi0vNwixhauYaReHSDxqWzpjId/OXpaIaXl2BFl2P9fH77LLYQ Upxd1/yXATeC3n1xuKS0XyLH2S8D67LtMrjlUD/QNxISjMmT2U5RsnBpMbIMczw9KXGdUaDX 2DWjAEsGhR68yTsrGUvudLJVrmt1QRkOjNoiDx9rUOZoBldygtegPqY5MaCJXsQ0tQOawdt9 Om7nY8AOdNGGGj/4F2aPiXRXzTJIzHeoK7keHx+3FhSlOH+hJ77UX+eI9/nrvbg0Ofgo+v05 upbj9/lLW/tqy3TNrvuDBQX+GrdvIKXWvhN1DsNV760iWFQjjVfDiMzyCA7q5T3qELxJwOJQ UnoLAULqAfretL5vRjb0+kAEfIfjJSNUhgzIt0yCButyxAnIV+KhwdIOTcIR6FCPQpMyAZos Yf1YyhGyA1EPTWoHNFlfhibplxzMVlyFOIM6PMEoD0C+2og+t2HfyfUmWH8+yhFiCzTJHry/ kkPZxSFWcDB7+WW4aQDqX9EuF2WncY4XUO9AhKBsGsq2II+GBjkPGgbmkv7Rj04ExizfjPLo /jgm81gsLvSFcSvoTz2OHPNT7kHcj+M05HfyXNXteP8NyJdBizUBdopYO4aBubCe+SOQPgwb 0WbjiFr8l4FnwCbhMM/ZnGck9nN8n53I7LqG2hB7v+4M9sO+0bcJUjZCtuPbbf8zqGUjgGd7 le/flO+CVcb9KfM1N9d9uN8/DPbf7kf/WJ4+HKqfY1D/r+EYlG+CDga2xmZ/HvIhEM5BuRAN 5ep8POmyJ/MDYoNk2A2j8ERsx95MfIQvikvZ+dtlg5wcNIkIV125Wiu9vjk3FdlWk5GjnD3D 2WHODnH2NGdPcHaAs/2c5XGWy9l8zjI5c3GWwdkczmZyJnMmciZwRlwLkJ9HnEP8EfE24hXE C4jnEc8hmhBHEYcQTyP2I/YhHkPUI7YiyhFLTZ/PcddNnB3h7CnODnL2JGf7OMvmbB5nN3CW zpnCmcQZ5QxcLuTvIt5CdCJeQ5xCdCBeRBxDtCCeRTQifoHYgKjITY2yRFlmBFrJOleeEjig BPYqgd1KYJUSqFYCy5RApRK4RQksUQJeJeBRJqoTVE0dp16jjlFj1Rg1So1Q7eoo1aZaVVWV VVGlKu5hI1IooAXFmaTAOFEOBWWa8WWx3kqsRUsMSc8kRkQBFJRkxhrpCQbdaZ5MWklfkJA9 2x3sUHIcCOnbvtvRz71eiEn4+hU7bFRQuKEdxpMZoGCb1qKMf1Vh0mKUBkxpgEkDpjSWNBdC akFpnW8sfIPjry7yndphlu7lLN1CT1CFTG/WLZy30BAr5uNzOL2ZMfaaDDO52c7YzY42EfAT PATfzTY87IUimCppXtI8psI3GlONYufAflXs5tlORxs51K+yozgcS4lPGb77hFp8FwiYpO4K U84S8Sx5AkDsA6lPOE4+BEju7bZ3w9xPsU2ZmhbuDI9zhjtrBeippdALUuc/02vFTrwDv9BB vht9heB38lj84qt8Cex0DoSRvc2jvpjQht1Yktci1Fps7WQvjMNxBD6AFpjYRjdDNJ3WIj8O +EhvbnYEQttRZCOpYCeZTFwxNrm751JvzwV7T88FmNtzoYt1Ll39NGVqpB2cGoSbrZQWLo1O I6kR06dRfQKNjoqYMSHcnpYabm8iQZKGFOxd0PsG0gLia+vtEO+jd+8KBnftam7etWnTRto1 zCL4iGA4brtqkMizZ0lU78Wzv++9eDd+A63u3dO7B2v3Vb6jIaEdokk+ljGM5B/DFMH2Esll P1f424LxJ/eYNew2azg0YhIFaakwfRro+K76WoyTiNiKvzK9V1tbe6+OjI6koXDAgB1h4v+v aN8gXfgREbto/yk1CrcTXmQMgh1P4Z5NP+Qfgj+aS4SJZiuy+nwm9PVhS1mLY/xlg3uA14dA BNaQAjvURwEUum/MXFycUFLiTsmaeX36tPS1U6eaVkAC7L8QP+z6DD7rGyYY+LIQvudGZkdr IWg817Y0bM5l1aGa4oNNe3sZD55WzgL0FlrfV1JwaBvw+28e2mbCCmVuZHN0cmVhbQplbmRv YmoKNTUgMCBvYmoKMzEwNAplbmRvYmoKMTEgMCBvYmoKPDwvU3VidHlwZS9UeXBlMS9CYXNl Rm9udC9UaW1lcy1Cb2xkL1R5cGUvRm9udC9OYW1lL1IxMS9FbmNvZGluZyA1NiAwIFI+Pgpl bmRvYmoKNTYgMCBvYmoKPDwvVHlwZS9FbmNvZGluZy9EaWZmZXJlbmNlc1sKMzIvLm5vdGRl Zl0+PgplbmRvYmoKNDAgMCBvYmoKPDwvVHlwZS9Gb250RGVzY3JpcHRvci9Gb250TmFtZS9U aW1lcy1JdGFsaWM+PgplbmRvYmoKNDEgMCBvYmoKPDwvU3VidHlwZS9UeXBlMS9CYXNlRm9u dC9UaW1lcy1JdGFsaWMvVHlwZS9Gb250L05hbWUvUjQxL0VuY29kaW5nIDU3IDAgUj4+CmVu ZG9iago1NyAwIG9iago8PC9UeXBlL0VuY29kaW5nL0RpZmZlcmVuY2VzWwozMi8ubm90ZGVm XT4+CmVuZG9iagoyMyAwIG9iago8PC9TdWJ0eXBlL1RydWVUeXBlL0Jhc2VGb250L1BFTUJX UytUVEUxQzk2ODQ4dDAwL1R5cGUvRm9udC9OYW1lL1IyMy9Gb250RGVzY3JpcHRvciAyMiAw IFIvRmlyc3RDaGFyIDAvTGFzdENoYXIgMy9XaWR0aHNbCjUwMCA1MDAgODkxIDg5MV0KPj4K ZW5kb2JqCjggMCBvYmoKPDwvVHlwZS9Gb250RGVzY3JpcHRvci9Gb250TmFtZS9UaW1lcy1S b21hbj4+CmVuZG9iago5IDAgb2JqCjw8L1N1YnR5cGUvVHlwZTEvQmFzZUZvbnQvVGltZXMt Um9tYW4vVHlwZS9Gb250L05hbWUvUjkvRW5jb2RpbmcgNTggMCBSPj4KZW5kb2JqCjU4IDAg b2JqCjw8L1R5cGUvRW5jb2RpbmcvRGlmZmVyZW5jZXNbCjMyLy5ub3RkZWYKMTQ3L3F1b3Rl ZGJsbGVmdC9xdW90ZWRibHJpZ2h0XT4+CmVuZG9iagoxMCAwIG9iago8PC9UeXBlL0ZvbnRE ZXNjcmlwdG9yL0ZvbnROYW1lL1RpbWVzLUJvbGQ+PgplbmRvYmoKMiAwIG9iago8PC9Qcm9k dWNlcihBRlBMIEdob3N0c2NyaXB0IDcuMDMpCi9UaXRsZShNaWNyb3NvZnQgV29yZCAtIG1p bmMyMF9wcm9wb3NhbF9kZWMyMC5kb2MpCi9DcmVhdG9yKFBzY3JpcHQuZGxsIFZlcnNpb24g NS4wKQovQ3JlYXRpb25EYXRlKDEyLzIzLzIwMDIgMTQ6Mjc6MzEpCi9BdXRob3IoamdzbGVk KT4+ZW5kb2JqCnhyZWYKMCA1OQowMDAwMDAwMDAwIDY1NTM1IGYgCjAwMDAwMzk0OTUgMDAw MDAgbiAKMDAwMDA0NDM3MyAwMDAwMCBuIAowMDAwMDM5MzYzIDAwMDAwIG4gCjAwMDAwMzk1 NDMgMDAwMDAgbiAKMDAwMDAzNzkwNyAwMDAwMCBuIAowMDAwMDAwMDE1IDAwMDAwIG4gCjAw MDAwMDA4MDYgMDAwMDAgbiAKMDAwMDA0NDA3MSAwMDAwMCBuIAowMDAwMDQ0MTMyIDAwMDAw IG4gCjAwMDAwNDQzMTIgMDAwMDAgbiAKMDAwMDA0MzU0OCAwMDAwMCBuIAowMDAwMDM5NTk4 IDAwMDAwIG4gCjAwMDAwMzk2MjggMDAwMDAgbiAKMDAwMDAzODA2NyAwMDAwMCBuIAowMDAw MDAwODI1IDAwMDAwIG4gCjAwMDAwMDM1NjMgMDAwMDAgbiAKMDAwMDAzOTY2OSAwMDAwMCBu IAowMDAwMDM4MjExIDAwMDAwIG4gCjAwMDAwMDM1ODQgMDAwMDAgbiAKMDAwMDAwODAyMSAw MDAwMCBuIAowMDAwMDQwMzM5IDAwMDAwIG4gCjAwMDAwNDAxMzcgMDAwMDAgbiAKMDAwMDA0 MzkxMyAwMDAwMCBuIAowMDAwMDM5NzEwIDAwMDAwIG4gCjAwMDAwMzgzNTUgMDAwMDAgbiAK MDAwMDAwODA0MiAwMDAwMCBuIAowMDAwMDExOTE2IDAwMDAwIG4gCjAwMDAwMzk3NjIgMDAw MDAgbiAKMDAwMDAzODQ5OSAwMDAwMCBuIAowMDAwMDExOTM3IDAwMDAwIG4gCjAwMDAwMTYz MzAgMDAwMDAgbiAKMDAwMDAzOTgxNCAwMDAwMCBuIAowMDAwMDM4NjQzIDAwMDAwIG4gCjAw MDAwMTYzNTEgMDAwMDAgbiAKMDAwMDAyMTYzNyAwMDAwMCBuIAowMDAwMDM5ODY2IDAwMDAw IG4gCjAwMDAwMzg3ODcgMDAwMDAgbiAKMDAwMDAyMTY1OCAwMDAwMCBuIAowMDAwMDI2MjI1 IDAwMDAwIG4gCjAwMDAwNDM2OTggMDAwMDAgbiAKMDAwMDA0Mzc2MSAwMDAwMCBuIAowMDAw MDM5OTE4IDAwMDAwIG4gCjAwMDAwMzg5MzEgMDAwMDAgbiAKMDAwMDAyNjI0NiAwMDAwMCBu IAowMDAwMDMwMjE4IDAwMDAwIG4gCjAwMDAwMzk5ODEgMDAwMDAgbiAKMDAwMDAzOTA3NSAw MDAwMCBuIAowMDAwMDMwMjM5IDAwMDAwIG4gCjAwMDAwMzQwNjIgMDAwMDAgbiAKMDAwMDA0 MDAzMyAwMDAwMCBuIAowMDAwMDM5MjE5IDAwMDAwIG4gCjAwMDAwMzQwODMgMDAwMDAgbiAK MDAwMDAzNzg4NiAwMDAwMCBuIAowMDAwMDQwMDg1IDAwMDAwIG4gCjAwMDAwNDM1MjcgMDAw MDAgbiAKMDAwMDA0MzYzOCAwMDAwMCBuIAowMDAwMDQzODUzIDAwMDAwIG4gCjAwMDAwNDQy MjEgMDAwMDAgbiAKdHJhaWxlcgo8PCAvU2l6ZSA1OSAvUm9vdCAxIDAgUiAvSW5mbyAyIDAg Ugo+PgpzdGFydHhyZWYKNDQ1NjAKJSVFT0YK --AqsLC8rIMeq19msA--