On Thu, Oct 19, 2006 at 09:55:03AM +1300, Hamish wrote:
Frank wrote:
> OK, I somewhat botched the last release numbering, so I've lept ahead
> to beta4 assuming the last was really beta3. This beta includes
> the Roussilhe Stereographic projection (copied from libproj4), as well
> as an update to EPSG 6.11.1 (at Markus Neteler's request).
do you want to sync GRASS to epsg 6.11.1 before releasing 6.2?
(no idea what the .1 adds, maybe current 6.11 is fine...)
Yes. I have done it already some hours ago.
Some important fixes have been done which were discovered
and needed for preparing the OSGeo North Carolina Edu dataset.
And the EPSG people really quickly responded and issued 6.11.1.
Hamish
ps - I would like one last (quick) RC version before release, as there
have been quite a few changes since rc2.
Agreed. I try to get it out tomorrow.
Honestly I am lost with the recent gis.m changes. Not sure which
patches need to be backported. The rest is sync'ed as appropriate.
On Thu, Oct 19, 2006 at 09:55:03AM +1300, Hamish wrote:
Frank wrote:
OK, I somewhat botched the last release numbering, so I've lept ahead
to beta4 assuming the last was really beta3. This beta includes
the Roussilhe Stereographic projection (copied from libproj4), as well
as an update to EPSG 6.11.1 (at Markus Neteler's request).
do you want to sync GRASS to epsg 6.11.1 before releasing 6.2?
(no idea what the .1 adds, maybe current 6.11 is fine...)
The latest EPSG version is buggy (the one Frank used in PROJ 4.5 beta).
At least 4 CS definition have been removed in error. I have informed
Roger Lott of EPSG about it and he confirmed this will be fixed. He
also said that the fix is going to be committed soon, maybe even this
week. First they need to find the reason of the erroreous deletion of
valid definitions, though. Frank Warmerdam is also informed about the
issue and kept posted, so he should be able to sync his software once
the fix is commited. GRASS should follow.
Yes. I have done it already some hours ago.
Some important fixes have been done which were discovered
and needed for preparing the OSGeo North Carolina Edu dataset.
And the EPSG people really quickly responded and issued 6.11.1.
On Thu, Oct 19, 2006 at 10:16:33AM +0200, Maciej Sieczka wrote:
Markus Neteler wrote:
> (cc list)
>
> On Thu, Oct 19, 2006 at 09:55:03AM +1300, Hamish wrote:
>> Frank wrote:
>>> OK, I somewhat botched the last release numbering, so I've lept ahead
>>> to beta4 assuming the last was really beta3. This beta includes
>>> the Roussilhe Stereographic projection (copied from libproj4), as well
>>> as an update to EPSG 6.11.1 (at Markus Neteler's request).
>>
>> do you want to sync GRASS to epsg 6.11.1 before releasing 6.2?
>> (no idea what the .1 adds, maybe current 6.11 is fine...)
The latest EPSG version is buggy (the one Frank used in PROJ 4.5 beta).
At least 4 CS definition have been removed in error. I have informed
Roger Lott of EPSG about it and he confirmed this will be fixed. He
also said that the fix is going to be committed soon, maybe even this
week. First they need to find the reason of the erroreous deletion of
valid definitions, though. Frank Warmerdam is also informed about the
issue and kept posted, so he should be able to sync his software once
the fix is commited. GRASS should follow.
Frank has re-added them in PROJ4-CVS.
Note that the nad/epsg file does not affect GRASS, we are using the
.csv files from gdal/data/.
Indeed, the nad/epsg file is only displayed in the startup GUI, unfortunately
the codes aren't even selectable from there.
> Yes. I have done it already some hours ago.
> Some important fixes have been done which were discovered
> and needed for preparing the OSGeo North Carolina Edu dataset.
> And the EPSG people really quickly responded and issued 6.11.1.
Maciek
So I don't see anything which needs to be done there unless
EPSG reaches a version > 6.11.1 .
On Thu, Oct 19, 2006 at 10:16:33AM +0200, Maciej Sieczka wrote:
Frank has re-added them in PROJ4-CVS.
Note that the nad/epsg file does not affect GRASS, we are using the
.csv files from gdal/data/.
And the .csv files from gdal/data/ are created based on EPSG database,
is that correct?
Indeed, the nad/epsg file is only displayed in the startup GUI, unfortunately
the codes aren't even selectable from there.
So I don't see anything which needs to be done there unless
EPSG reaches a version > 6.11.1
I'm only wondering if other codes, besides the 2172-2175 I know of,
were removed too. Because if that would be the case, they would be
missing in GDAL too, thus in GRASS as well. Frank would need to
manually add them back like he did with 2172-2175, or wait for the
fixed EPSG release.
Let's at least wait until Roger Lott finds out whether any more
corruption, besides that we know of, took place in EPSG 6.11.
And the .csv files from gdal/data/ are created based on EPSG database,
is that correct?
Maciej,
That is correct.
Indeed, the nad/epsg file is only displayed in the startup GUI, unfortunately
the codes aren't even selectable from there.
So I don't see anything which needs to be done there unless
EPSG reaches a version > 6.11.1
I'm only wondering if other codes, besides the 2172-2175 I know of,
were removed too. Because if that would be the case, they would be
missing in GDAL too, thus in GRASS as well. Frank would need to
manually add them back like he did with 2172-2175, or wait for the
fixed EPSG release.
The 2172-2175 codes that were manually added into nad/epsg will not
appear in the current GDAL. It should be sufficient to add the
corresponding records in gdal/data/pcs.csv from an older version of
GDAL that includes them.
Best regards,
--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up | Frank Warmerdam, warmerdam@pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush | President OSGeo, http://osgeo.org
On Thu, Oct 19, 2006 at 03:24:18PM +0200, Markus Neteler wrote:
...
So I don't see anything which needs to be done there unless
EPSG reaches a version > 6.11.1 .
I just learnt that that statement is wrong
So: could you check in the missing lines into gdal/data/pcs.csv
from an older GDAL version?
I'll wait with RC3 to get that fixed.
Markus
PS: Frank indicated the process of (re-)generating these EPSG files
as this (from today's IRC chat in gdal):
1) edit postgres .sql files to fix up a bit (hack in LATIN1 encoding statement).
2) load into postgresql.
3) run a script to dump out .csv files.
4) mv into right place (in libgeotiff/csv)
5) run build_pcs.py script.
6) mv updated stuff into gdal/data/ (and make sure it gets installed in
all /usr/local places on working system)
7) use epsg_tr.py to generate postgis .sql file, and proj nad/epsg
file. (multistep for each)
I'm unclear why the subject of this email train is 6.2 epsg version - I presume it is a typo for 6.11.2.
Version 6.11.2 of the EPSG dataset has been sent to our web master for public release. He is usually pretty quick at doing what is necessary and it would not surprise me if the release is available some time on Monday.
The only change to v6.11.1 is the reinstatement of the 4 CRSs 2172-2175 for Poland. We have run two independent checks which confirmed that these seem to be the only records that went missing, although we have been unable to trace the cause - I would guess human error!
v6.11.1 corrects errors in v6.11 for Carolina and Hong Kong CRSs. It also adds additional new data, as summarised in the release note. v6.11 has been withdrawn and should not be used.
Regards,
Roger Lott
----- Original Message ----- From: "Maciej Sieczka" <tutey@o2.pl>
To: "Maciej Sieczka" <tutey@o2.pl>; "GRASS developers list" <grass-dev@grass.itc.it>; "Frank Warmerdam" <warmerdam@pobox.com>; "Roger Lott (EPSG)" <epsg.rl@btinternet.com>
Sent: Thursday, October 19, 2006 2:43 PM
Subject: Re: [GRASS-dev] Re: 6.2 epsg version
Markus Neteler wrote:
On Thu, Oct 19, 2006 at 10:16:33AM +0200, Maciej Sieczka wrote:
Frank has re-added them in PROJ4-CVS.
Note that the nad/epsg file does not affect GRASS, we are using the
.csv files from gdal/data/.
And the .csv files from gdal/data/ are created based on EPSG database,
is that correct?
Indeed, the nad/epsg file is only displayed in the startup GUI, unfortunately
the codes aren't even selectable from there.
So I don't see anything which needs to be done there unless
EPSG reaches a version > 6.11.1
I'm only wondering if other codes, besides the 2172-2175 I know of,
were removed too. Because if that would be the case, they would be
missing in GDAL too, thus in GRASS as well. Frank would need to
manually add them back like he did with 2172-2175, or wait for the
fixed EPSG release.
Let's at least wait until Roger Lott finds out whether any more
corruption, besides that we know of, took place in EPSG 6.11.
