Hmm, interesting !

I have to study this further and read the specs.

On 17 Aug 2012, at 22:29, Paul DeBruicker <[email protected]> wrote:

> Hi Sven,
> 
> Updating certainly lets me download it but I'm then faced with the problem 
> that the resulting file is a tar archive and is no longer gzipped.
> 
> 
> E.g.
> 
> wget http://www.iana.org/time-zones/repository/tzdata-latest.tar.gz
> 
> 
> paul@paul-laptop:~/pharo/Pharo-1.4$ls -la tzdata-latest.tar.gz
> -rw-rw-r-- 1 paul paul 206251 Aug  2 21:08 tzdata-latest.tar.gz
> 
> paul@paul-laptop:~/pharo/Pharo-1.4$ file tzdata-latest.tar.gz
> tzdata-latest.tar.gz: gzip compressed data, from Unix, max compression
> 
> 
> 
> 
> and after running the download in pharo, as you saw, the file is 686kB.
> 
> paul@paul-laptop:~/pharo/Pharo-1.4$ls -la tzdata-latest.tar.gz
> -rw-rw-r-- 1 paul paul 686080 Aug 17 13:23 tzdata-latest.tar.gz
> 
> paul@paul-laptop:~/pharo/Pharo-1.4$ file tzdata-latest.tar.gz
> tzdata-latest.tar.gz: POSIX tar archive (GNU)
> 
> 
> Is there a way to not unzip the file when downloading it through Pharo?  I 
> admit its a minor quibble, but its not expected.
> 
> 
> Thanks
> 
> Paul
> 
> 
> On 08/17/2012 11:47 AM, Sven Van Caekenberghe wrote:
>> Paul,
>> 
>> On 17 Aug 2012, at 19:55, Paul DeBruicker <[email protected]> wrote:
>> 
>>> Thanks for looking into it Sven
>>> 
>>> Pharo 1.4 (Summer) 14457.  Eliots 2585 VM.  Ubuntu 12.04 64bit.
>> 
>> You seem to have hit a problem in Zn that was recently fixed (beginning of 
>> august), if you update Zn it works.
>> 
>> Regards,
>> 
>> Sven
>> 
>>> On 08/17/2012 10:49 AM, Sven Van Caekenberghe wrote:
>>>> Works for me, Pharo 2.0 ##20201
>>>> 
>>>> [sven@voyager:~/smalltalk]$ tar tvfz tzdata-latest.tar.gz
>>>> -rw-r--r--  0 0      0       44941 Jul 19 02:45 africa
>>>> -rw-r--r--  0 0      0       15827 Jul 19 02:30 antarctica
>>>> -rw-r--r--  0 0      0      112760 Jul 25 16:13 asia
>>>> -rw-r--r--  0 0      0       68423 Jul 25 16:40 australasia
>>>> -rw-r--r--  0 0      0      121390 Jul 25 16:13 europe
>>>> -rw-r--r--  0 0      0      135756 Jul 25 16:13 northamerica
>>>> -rw-r--r--  0 0      0       73854 Jul 25 16:13 southamerica
>>>> -rw-r--r--  0 0      0        1190 Jul 19 02:30 pacificnew
>>>> -rw-r--r--  0 0      0        2955 Jul 19 02:30 etcetera
>>>> -rw-r--r--  0 0      0        4083 Jul 19 02:30 backward
>>>> -rw-r--r--  0 0      0        1546 Jul 19 02:30 systemv
>>>> -rw-r--r--  0 0      0         393 Jul 19 02:30 factory
>>>> -rw-r--r--  0 0      0       19306 Jul 19 02:30 solar87
>>>> -rw-r--r--  0 0      0       19324 Jul 19 02:30 solar88
>>>> -rw-r--r--  0 0      0       19600 Jul 19 02:30 solar89
>>>> -rw-r--r--  0 0      0        4326 Jul 19 02:30 iso3166.tab
>>>> -rw-r--r--  0 0      0       19913 Jul 19 02:30 zone.tab
>>>> -rw-r--r--  0 0      0        3186 Jul 25 16:13 leapseconds
>>>> -rw-r--r--  0 0      0         680 Jul 19 02:30 yearistype.sh
>>>> [sven@voyager:~/smalltalk]$ file !$
>>>> file tzdata-latest.tar.gz
>>>> tzdata-latest.tar.gz: POSIX tar archive (GNU)
>>>> [sven@voyager:~/smalltalk]$ ls -la !$
>>>> ls -la tzdata-latest.tar.gz
>>>> -rw-r--r--@ 1 sven  sven  686080 Aug 17 19:47 tzdata-latest.tar.gz
>>>> 
>>>> What image, platform, VM are you on ?
>>>> 
>>>> On 17 Aug 2012, at 19:19, Paul DeBruicker <[email protected]> wrote:
>>>> 
>>>>> I'm trying to download the Olson time-zone database so Chronos can 
>>>>> automatically update its ruleset.
>>>>> 
>>>>> In a workspace, when I do:
>>>>> 
>>>>> ZnClient new
>>>>>   url:'http://www.iana.org/time-zones/repository/tzdata-latest.tar.gz';
>>>>>   downloadTo:FileDirectory default pathName.
>>>>> 
>>>>> 
>>>>> I get a SubscriptOutOfBounds error .  The file is ~207kB.  The 
>>>>> GzipReadStream>#getFirstBuffer gets the first 65,536 bytes and then hits 
>>>>> the out of bounds error.  Moving the limit in that method to 262,144 (1 
>>>>> << 18 rather than 1 << 16) gets us to another error from which I don't 
>>>>> know the right way to proceed.
>>>>> 
>>>>> 
>>>>> I think that because its an content type 'application/x-tar' it shouldn't 
>>>>> hit the GzipReadStream class at all but should just be a ReadStream that 
>>>>> does not get decompressed in the image and instead sent straight to the 
>>>>> disk.  I'm not sure how to implement that though inside Zinc and would 
>>>>> welcome ideas to try.
>>>>> 
>>>>> Thanks
>>>>> 
>>>>> Paul
>>>>> 
>>>> 
>>>> 
>>> 
>>> 
>> 
>> 
> 
> 


Reply via email to