I do not zip files, but rather XML. As you know the performance and speed of CURSORTOXML and XMLTOCURSOR get very bad when you have a cursor with many records. So in our case for example we can make a cursor with a memo field and zip up (for example) 1000 records, put into a memo and do this for 1000 records. zip this up and send it someplace and on the other side unzip and get the resulting 1000X1000 records in very little time. ie performance is fantastic and we do this by using Craig Boyds compression FLL. Very powerful and useful.

The same concept that Allan Stevens presented at SWF a couple of years ago.





>Use Windows's own zip. I got it at this forum from someone.
>
>

>**** zip *****
>cFileZip = "C:\prg\xpzip\dene.ZIP"
>cSource = "c:\prg\xpzip"
>strtofile(chr(0x50)+chr(0x4B)+chr(0x05)+chr(0x06)+replicate(chr(0),18),cFileZip)
>o=CREATEOBJECT("shell.application")
>FOR EACH ofile IN o.NameSpace(cSource).items
>   o.NameSpace(cFileZip).copyhere(ofile)
>ENDFOR
>
>**** unzip ****
>LPARAMETERS cFileZip,cDestination
>
>o=CREATEOBJECT("shell.application")
>FOR EACH ofile IN o.NameSpace(cFileZip).items
>    o.NameSpace(cDestination).copyhere(ofile)
>ENDFOR
>

>
>
>>Is anyone using Craig Boyds "vfpcompression.fll'?
>>
>>I have one Windows 7 professional workstation that is failing on the "unzipstring" call.
>>
>>
>>other windows 7 prof. work fine, just this one workstation and I am lost as to what could be different.
>>
>>
>> lcCompressString2 = unzipstring(lcCompressString1)
>>
>>
>>The length of the lcCompressString1 is 2,957,129 and the resulting lcCompressString2 is 3,984,807 on those that are working.
>>
>>Any suggestions or advice to determine what is wrong is greatly appreciated.
>>
>>Thanks in advance