>>>I'm generating webpages with some images in them. The slow part is getting the image size, which I do by
>>>
>>>

>>>loImg=LoadPicture(lcFajl)
>>>lnWidth=Round(loImg.Width/26.458,0)
>>>lnHeight=Round(loImg.Height/26.458,0)
>>>

>>>
>>>The first line is the slow part. I seem to remember these attributes can be extracted, given the filename, by some feature of scripting.filesystemobject that I can't find now.
>>>
>>>I think I saw something like that in the last ten days or so, but me.memory.refresh() returns .f. - so, is there anyone with better memory?
>>
>>I think FileSystemObject loads the file anyway so you wouldn't see much of a speed improvement.
>>
>>The fastest way is to read in the header and extract the dimensions from there. What type of file are you looking at ?
>
>About 99% of it is jpg, so a jpg-only solution would work perfectly. The remainder can be converted to jpg :).

IIRC, JFIF is the hardest header format to read :-{
Link to a C version you might be able to convert : http://www.64lines.com/jpeg-width-height
No idea how reliable it may be tho.
Occurs to me that if the list of images changes only rarely you could use LoadPicture() when a file is added and store filename,width and height in a lookup table for future use?