Bug - Won't Fix Cached images folder caches empty images

[mredge73]

I use the "cache kol images feature" for the relay browser. The problem with the feature is that it will cache empty images if you don't have access to kol's image server. After that you are stuck forever with these empty images unless you manually delete them all and re-cache them when you have access to the image server.

I am asking if you guys can amend the function so that empty images are not cached. This alone will ease my pain but, as a bonus, could there be a way to auto-delete the ones that already exist. On my machine the empty images all have the same size, 800bytes.

My corporate nanny software (websence) blocks access to kol's image server so I rely on these cached images during casual relay adventuring during long hours when we are really slow.

 

[Veracity]

Revision 10451 will not write a local copy of an image file if the connection to the URL fails. If the conenction succeeds and 0 bytes are retrieved, it will delete the resulting image file.

I have no idea where you are getting 800 byte files from a failed connection. When I tried to fetch a non-existant file, I got a connection failure and previously we ended up with a 0-byte file. Perhaps your nanny software is allowing the connection but returning 800 bytes worth of ... something. Can you tell me what is in one of these 800-byte files?

 

[Veracity]

Ha. I looked at websense.com to see what happens when it blocks something. It returns an HTML & Javascript page telling you what it blocked and why. That page can include company-specific content.

Basically, if we ask for an image and Websense blocks it, we get who-knows-what back. There's no general way that we can detect that we didn't get back what KoL actually sent. The problem isn't that we cache "empty" images. The problem is that we cache bogus non-images returned by your company's nanny software.

I can't fix this for you. Sorry.

 

[Theraze]

Only solution would be for mredge to turn on caching at home and then copy the resultant cached files to his work machine, but... again, that's on him, not the developers.