pdf file checks in as ANSI instead of Binary

The problem I ran into is that the sourcesafe 2005 client would not allow me to check a .pdf file in as binary. I tried over and over and could not force it to binary. I then tried running the old 6.0d client and it did allow me to checkin to the same database as binary. What happened in the 2005 client I even tried specifying that *.pdf files were in the "binary" group and it still checks in as ANSI instead of Binary.

If I go back to the 6.0d client, and select Binary while checking in the pdf file, it works like a champ.

Answer this question

pdf file checks in as ANSI instead of Binary

jayk
David,

We are having the EXACT same problem. No matter what we do -- just as you've suggested -- changing filetype to binary reverts back to "auto-detect" and adding the .pdf extension to the list of binary files doesn't help.

The only fix we've been able to come up with is to revert back to checking .pdf files in with the old VSS6 client (just like you).

I consider this a complete showstopper for us as we use lots of .pdf files and when they are mangled, it's usually only by a few bytes, but enough that customers can't open them (we get errors like "missing token," something about "image data" -- basically all problems related to corrupted .pdf files).

I'm happy to vote for your bug if you've submitted one. To me, it's a biggie.
Fraas
Hi David,

I tried with Beta2 and latest builds of VSS2005 and I cannot reproduce this behavior. I'd suggest opening a product bug and attaching the pdf file you're using. You should provide details of what build of VSS you're using (from Help/About box). Also, attaching your srcsafe.ini and ss.ini files may help, too.

You can correct the file type of the already-added pdf files by selecting the files in SourceSafe Explorer, displaying the File/Properties dialog and changing the Type value in the drop-down combobox (you need to have the file checked in to have the drop-down enabled).

Thanks,
Alin
lym51
David,

I just found a reference to this in the bugbase and apparently it was fixed in builds AFTER beta 2. Here's the link:

http://lab.msdn.microsoft.com/productfeedback/viewfeedback.aspx feedbackid=72e070e6-ffdf-40f3-8a22-2bef4d033185

So I guess now we just have to get updated, don't we!

Glad they fixed it, life can now continue :)

--Buck
pdf file checks in as ANSI instead of Binary

Answer this question