whdata, whgdata, and whxdata output folders

[Gravenstein]

I'm wondering if I can safely remove the whdata and whgdata output folders from my projects. The help runs on our clients' servers, which are IE 5 or better (we don't have to worry about supporting non-IE browsers). We're using the DHTML version of WebHelp--at least I assume that we fall in this category, because we use some DHTML features, and we specify "Java Applet: DHTML>Java Applet>Pure HTML" in our WebHelp SSL. So if I understand correctly, we should need only the whxdata folder to support our help.

Is that correct, or am I missing something? The whgdata folder in particular gets to be quite large. Would there be a danger in my omitting it?

G

[Peter Grainge]

Would you expect your car to run if you took the engine out? Webhelp is driven by all the wh* stuff so you might find it doesn't run too well without them.

Let me refine that. You might be able to remove some but you would have to do some serious testing, is it worth the effort just to save a bit of disk space?

[MergeThis]

Those folders only contain all the information required to populate the navigation pane (Contents, Index, and Search).

If you feel your help will be useful to the user without that information, then cut away!


Good luck,
Leon

[HKabaker · Accepted solution]

Just to amplify a bit,

Some of these files are "duplicates" -- that is, they have the same text content -- but code is customized for browsers other than IE and for IE browsers earlier than 5.5. There may be some differences between IE 5.5 and IE 6; I'm not sure.

And you shouldn't assume that the latest IE browsers refer only to the whxdata folder. As Peter says, you would need thorough testing.

Finally, as a reminder and for forum readers who may not have seen this advice:

Each merged project has its own set of these three folders, which the master project uses to assemble TOC, Index and Search databases on the fly. So even though they have identical names, they are not duplicate or redundant. Let them be.

Harvey

[Gravenstein]

Thanks, all. I had hoped to eliminate files associated with browsers that we didn't support anyway, but I agree that it's better to take the more cautious approach. Even if I could devote lots of time to testing, there's no guarantee that I'd be able to simulate all the possible variations in user environments.

Cheers ,
G