Sourcesafe and RH8

[Guest]

Hi everyone,

Currently we're looking into using multiple authors for a merged web help project.

So we have

Front end "master" project

Slave project 1

Slave project 2

etc.

We want to be able to check out each slave project to a different author then when they're all finished check them back in so lead author can compile the whole project. It seems our options are either source safe which our developers use or robosource.

Does anyone have experience of using source safe? If so what problems have you encountered if any please?

We are of course creating and editing locally on a real folder off the C drive. The source safe et al are for storage only...

[RoboColum_n_]

We used to use SourceSafe but recently migrated to SubVersion. We only used Sourcesafe because this is what our developers used and we migrated with them when they moved to SubVersion. We never had any problems with SourceSafe but we did not integrate it with our RH projects. We simply checked out the source in SourceSafe, made the changes in RH and checked them back in in SourceSafe. You can if you wish do the check in/check out directly from your RH project in the same way as RoboSourceControl.

---

The RoboColum(n)

@robocolumn

Colum McAndrew

[Guest]

Thanks folks,

Colums apporach is what I had in mind. Also what about Documentum? meant to ask if anyone had used it with RoboHelp..

[MergeThis]

Whichever product you use, do not add these RH project files to the Source Control database:

.cpd

.hhp

.ldb

.pss

Good luck,

Leon

[Guest]

All we want to do is to copy each project contained in a folder say "printing guide" which includes the whole robohelp project into source safe for storage and check the whole thing out when needed. Are we saying then that this is a problem?

[MergeThis]

The problem is that you'd be turning this into an overly complex undertaking.

The process should be this:

Source Files on User Machine	Source Files  in Source Control Database	Output Files Generated to User Machine and Published to Server
C:/xxx_merge folder	$xxx_branch	xxx_path
parent_project folder	parent_project folder	parent_project output files
projects (folder "shell" only)	child1_project folder	mergedProjects folder (folder "shell" only)
child1_project folder	child2_project folder	child1_project folder
child2_project folder	etc.	child2_project folder
etc.		etc.

Therefore, if User A needs to work on the child2_project, User A "Gets" only the child2_project, "Checks Out" the files to edit, Generates and Publishes output to the child2_project output folders, and "Checks In" the edited files. No need to check out the entire project, no need to generate the entire project, no need to publish the entire project.

Good luck,

Leon

[Guest]

Folks,

I have been  looking into this further and have posted some more questions in the source area under javascript:;

I'm afraid I'm being a bit thick about this but I want to make sure I understand thoroughly and properly.

Thanks

[MergeThis]

You're assuming an extra step that might not always be needed, which is the hidden beauty of merged WebHelp.

That is, the parent project does not need to be "run-time" generated when changes are made to the child project(s). Each of the child project output folders are actually self-contained entities, and the parent project "pulls up"  the current output at viewing time.

If you haven't already, you absolutely want to read http://www.grainge.org/pages/authoring/merging_webhelp/merged_help.htm (Peter Grainge's excellent merged Help tutorial).

(Oh, and I used SourceSafe with RH seven years ago (with RH X5), and the process was just as efficient as using RoboSource Control is today (with RH8).)

Good luck,

Leon