• Global community
    • Language:
      • Deutsch
      • English
      • Español
      • Français
      • Português
  • 日本語コミュニティ
    Dedicated community for Japanese speakers
  • 한국 커뮤니티
    Dedicated community for Korean speakers
Exit
0

RH8 HTML compiling to Webhelp - how to shorten cycle?

Guest
Oct 15, 2010 Oct 15, 2010

Copy link to clipboard

Copied

Good morning!

We have 20 projects being authored in RH8.  All of them have a healthy supply of baggage files - mostly pdf files, but also some .doc and .xls files.  All projects live on the C:\ drive - no server authoring.

We are seeing a huge range of compiling times...from 5 minutes to nearly 30 minutes per project.  I understand full-text search is now looking at the baggage files...and that's fine...but WHAT is it looking at?  Is there anything I can do to the files to make them index quicker?  Something is clearly different and its very frustrating.  We have put in Titles within the properties - no difference.  The ones that compile quickly do so each time...the ones that take 'forever' always take forever.  And the time is always within the Full-Text search part of the compiling.  Those dotted lines either appear quickly...or very, very slowly.

And our baggage file names are 'clean' - no illegal characters, and underscores instead of spaces.  We treat them like topic file names.

Is there anything that can be done to help the search process do its job quicker?

Thank you!

Views

5.8K

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Community Expert ,
Oct 15, 2010 Oct 15, 2010

Copy link to clipboard

Copied

I nearly wet myself laughing when I saw who posted this! Your projects are large to say the least and what you have accumulated over the years is not going to help.

Let's have some information about the project that takes 30 minutes. No of topics, total size, no of images, you know the drill.


See www.grainge.org for RoboHelp and Authoring tips

@petergrainge

Help others by clicking Correct Answer if the question is answered. Found the answer elsewhere? Share it here. "Upvote" is for useful posts.

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Guest
Oct 15, 2010 Oct 15, 2010

Copy link to clipboard

Copied

Peter...that hurts!  Just because I have 50+ people authoring ....many who know little about RH...and dumping into these projects...come on...are you saying there might be "junk" in here...??!!!  probably!

I'm gathering info now.  Will post shortly.  Recompiling one monster so I have an accurate time count for you.  Give me an hour please...!

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Community Expert ,
Oct 15, 2010 Oct 15, 2010

Copy link to clipboard

Copied

I'm not blaming those people or you, just stating what you know to be true. I seem to recall some dodgy tables a few weeks back!

You say that some same size projects compile in vastly different times. Something has to be different in them.


See www.grainge.org for RoboHelp and Authoring tips

@petergrainge

Help others by clicking Correct Answer if the question is answered. Found the answer elsewhere? Share it here. "Upvote" is for useful posts.

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Guest
Oct 15, 2010 Oct 15, 2010

Copy link to clipboard

Copied

Understand!  Actually getting topic counts / compiling times for top 10 projects right now...will post back here when I have it all.  It takes a bit since I have to compile each one to get its 'real' time.

Thanks!!!

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Guest
Oct 20, 2010 Oct 20, 2010

Copy link to clipboard

Copied

I've managed to get file counts, compiling times, calculate which take longer than others...for same relative size and my numbers are as follows:

compiling.jpg

The numbers in YELLOW are our 'slow projects'...and the ones in RED are our 'fast projects'.  Everything else is just there.

Questions:  is nearly .7 seconds per file reasonable for machines with 3GB RAM, plenty of space and 2.53 Ghz processors?  As you can see...we're getting compiling times of 16-38 minutes for our larger projects and these are not really huge projects.  They will get bigger...!  And,  of course with RH7 we're talking 2-3 minutes compiling time per project so this is a huge change for us.  All of these are compiled and published to Production weekly and many twice-weekly.

Why are the 'red' projects so fast?  DMS is .35 seconds per file yet has a large number of baggage files in addition to the htm files.

Any thoughts on this would be appreciated.

Have a great day!

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Community Expert ,
Oct 20, 2010 Oct 20, 2010

Copy link to clipboard

Copied

Is Republish All blank for all projects, or ticked for all?


See www.grainge.org for RoboHelp and Authoring tips

@petergrainge

Help others by clicking Correct Answer if the question is answered. Found the answer elsewhere? Share it here. "Upvote" is for useful posts.

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Guest
Oct 20, 2010 Oct 20, 2010

Copy link to clipboard

Copied

its 'blank for all'

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Community Expert ,
Oct 20, 2010 Oct 20, 2010

Copy link to clipboard

Copied

You are using the term Compiling but strictly speaking WebHelp is Generated and then Published. Are the times you quote just for generating or both?


See www.grainge.org for RoboHelp and Authoring tips

@petergrainge

Help others by clicking Correct Answer if the question is answered. Found the answer elsewhere? Share it here. "Upvote" is for useful posts.

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Guest
Oct 20, 2010 Oct 20, 2010

Copy link to clipboard

Copied

Peter,  I'm quoting the total time since that's really what matters to us - the final number showing in the output window.  Example:

:

"Finished compiling WebHelp in 1316 sec(s)"

"Compilation complete."

For that number 896 seconds was Full-Text Search and 695 seconds was updating topics.  This is from PS for the current compiling....which is nearly TWICE as long as the number in the table above!  Our fun...the compiling time ranges widely and almost never for the good. 

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Community Expert ,
Oct 20, 2010 Oct 20, 2010

Copy link to clipboard

Copied

Of course it is total time that matters to you but where the problem is occurring is what matters to solving it. It could be that local generation times are comparable and the extra time is all in the publishing.


See www.grainge.org for RoboHelp and Authoring tips

@petergrainge

Help others by clicking Correct Answer if the question is answered. Found the answer elsewhere? Share it here. "Upvote" is for useful posts.

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Guest
Oct 21, 2010 Oct 21, 2010

Copy link to clipboard

Copied

Peter, good morning!

what information do you need?  You want a breakdown for each of the projects of the for the compiling 'pieces'?  Remember...we are not publishing to a site, simply compiling locally on the C drive.  Let me know and I'll track it down.  Thanks!

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Community Expert ,
Oct 21, 2010 Oct 21, 2010

Copy link to clipboard

Copied

OK so you are not generating and publishing, you are generating only.

Ordinarily I would say send me a good and a bad project but I know it will be at least a week and probably nearer two before I can devote the time to it.

Have you tried deleting the CPD and have you monitored the Output View while it is generating to confirm it is the search where the time is going?

If that doesn't reveal anything, try things like applying an Exclude build tag to topics with links to lots of baggage files and generate excluding those files.

Post back to say how that goes.


See www.grainge.org for RoboHelp and Authoring tips

@petergrainge

Help others by clicking Correct Answer if the question is answered. Found the answer elsewhere? Share it here. "Upvote" is for useful posts.

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Guest
Oct 21, 2010 Oct 21, 2010

Copy link to clipboard

Copied

Peter, clearly full-text search is taking a lot of time - we can see that. MORE than the compiling process.  I can provide the number breakdown between "updating HTML topics" and full-text search if that's helpful.

The topics update in a continual stream quickly.  Then when full-text search begins those dotted lines appear quickly in some projects (DMS) and very, very slowly in others (PS).  This is the point of my question - can we do something to the files (guessing its the baggage files?) to make them index into full-text search quicker?  DMS has over 100 PDF files yet full-text search breezes right through it.

We rebuild the .cpd file often so that's not the issue.

Finding topics that use lots of baggage files will be a challenge since RH will not tell you where they are used like images.  ??

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Advisor ,
Oct 21, 2010 Oct 21, 2010

Copy link to clipboard

Copied

In your original post, you said: "I understand full-text search is now looking at the baggage files" without, I think, truly understanding the magnitude of that.

Oh, let's see...2225 topics, 169 .pdf files, 36 .xls files, and 87 .doc files in the LSMT project? 1788 topics, 85 .pdf files, 35 .xls files, and 141 .doc files in the Annuities project? All being indexed for full-text search? I'm surprised that RH isn't barfing up hair balls and refusing to generate at all! If RH is doing anything similar to the Zoom Search engine some of us use, it's using plug-ins to make temp ASCII files of each of the files, indexing those, and establishing keys to the originals. Takes a little time, you know?

As a test, you could make a duplicate of your rhbag.apj file, strip all but a few baggage files from the list, and swap out the original. Generate the project and see what your times are. Another thing you could try is to completely delete the output folder you're generating to, then generate again; that sometimes helps.

Might be time to think about breaking up some of those monsters into merged WebHelp. Yet another idea might be to exclude those files from the RH Search, and run separate indexing on those with a product like Zoom, adding a toolbar link to its separate search page.

Good luck,

Leon

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Guest
Oct 21, 2010 Oct 21, 2010

Copy link to clipboard

Copied

Leon, thanks for replying.

Unfortunately this doesn't really help.  Our output folder is never present when we compile, its removed each week after the deliverable is done.

I'll look at the rhbag suggestion and maybe with trial and error I can see where we crater but that takes lots of time,  which I don't have.

I was hoping that there was a 'known issue' - like exclude all PowerPoint files or all Excel files and things will be fine.  No luck.

Our projects could be much larger...one block of these projects was one single project and is now merged webhelp.  But that has its issues too!

Our client audiences deal with a wide range of issues and want fairly comprehensive blocks of content together in each volume.  The joys of the Life/DI/Investment insurance industry!

Too bad there is not a visible 'auditing tool' where we could see what it was that caused it to start putting up those dots so slowly...and then deal with it.

Take care...

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Advisor ,
Oct 21, 2010 Oct 21, 2010

Copy link to clipboard

Copied

Amen, sister.

The only "known issue" is the sheer overload of these monster projects of yours. The beauty of the merged WebHelp, of course, is that you're never generating the whole thing, only the individual projects with changes. Even that, however, might be small consolation in your environment.

Good luck,

Leon

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Community Expert ,
Oct 21, 2010 Oct 21, 2010

Copy link to clipboard

Copied

Teri


If you are able to exclude some file types from the search, even just as a trial, see Item 6 at http://www.grainge.org/pages/authoring/rh8/using_rh8.htm which quotes Colum's instructions for removing specific file types from the search. I cannot tell you which file types are causing the problem, if indeed it is down to one file type, but trial and error should help.

Looking at the file type numbers though, there is no obvious type to try.

PS us the project that stands out. Hardly anything to it yet it takes so long. If you want to send it to me the usual way, I will at least generate it to get a comparison.


See www.grainge.org for RoboHelp and Authoring tips

@petergrainge

Help others by clicking Correct Answer if the question is answered. Found the answer elsewhere? Share it here. "Upvote" is for useful posts.

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Guest
Oct 21, 2010 Oct 21, 2010

Copy link to clipboard

Copied

Peter, as always...THANK YOU for extending a hand and offering to look!  I will package and send the normal way.

I will look at the article and play with that...might well be worth a try.  We are implementing a separate search engine - Ultraseek - but I hate to give up the Search tab since its how I find topics quickly out in Production when I need to.  Ultraseek finds more than I need it to!

Take care...look for a transfer message shortly...

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Community Expert ,
Oct 21, 2010 Oct 21, 2010

Copy link to clipboard

Copied

I am downloading the project but I had to laugh when you said you have stripped out the images to get the project below 100mb!

The number of files in PS in your spreadsheet is small but the size of them must be massive. I'll see how what I have got generates but if it is only part of the project, it is not really a valid test.

Using a different search engine may have some advantages if it has features RoboHelp does not offer you but it is not going to stop RoboHelp creating its search index.


See www.grainge.org for RoboHelp and Authoring tips

@petergrainge

Help others by clicking Correct Answer if the question is answered. Found the answer elsewhere? Share it here. "Upvote" is for useful posts.

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Community Expert ,
Oct 21, 2010 Oct 21, 2010

Copy link to clipboard

Copied

Under five minutes to generate! How much are you willing to pay for my laptop? Oh, I did have the project on an external USB portable hard drive (unpowered) and that, in theory, should have slowed it somewhat.

Sounds like your environment might have something to do with this.


See www.grainge.org for RoboHelp and Authoring tips

@petergrainge

Help others by clicking Correct Answer if the question is answered. Found the answer elsewhere? Share it here. "Upvote" is for useful posts.

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Community Expert ,
Oct 21, 2010 Oct 21, 2010

Copy link to clipboard

Copied

4 minutes 15 seconds on the internal hard drive.


See www.grainge.org for RoboHelp and Authoring tips

@petergrainge

Help others by clicking Correct Answer if the question is answered. Found the answer elsewhere? Share it here. "Upvote" is for useful posts.

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
New Here ,
Oct 29, 2010 Oct 29, 2010

Copy link to clipboard

Copied

Hi All,

Short of my sending to you, Peter, our mongo project, would you give me a rough estimate of the specs of the PC used for your build of this project?

We are looking at new PCs specs right now, and our person trying to build FM -> PDF and FM-> WebHelp (ugh on that for time!!)...is saying the minimum specs stated by RH is just not enough for these robust projects. And yes, we also have it single-sourcing on the network.

Anyway, a good idea of specs would be helpful right about now.

thanks!

Phyllis

Message was edited by: peppy3

Message was edited by: peppy3

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Community Expert ,
Oct 31, 2010 Oct 31, 2010

Copy link to clipboard

Copied

I am not sure what you mean by the first sentence.

You don't give any description other than robust. I've seen people say here they have a large project of a hundred topics. Large is when you get into thousands.

Pretty much any half decent modern machine should do, the key thing is RAM.


See www.grainge.org for RoboHelp and Authoring tips

@petergrainge

Help others by clicking Correct Answer if the question is answered. Found the answer elsewhere? Share it here. "Upvote" is for useful posts.

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
New Here ,
Nov 01, 2010 Nov 01, 2010

Copy link to clipboard

Copied

Hi Peter,

Yes, we are in the thousands of topics. One of our main projects starts as a PDF of 2300 pages (roughly) and also needs to be converted to a WebHelp project. If I am talking specifics, in terms of RAM, I was hoping for someone to state something like: 4GB, etc.

The best thing I can do to describe how our software applications work, is to say, "it's like running multimedia." Then, they start to catch the drift of the power we need. We're also mapping out the amount of time the RH and FrameMaker processes take to give them an idea how long we wait for the processes to run.

thanks for your help.

Phyllis

Votes

Translate

Translate

Report

Report
Community guidelines
Be kind and respectful, give credit to the original source of content, and search for duplicates before posting. Learn more
community guidelines
Resources
RoboHelp Documentation
Download Adobe RoboHelp