Minor Documentation Corrections

[Ken Carpenter]

Hi Michi,

Here are a few more minor corrections to the documentation. By the way, I got a sense of deja vu reading this document! The organization and some of the diagrams seemed really familiar, so I pulled out my copy of Adv. CORBA Programming with C++. Ice provides not only code reuse, but also documentation reuse!

I really have to thank you for writing that book. After seeing the nasty CORBA C++ mapping I ran shrieking in the opposite direction! Ice looks much nicer.

Corrections:

- TOC page numbers need to be regenerated

- Label the page numbers of the PDF (Document->Number Pages...) so you can go directly to a logical page number with the page number edit box at the bottom of the Acrobat Window.

- Section 19.4.1, page 479:
"So that the service manager to load it properly" --> "So that the service manager CAN load it properly"

- Section 14.5 Mutexes, page 301 (under the "lock" bullet):
"the calling threads has acquired" --> "the calling THREAD has acquired"

- Section 14.8 Timed Locks, page 313, first sentence:
"functions the operate" --> "functions THAT operate"

- Section 14.9 Monitors, page 316, first paragraph:
"when condition becomes true" --> "when A condition becomes true"

- Section 14.9 Monitors, page 317, first paragraph after code block:
"Note that Monitor is template class" --> "Note that Monitor is A template class"

- Section 14.9 Monitors, page 320, second paragraph after code block:
"template class with automatically locks" --> "template class WHICH automatically locks"

- Appedix D.1 Proxies, page 696, 3rd-to-last paragraph:
The first sentence is left incomplete. Looks like it's missing an XREF.



Ken Carpenter

[michi]

Originally posted by Ken Carpenter

Here are a few more minor corrections to the documentation. By the way, I got a sense of deja vu reading this document! The organization and some of the diagrams seemed really familiar, so I pulled out my copy of Adv. CORBA Programming with C++. Ice provides not only code reuse, but also documentation reuse!

Hi Ken,

Yes, there is some reuse in the book, in style and organization. Basically, it worked for the first book, so I figured it would work for the second one

I really have to thank you for writing that book. After seeing the nasty CORBA C++ mapping I ran shrieking in the opposite direction! Ice looks much nicer.

Thanks, glad you like it! Yes, the CORBA C++ mapping really is an abomination. The complexity of that thing is insane, and totally unnecessary. Historically, what happened was that there were two competing submissions for the C++ mapping, the one we have now and one by Hyperdesk. The Hyperdesk mapping was much more like the Ice mapping in look and feel. Unfortunately, people decided to benchmark the two mappings and found that the Hyperdesk one was five times slower than the raw C++ mapping we have. So, they decided to adopt the abomination we have now.

The joke is that the speed difference is artificial: the raw mapping does very little, so the application ends up having to do all the things that should be done by the mapping. It turns out that these are the expensive things (such as hitting the memory allocator all the time). In other words, the actual speed difference is almost non-existent when you benchmark the mappings correctly, instead of comparing apples and oranges...

Corrections:

- TOC page numbers need to be regenerated

- Label the page numbers of the PDF (Document->Number Pages...) so you can go directly to a logical page number with the page number edit box at the bottom of the Acrobat Window.

We had some problems with the TOC -- I don't know why the page numbers have this nasty habit of going out of sync. We'll fix it for the next version.

Thanks for the suggestion about the PDF page numbers -- I'll do that for the next version as well.

Thanks for all the other corrections -- they'll be fixed in the next version.

Cheers,

Michi.

[Ken Carpenter]

Originally posted by michi
Thanks for the suggestion about the PDF page numbers -- I'll do that for the next version as well.

You'll want to number the first 15 pages with roman numerals (as you have already in the actual page contents), otherwise when you type in 5, for example, it will still go to the 5th physical page.

Ken Carpenter

[michi]

Originally posted by Ken Carpenter
You'll want to number the first 15 pages with roman numerals (as you have already in the actual page contents), otherwise when you type in 5, for example, it will still go to the 5th physical page.

Yep, I worked that out, no sweat

Cheers,

Michi.