Journal Editorial + CVu Journal Vol 16, #5 - Oct 2004
Browse in : All > Journal Columns > Editorial
All > Journals > CVu > 165
Any of these categories - All of these categories

Note: when you create a new publication type, the articles module will automatically use the templates user-display-[publicationtype].xt and user-summary-[publicationtype].xt. If those templates do not exist when you try to preview or display a new article, you'll get this warning :-) Please place your own templates in themes/yourtheme/modules/articles . The templates will get the extension .xt there.

Title: Editorial

Author: Administrator

Date: 09 October 2004 13:16:08 +01:00 or Sat, 09 October 2004 13:16:08 +01:00

Summary: 

Body: 

This is my third edition editing C Vu. It's been a cracking experience which has been made all the more pleasant by the help I've had along the way. It's always good when material comes in well in time and more rewarding when readers write in to the letters page with comments on the material. This edition is a case in point - there are more letters in this edition than I've seen in quite a long time.

Is this down to the material being more thought provoking, the range of articles being more varied or simply people feel like emailing? It's hard to say.

Not all emails get published though. For example, I've had some through expressing concern on the focus I've brought to open source coding (the GTK articles, wxWidgets [starting this edition] and Qt have raised a few eyebrows). This is in part quite deliberate, but also down to one other reason : nobody has submitted material on the likes of MFC. We can only publish what is submitted!

Why Deliberate?

C,C++, C#, Python and Java (to name a few) are platform independent languages. It makes no difference which platform code is developed on, as long as they use the pure language and the compiler used at either end is standards compliant (or close to it!), the outcome will be the same. C on my RiscPC will compile on my Linux box and on my friend's OS X machine. The outcome will be the same. That is part of the beauty of these languages. Write once, compile many, outcome same.

The same applies for widget libraries. I enjoy writing code which compiles on my Linux box, take it into work and compile the results on MSVS.NET and see the same results. The independence of the widget library is great in that way. To me, this is an extension of the same ideas as are behind the platform neutrality of the languages we all love and know.

By using these cross platform libraries, it is my firm opinion that there can be a massive increase in productivity as well as stopping some of these "we have this, you don't" arguments you see from time to time. Imagine something like 3D Studio Max or Sibelius 2 being written using (say) wxWidgets - the companies behind them could very quickly and easily produce versions for many platforms (for wxWidgets, it ranges from 16 to 64 bit platforms). Upshot would be a great deal more money for the companies.

What does need to be asked though is why is this not done? Many pieces of software are available for both OS X and Win32. They're not using the same widget sets and required different teams of programmers. To me, this seems very wasteful. Fair enough for something which is Win32 only and requires MFC and other proprietary libraries, but for the rest where the two platform system is used, it would make more sense to use an independent widget set. One code base, compile many, lots of money!

Despite what some of the worlds largest companies say, the use of cross platform libraries is gaining in popularity and moreover, gaining in speed.

One of the largest problems though with some of the cross platform libraries are the licences. Qt is free for X11 and OS X, yet the Windows version requires licences. Many managers don't understand the implications and ramifications of using GPL libraries. I don't know the solution for this, though a simplification of the licences would certainly help.

Books Books Everywhere and Not a Drop to Read

One of the problems you have when you review lots of books is what to do with them after you've reviewed them. Now, this is not a problem for the better books. They are on my bookshelves, waiting for their next set of being used.

The problem comes with the books which are dangerous. You know the sort - they have "not recommended" in the book reviews and after you read the review, you can imagine the reviewer dancing around a small bonfire made from the dead trees wasted on such a pile of rubbish. No responsible person would try to sell them on eBay or put them into a recycling box in case some poor person at the recycling plant gets hold of the book and decides to read it.

You could leave them on the shelf, but then they're taking up valuable good book space. I suppose putting them in the loft would be an idea - however even that has its drawbacks (mice).

What would be quite fun would be to get all of the authors of these utter turkeys in a field and have people pelt them with pages from the books while chanting "you shall not write such utter tosh in future". For good measure, some of the technical editors who have supposed to have read these books should also be put in the same field. Okay, it wouldn't be very productive, but it's one way of getting rid of the books! That said, I have a feeling that one or two authors in particular would pay little or no attention to such activity... (no names, no pack drill).

It's actually a pity that book companies don't do the same as record companies. I have two books by Ammeral; C++ for Programmers and STL for C++ Programmers. Both very worthwhile books and both of which (in their time) have been frequently referenced. There is a lot of crossover between the two books (sorry Francis, I know you'd disagree with me here - but there is).

What would be great is if there was a sort of "Best of" for these books. One volume without the crossover material, but all of the great information.

This idea could be applied to a number of other books - some of the XP ones have a good chunk of similar (not the same) material. A bit of rewording and instead of 4 books, it becomes 2. Less space occupied and more information for the page count.

All right, some books you would never dream of doing that to. Josuttis's C++ Standard Library being one of them. That book is just so crammed pack full, it would be pointless to try and merge it with (say) C++ Templates (which was co-written with Vandevoorde).

Dead Websites (or A Tale of Two Websites)

One of the most annoying aspects of any book is when they reference a website in the text or on the book itself. Now, I'm not that daft to imagine for one moment that a book company or a person's ISP is going to exist forever. However, book companies take over other book companies, so at least some material should still exist.

Case 1

Company 'A' have produced a lot of books. I have 7 of their books on my shelf currently. Their books make references to a number of websites, all of which are required to some extent to service the code in the book and in one particular book, an entire chapter is pretty pointless without one of the libraries listed.

The original company was bought out and the new company doesn't recognise the book as being one of theirs, leaving the person with a book which is practically useless for a couple of the most important chapters. There is a CD ROM with the book, but in a break with tradition, it is filled almost entirely with material that someone at the book's original company thought would be a good idea at the time.

The original author's website has vanished. Waybackmachine can't find the download and even Google draws a blank.

Case 2

Company 'B' publishes a book which is a couple of years old and hasn't been updated. The libraries referenced still all exist but have been greatly modified since the original release of the book. A person undertakes the job to update the codebase, tells both the author and the book company of the update and where it can be found. Company B takes a copy and posts it on the support area for the book and drop a quick email back to say thanks. The download is amazingly popular for both the company and the person who has done the update.

Company B publishes their books through another company.

I suppose there is only one thing worse than a company like company 'A' and that is one which has updates but the updates are broken and refuses to even email back to say "thanks, but we're not going to fix it for reason a, b and/or c".

And So It Begins

By the time this edition hits the doormats, the new academic year in the UK Universities will be well underway. A fresh intake, all ready and eager to learn. Plenty of parents worry about their offspring being away from home for the first time and hoping they'll be fine.

While my child is only 6 (and so is not ready for University yet!), I can say that they will be. First year student life is a gas. Stop worrying - the worst they can do is have some really weird tattoo done and miss a couple of lectures.

With all of this spare time you now have as you've stopped worrying (a bit), what should you do? Watch TV? Listen to that collection of I'm Sorry I'll Read That Again you have on CD? Have a meal out?

Why not write for C Vu or Overload? We're always after new articles, book reviewers and contributors to the Student Code Competition. Let's make both magazines even bigger and better value for everyone!

Notes: 

More fields may be available via dynamicdata ..