THtmlViewer component

[Alistair Ward]

When building Templot3 (using Lazarus 2.0.10), I'm getting duplicate unit warnings for all the HtmlViewer files, as they have been included in the templot3 source.

I've had a look at the changes, and there are 2 differences:

1. Hooks on clicking on links.
2. Font/measurement scaling.

I've modified templot to use the OnHotSpotClicked event, which means the internal hooks in htmlview.pas can be removed (committed in GitHub).

I'm still looking at the scaling, and have an idea for addressing it, but it will still require changes to the component. I'm hoping the idea I have will be sufficiently general that it can be submitted back to the THtmlViewer project for inclusion in the standard release.

If the scaling works, then I hope to eventually remove the htmlviewer source entirely from templot3.

Cheers,
Alistair.

 

[Martin Wynne]

Hi Alistair,

Is the HtmlViewer package now installed by default in Lazarus 2.0.10 ?

The reason I included all the HtmlViewer files is so that dabblers in TemplotMEC could compile T3 straight off without needing to go through the package installation process, see:

http://85a.co.uk/forum/view_topic.php?id=3283&forum_id=8#p24933

If possible I would want to keep the option for the TemplotMEC dabblers to download a ZIP of T3 and compile it straight off in Lazarus. Github is likely to put them off (it puts me off), and we don't want to lose inspired input from actual Templot users.

There are also some of my // 85A changes in StyleUn.pas .

They are related to text scaling, and it's important for the program size slider function (bottom right of trackpad). That's especially useful for users on Mac/Linux/Wine.

Originally I purchased a copy of HtmlViewer from David Baldwin before it became open-source and was incorporated into Delphi. There may be some code remaining which requires the original files. It will need a lot of testing, especially the File Viewer and Saved Program Preferences functions if you are removing the included files from T3.

I confess to getting very nervous of talk of removing and replacing existing stuff in T3. I don't think I have enough lifetime left to re-sort out all over again stuff which I originally sorted out 20 years ago. My instinct would be to concentrate on filling in the gaps in T3 rather than re-doing what is already there. If it ain't broke don't fix it.

But of course your fork is entirely for you to decide.

The big mystery remains, and you might want to do some detective work on it (I gave up). Why does T3 run fine inside the Lazarus IDE, but gobble 30% CPU continuously when run natively? I bodged a fix for that by running it from within a command window:

But that obviously can't remain if T3 is ever to replace T2.

(I put the reminder message function on there just to give it a supposed purpose.)

I'm fairly sure it relates to the Lazarus implementation of the Application.OnIdle event (a core function of Templot), but I can't find the problem.

cheers,

Martin.

 

[Alistair Ward]

Hi Martin,

No, the THtmlViewer component is not installed by default in Lazarus, but I found I had to install it before I could load the templot3 project properly.

I'll be looking at the scaling code in StyleUn.pas next. I have an idea for how to do it so that it removes the dependency from the viewer component on the help_sheet unit. I'm hoping that any change I make can be incorporated back into the official component.

I too purchased a licence to the THtmlViewer component from Dave Baldwin many years ago... I remember making some fairly serious changes to it to support using it for reporting, including automatically populating a table of contents, and printing embedded graphics to printers at printer resolution.

I understand your nervousness about making changes, but every improvement in the code design makes it easier to understand what the program is doing, which then makes it easier to add new features...

Cheers,
Alistair.

 

[Martin Wynne]

I understand your nervousness about making changes, but every improvement in the code design makes it easier to understand what the program is doing, which then makes it easier to add new features...

Hi Alistair,

I'm trying to decide whether to hold on to 40 years worth of string and blu-tack bodging, or accept whatever you and others say is best for me.

Perhaps it's important to do whatever will attract others to get involved. T3 has hardly progressed in 3 years, so I can see something has to change if it is ever going to.

cheers,

Martin.