Lazarus Release 2.0.6

[Martin_fr]

I don't know about the individual issues on your list (the 2nd sounds like a feature not a bug, see below). The individuals that have worked on those issues would have to comment on your items.... But in general:

The list of fixes that "made it" can be found here: https://wiki.lazarus.freepascal.org/Lazarus_2.0_fixes_branch#Fixes_for_2.0.6_.28merged.29

If the fix is on that list, then there are the following possibilities (most are rare):
- the fix is incomplete, but no one has noticed that before now. 
  It could be incomplete in "trunk" or it could be that there is another bug that does make it incomplete in fixes only.
- The fix works in trunk, but not or not always in fixes  (this is really rather rare)

If it is not on that list, but the issue was marked as fixed on the bugtracker, then the issue was not merged.
That could be
- an oversight (unfortunately that happens, I have myself squeezed issues into a busy day, and forgotten to mark them for merging)
- technical problems with merging. Maybe the fix did not work in fixes and would have needed extra tweaks, but no-one has done them.
- it was not a bug fix. Normally only bug fixes are merged.

"trunk" is the preparation for the next major release 2.2
"fixes" is a subset of the trunk changes and what goes into 2.0.6 (or 2.0.
"merge" means a change from trunk was also added to fixes

Some issues about what might be merged were discussed here: https://forum.lazarus.freepascal.org/index.php/topic,47061.0.html
Some items only got to our attention after the deadline. (After building had commenced)

[Martin_fr]

* We have just ended the update to 2.0.4 (a bare month or so ago) in all our dev machines. You release too quickly for us 

Well it's free software (as in free = freedom to ...).
You are free to update, but there is no force.

If 2.0.4 works well for you, you can keep using it.

[jamie]

2. The TShellTreeView was suppose to have an added OnAddItem event, never made it.

No, it was NOT supposed to have that in fixes branch, as I have explained to you before.
It is not a bugfix but a new feature, and therefore does not belong in the fixes branch.

If you really need this, you're probably OK with just copying shellctrsl.pas from trunk (nothing much else has changed).

Bart

Ok, I can go with that, but I also showed a problem with the LazHtmlhelp example and the problem file, which is a BUG!
I uploaded the fixed file with what I think was plenty of time to get it in there.

  Oh well, no worries, I am the middle of overwriting a bunch of files now with fixes...

 --
 The issue with me doing the work in the Shell code is that when one day when you decide to change it then my work goes out the door and compiled projects fail.. This is why I was waiting (Hoping) that you put it in there.

 Oh well. Guess I can keep the distro (corrected one) with projects compiled by it.

[Bart]

Ok, I can go with that, but I also showed a problem with the LazHtmlhelp example and the problem file, which is a BUG!
I uploaded the fixed file with what I think was plenty of time to get it in there.

And that fix was wrong (or in the wrong file).

Bart

[lucamar]

* We have just ended the update to 2.0.4 (a bare month or so ago) in all our dev machines. You release too quickly for us 

Well it's free software (as in free = freedom to ...).
You are free to update, but there is no force.

If 2.0.4 works well for you, you can keep using it.

We try to keep up to date but there are two problems: 1) Too many machines (close to 50, IIRC) and 2) machines that should be kept at some release or with alternate, fully functional installs (say, the ones where we still do maintenance for Windows 9x). And that's not counting the pure DOS (only FPC updates, thank the Gods!) and the mostly "static" CP/M dinosaurs. All in all, a little nightmare for updates: 95% in a week and the next three (or six!) months for the rest 

[jamie]

Ok Bart, what ever you say...

 All I did was change the TProcess procedural code over to the ShellEexcute code when ever a shell Link was detected in the file name... The existing code still works with TProcess and I have tested this all the way back to W2k. Works in all windows targets and up to Win10..

 If you think that was the wrong code then you must know what the correct code is which only means its a bug...

 I don't want to get into a battle of wits over this, I normally fully test changes with all versions of windows before offering a fix..

  Carry on, I'll just craw back in my corner an change the 2.0.6 distro files so they work
properly.

 Have a nice day and thanks for the update, what ever it was that got updated.


 

[Bart]

Have a nice day and thanks for the update, what ever it was that got updated.

As posted in the first post in this thread:

Here is the list of fixes for Lazarus 2.0.x:
http://wiki.freepascal.org/Lazarus_2.0_fixes_branch

So, that's what got changed (fixed actually).

Bart

[dbannon]

We try to keep up to date but there are two problems: 1) Too many machines (close to 50, IIRC) ..... and the mostly "static" CP/M dinosaurs.

Oh, wow, CP/M, Lazarus on 50 machines ?  Lucamar, you will have to tell us a bit more about what you do please ?

Anyway, given that 206 is (almost) exclusively fixes, if you are not being bitten by those particular bugs, why would you bother to update ?

CP/M ? really ?

Davo

[willbprog9933]

Unable to install on macOS Catalina.

[Rusik]

It seems that there is misprint.
                 
"Topic: Lazrus Release 2.0.6"

[avra]

We try to keep up to date but there are two problems: 1) Too many machines (close to 50, IIRC)

I set one fpcupdeluxe based Lazarus on main dev pc, and when finished I just copy to the same dir on all other dev pc. Shortcut is copied, too. Lazarus is 32 bit (64 has some cross compilation issues), and running on WinXP, Win7 and Win10 machines. Some machines with 32-bit and some with 64-bit OS. I compile IDE with -O2 -g- -Xs -CpPENTIUM4 -CfSSE3 to have it running even on some very old WinXP machines with old CPUs. Once I set up the IDE, running it on all other machines takes less then an hour. Similar thing is with linux machines.

[lucamar]

Oh, wow, CP/M, Lazarus on 50 machines ?  Lucamar, you will have to tell us a bit more about what you do please ?

We are just kind of a loosy group of programmers working together and some of us have got too many machines. I myself have an old 386 (FreeDOS/Minix), a Pentium (FreeDOS/Windows 9x/Linux), a Pentium 4 laptop (WinXP/Linux), a couple Core Duo laptops (Win7/Linux), an Atom notebook (Win7/Linux), an I5 laptop (Win10/Linux), etc.

The thing is that between us all we should have exact (or almost) duplicates of the machines our customers have got, so we can program/test in the "peace of home" and be (mostly) sure the customer's machines won't have (many) surprises waiting for us.

Anyway, given that 206 is (almost) exclusively fixes, if you are not being bitten by those particular bugs, why would you bother to update ?

We are rather conservative in that respect (one or two of us check/test the new release for as long as needed before the rest updates) but we like to use the latest release so that if a problem arises we can make meaningfull reports. In that respect, the slow update cycle of Free Pascal helps.

CP/M ? really ?

Yeah, we have two customers who are mighty happy with their 80s computers. One of them still has (and requires support for) some Amstrad PCW. The other one has a couple Amstrads too but also uses some computers/controllers based around a Z80 which we designed for him and which run a CP/M-compatible "OS", so we can program them with Pascal MT+/Turbo Pascal. I do have a couple of heavily modified Amstrad PCW8256's for these CP/M related tasks.

As you see, pretty much your "standard" software shop

[Martin_fr]

Unable to install on macOS Catalina.

We are aware of this. Last I heard was that there were some issues to create that installer. So some work is needed. No idea how long that will take.

[wittbo]

@jwdietrich:  Did you recompile as 32Bit Carbon-based or  64bit Cocoa based IDE?
If 64 Bit, how are your experiences on MacOS Mojave?

[Ariador]

Regretfully couldn't manage to be so glad with new Lazarus 2.0.6 at my Ubuntu 19.10.

I have downloaded all 3 files from SourceForge via your link, installed FPC 3.0.4 and its sources then installed "lazarus-project_2.0.6-0_amd64.deb" and started Lazarus. And Lazarus IDE had immediately frozen my desktop at all. I had to switch to another one with Ctrl-Alt-F8 and restart. Tried then to recompile Lazarus to Qt widgetset with "make clean all LCL_PLATFORM=qt" but had finally just a fatal error:

    ...
    (9022) Compiling resource ../units/x86_64-linux/qt/lazarus.or
    (9015) Linking ../lazarus
    /usr/bin/ld: warning: ../link.res contains output sections; did you forget -T?
    /usr/bin/ld: cannot find -lQt4Pas
    /usr/share/lazarus/2.0.6/ide/lazarus.pp(165,1) Error: (9013) Error while linking
    /usr/share/lazarus/2.0.6/ide/lazarus.pp(165,1) Fatal: (10026) There were 1 errors compiling module, stopping
    Fatal: (1018) Compilation aborted
    make[2]: *** [Makefile:4144: lazarus] Error 1
    make[2]: Leaving directory '/usr/share/lazarus/2.0.6/ide'
    make[1]: *** [Makefile:4572: ide] Error 2
    make[1]: Leaving directory '/usr/share/lazarus/2.0.6/ide'
    make: *** [Makefile:3381: ide] Error 2

So I must assume, new Lazarus is not happy with GTK2 and GTK3 (my current distro provided GTK 2.24.32 and GTK 3.24.12) and also it's not friendly with some Qt5 (namely Qt 5.12 with 5.62 libs as it comes with distro). For there's only Qt5 at my system, lQt4Pas can't be found in /usr/bin/ld which is a link to x86_64-linux-gnu-ld.bfd, I guess.

So what can you suggest for the case?
Do I need to downgrade to older GTK or Qt or there is a fix? I'd not like to dig into all these makefiles myself