Hi,

I'm trying to debug a issue I found in a recent 2.6.1 update under Windows. What is the difference between the branches/fixes_2_6 and branches/fixes_2_6_0?

Also, how do I find out in SubVersion at what revision the 2.6.1 branch started?

G.

Graeme Geldenhuys wrote

Also, how do I find out in SubVersion at what revision the 2.6.1 branch started?

2.6.1 did not start yet? It will be build from fixes_2_6. According to the TortoiseSVN revision graph, fixes_2_6 has been branched from r18073.

Sometimes a 2nd or 3rd tier release is late and no release candidates were built but the builder notices some necessary changes, so another branch from fixes_2_6 is created because some people did not like when we moved tags because their scm cannot cope with this ;)

Florian Klämpfl wrote

> Also, how do I find out in SubVersion at what revision the 2.6.1 branch started? 2.6.1 did not start yet?

I meant the FPC that currently has the 2.6.1 version number... this seems to be the code in the 'fixes_2_6' branch.

It will be build from fixes_2_6. According to the TortoiseSVN revision graph, fixes_2_6 has been branched from r18073.

Thanks, that's the info I needed.

In our previous episode, Graeme Geldenhuys said:

> It will be build from fixes_2_6. According to > the TortoiseSVN revision graph, fixes_2_6 has been branched from r18073. Thanks, that's the info I needed.

(but it was still 2.5.1 for a while. It only got changed to 2.6.1 around jan 1st)

2.6.1 = fixes_2_6 - branchmoment RELEASE_2.6.0 - (fixes -> release merged revs)

So all patches before 2.6.0 branched off are in 2.6.0 (and thus not 2.6.1 specific)

Hi,

Marco van de Voort wrote

So all patches before 2.6.0 branched off are in 2.6.0 (and thus not 2.6.1 specific)

Thanks for that info. There was a bug in FPC 2.6.1 in a revision of about 2-3 weeks ago that caused all my Windows apps to crash randomly. Yet the Linux equivalent of fixes_2_6 didn't produce the same issue. It seems whatever it was, has since been fixed, because after getting another fixes_2_6 update yesterday, and a recompile of all my projects, they now work fine again. Weird, but at least everything is working again. :)

In our previous episode, Graeme Geldenhuys said:

Very strange. fixes_2_6 has hardly mutated in that period, if at all.

Hi,

Very strange. fixes_2_6 has hardly mutated in that period, could possibly be off with when I got the old fixes_2_6 checkout.

Or maybe something just went wrong with the FPC build under Windows that time. Anyway, it sure gave me a wake-up call when I was preparing a new release of our product suite. Damn lucky we test well before we push out a release. :) In the end I simply switched to using FPC 2.6.0 (due to time constraints) for that release of our software, and that worked out fine.