[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Libcdio-devel] [RFC] New API iso9660_statv2_t as API/ABI compatible

From: Pete Batard
Subject: Re: [Libcdio-devel] [RFC] New API iso9660_statv2_t as API/ABI compatible way to read files >= 4 GiB
Date: Tue, 10 Jul 2018 11:39:39 +0100
User-agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:52.0) Gecko/20100101 Thunderbird/52.9.0

Works for me.

I've only cast my vote and, at this stage, I think I have elaborated enough on the reasons why, IMO, trying to preserve the ABI at all cost has a lot more negative consequences in the long run than the *assumed* positive consequences we think everybody might gain in the short term. Plus I can also see "negative" consequences in the short term, such as the sheer amount of work Thomas has spent and continues to spend on the "preserve the ABI" path, but that's a different story...

But that's just my vote. If the consensus is with preserving the ABI, go for it.



On 2018.07.10 11:20, Rocky Bernstein wrote:
In my view, the person doing the work has priority over how it should be
done, except when there are serious negative consequences.

While I appreciate Pete's solution is in a sense complete,  Pete has also
said that he doesn't have time to work further on it.

Thomas has offered to improve it, and seems to be vested in it, so my take
is that we go in that direction.

On Tue, Jul 10, 2018 at 6:01 AM, Thomas Schmitt <address@hidden> wrote:


Pete Batard wrote:
Then, as a pragmatic person, I vote for the "only add total_size"

I agree that pragmatism should not be half-hearted.

It will last a few days until i am done with polishing the statv2
proposal. If you did not do it yourself until then, i plan to strip
down a copy of pbatard-multiextent2 to the version without extent array
At latest at that point i will ask you for some testing effort.

Then it is nearly only about the decision whether ABI is worth to be
kept compatible or whether it may change with each release.
(Pete and i made clear our differing views.)

If the decision is made for ABI change and pragmatic proposal, then
we could still bring in statv2 in the unlikely case that we ever encounter
an ISO which does not fulfill the pragmatic assumptions.
(This would of course become increasingly cumbersome after more changes
  in libiso9660.)

Have a nice day :)


reply via email to

[Prev in Thread] Current Thread [Next in Thread]