<div id="__MailbirdStyleContent" style="font-size: 10pt;font-family: Arial;color: #000000;text-align: left" dir="ltr">
Hi Jakob.<div><br></div><div>> <span style="font-family: Arial, Helvetica, sans-serif;font-size: 13px"> </span><span style="font-family: Arial, Helvetica, sans-serif;font-size: 13px">I remember something like</span></div><span style="font-family: Arial, Helvetica, sans-serif;font-size: 13px">> improved headless operations or being able to embed the VM into other</span><br style="font-family: Arial, Helvetica, sans-serif;font-size: 13px"><span style="font-family: Arial, Helvetica, sans-serif;font-size: 13px">> languages (i. e. embed a Smalltalk in a C++ program or similar),</span><br style="font-family: Arial, Helvetica, sans-serif;font-size: 13px"><span style="font-family: Arial, Helvetica, sans-serif;font-size: 13px">> though it has been a while since I've read it and I might remember it</span><br style="font-family: Arial, Helvetica, sans-serif;font-size: 13px"><span style="font-family: Arial, Helvetica, sans-serif;font-size: 13px">> wrong.</span><div class="mb_sig"></div>
<div><span style="font-family: Arial, Helvetica, sans-serif;font-size: 13px"><br></span></div><div><span style="font-family: Arial, Helvetica, sans-serif;font-size: 13px">The internal "minheadless" branch is still there. I am currently working on Squeak FFI and the ThreadedFFI plugin so that Squeak's image could also benefit from that branch's ideas without having to ship the SDL2 library, which might be one the "roadblocks" you remembered. There are several paths to achieve what "minheadless" points at. It takes a little bit more time to follow such smaller steps. Yet, I think it's worth to be careful here to not lose people in the process. :-)</span></div><div><span style="font-family: Arial, Helvetica, sans-serif;font-size: 13px"><br></span></div><div><span style="font-family: Arial, Helvetica, sans-serif;font-size: 13px">Best,</span></div><div><span style="font-family: Arial, Helvetica, sans-serif;font-size: 13px">Marcel</span></div><blockquote class="history_container" type="cite" style="border-left-style: solid;border-width: 1px;margin-top: 20px;margin-left: 0px;padding-left: 10px;min-width: 500px">
<p style="color: #AAAAAA; margin-top: 10px;">Am 24.07.2021 20:26:29 schrieb Jakob Reschke <jakres+squeak@gmail.com>:</p><div style="font-family:Arial,Helvetica,sans-serif"> <br>That comforts me a bit more. I have mostly the case in mind where a<br>Squeaker wants to pull a feature of Pharo back into the OpenSmalltalk<br>repo. Some of the stuff they presented as "We wanted to do this but<br>Squeak put roadblocks in our way" in their document stating the<br>reasons for the fork sounded interesting. I remember something like<br>improved headless operations or being able to embed the VM into other<br>languages (i. e. embed a Smalltalk in a C++ program or similar),<br>though it has been a while since I've read it and I might remember it<br>wrong.<br><br>But now that I have had another look at their default branch, I see<br>what Nicolas means. Seems like cherry-picking will already run into<br>inevitable conflicts at the file level before one could even begin<br>merging at the line level.<br><br>So I guess it will not become worse if we move our directories around as well.<br><br>Kind regards,<br>Jakob<br><br><br>Am Sa., 24. Juli 2021 um 16:29 Uhr schrieb David T. Lewis <lewis@mail.msen.com>:<br>><br>><br>> Also, the src directories only store the generated sources from VMMaker,<br>> so reorganizing that directory structure will not make it harder to<br>> merge or compare the actual VMMaker source, and it does not impact<br>> the ./platforms directories.<br>><br>> So no worries :-)<br>><br>> Dave<br>><br>><br>> On Sat, Jul 24, 2021 at 04:14:51PM +0200, Nicolas Cellier wrote:<br>> ><br>> > Hi Jakob,<br>> > Pharo fork already has made its own incompatible changes well before us<br>> > with this respect.<br>> > We can't make it fully compatible because they simplify the architecture so<br>> > as to maintain only one flavour, which we don't want to.<br>> > Since there is close to zero retro-contribution (except from Ronie), i<br>> > would not worry too much.<br>> ><br>> > Nicolas<br>> ><br>> ><br>> > Le sam. 24 juil. 2021 ?? 13:24, Jakob Reschke <jakres+squeak@gmail.com> a<br>> > ??crit :<br>> ><br>> > ><br>> > > One other thing has crossed my mind: will the moving of the directories<br>> > > make it harder to copy features and refactorings from or to the Pharo fork?<br>> > ><br>> > > Kind regards,<br>> > > Jakob<br>> > ><br>> > ><br>> > > David T. Lewis <lewis@mail.msen.com> schrieb am Di., 20. Juli 2021, 18:51:<br>> > ><br>> > >><br>> > >> On Tue, Jul 20, 2021 at 02:35:38PM +0200, Marcel Taeumel wrote:<br>> > >> ><br>> > >> > Hi all --<br>> > >> ><br>> > >> > "It's ready" again. Now with an explanation on how to review this:<br>> > >> > https://github.com/OpenSmalltalk/opensmalltalk-vm/pull/573<br>> > >> ><br>> > >> ><br>> > >><br>> > >> Hi Marcel,<br>> > >><br>> > >> The various shell scripts in the ./building directory are missing execute<br>> > >> permission, possibly because they were moved to their new location on a<br>> > >> non-unix system.<br>> > >><br>> > >> On Linux you can see the problem files with:<br>> > >><br>> > >> $ grep -rl '#!.*/bin/' building | xargs ls -l | less<br>> > >><br>> > >> I did not check each file versus the Cog branch, but a quick fix is this:<br>> > >><br>> > >> $ grep -rl '#!.*/bin/' building | xargs chmod +x<br>> > >><br>> > >> This finds all the files in the ./building directory that start with a<br>> > >> shebang line and sets them executable.<br>> > >><br>> > >> Finally, do a git commit, which will save the files with execute<br>> > >> permissions.<br>> > >><br>> > >> Dave<br>> > >><br>> > >><br>><br></lewis@mail.msen.com></jakres+squeak@gmail.com></lewis@mail.msen.com></div></blockquote></div>