<div>On Thu, Mar 22, 2012 at 00:04, David Edmundson <span dir="ltr"><<a href="mailto:david@davidedmundson.co.uk">david@davidedmundson.co.uk</a>></span> wrote:</div><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
2012/3/21 Martin Klapetek <<a href="mailto:martin.klapetek@gmail.com">martin.klapetek@gmail.com</a>>:<br>
<div class="im">> Hi all,<br>
><br>
> I'd like to bring up our current repos status. Right now we have 19 ktp<br>
> repos and telepathy-logger-qt repo, totaling it to 20 repos for ktp. From<br>
> these are 5 repos unmaintained/not developed (the old nepomuk stuff, kded<br>
> launcher and testlib/tool). While I do understand why we have so many repos,<br>
> I think it's good time to step back and look at it.<br>
><br>
> I know our philosophy is to have small, separate components, which can be<br>
> easily exchanged for others. But let's face it - there are no others and<br>
> probably won't be anytime soon. Of course you can use Empathy in combination<br>
> with ktp, but I've tried it few times and it does not work that well and I<br>
> don't know anyone using it like that (also there wasn't a single bug<br>
> report/wish/mail mentioning cooperation with Empathy). I don't want to<br>
> create one single app, but just group few repos together, the components are<br>
> still going to be separated and fully exchangeable. We'll just provide<br>
> smaller amount of packages with easier way to install (and everybody<br>
> installs all our stuff anyway).<br>
><br>
> Do we really need a separate repo for every single tool/utility we add to<br>
> our suite? These could be easily grouped under one single repo, say<br>
> ktp-utils. For all our plasma-stuff, David is going to create one single<br>
> repo. I think it would be good to extend it to others as well.<br>
><br>
> I propose to merge some repos into one and create several "super repos",<br>
> basically just a repo with simple subfolders, compilable all at once (super<br>
> CMakeLists.txt), here's the scheme:<br>
<br>
</div>When? 0.4 or just long term thinking? We've already done a massive<br>
shift around which was a massive pain in the ass so that we (quote)<br>
"on't have to change them again".<br></blockquote><div><br></div><div>I'd go ktp-utils for 0.4, the rest is long-term (if at all). However this is not as massive change as renaming, it's just moving stuff around.</div>
<div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im"><br>
> ktp-utils<br>
> - ktp-kipi-plugin<br>
> - ktp-send-file<br>
> - ktp-ssh-contact<br>
> . ktp-kopete-logs-import(?)<br>
<br>
</div>We need to be a bit careful about making an app for every single stream tube.<br>
<div class="im"><br>
><br>
> ktp-workspace-integration<br>
> - ktp-contact-runner<br>
> - ktp-kded-module<br>
><br>
> ktp-plasma<br>
> - ktp-contact-applet<br>
> - ktp-presence-applet<br>
</div>> - ktp-contact-list-applet(?) - this is already being merged into ktp-contact-applet. I am also intending on merging the chat plasmoid in here too. I think this is already underway.<br>
<div class="im"><br>
<br>
><br>
> ktp-base<br>
> - ktp-accounts-kcm<br>
> - ktp-approver<br>
> - ktp-auth-handler<br>
> - ktp-call-ui<br>
> - ktp-common-internals<br>
> - ktp-contact-list<br>
> - ktp-filetransfer-handler<br>
> - ktp-nepomuk-service<br>
> - ktp-text-ui<br>
<br>
</div>I'm not so sure about this.<br></blockquote><div><br></div><div>This set could actually stay the way it is now (ie. no super-repo).</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
<br>
> ktp-unmaintained<br>
> - ktp-kde<br>
Controversial! Aren't you the one maintaining it?<br></blockquote><div><br></div><div>No, we're working on kpeople, which is a different library.</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im">> - ktp-launcher-kded<br>
> - ktp-presence-dataengine<br>
> - ktp-test-tool<br>
> - ktp-testlib<br>
><br>
</div>This is pointless, they should just die. No point merging them to kill them.<br></blockquote><div><br></div><div>Last time we spoke about this George said it's bad to "just kill them" as they still contain valuable code. This is not exactly true for the test-tool, which was extended into full contact list. The testlib is currently being used by Vishesh to test the Nepomuk stuff. The dataengine is afair broken and crappy. The launcher is not being shipped nor used.</div>
<div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im">><br>
> It's just an idea and I'd like to hear your opinions, especially from<br>
> packagers, so please speak up :)<br>
><br>
<br>
</div>I agree our situation is a bit annoying, but it's not /that/ broken either.<br>
<br>
In theory we make:<br>
Packagers lives easier (not splitting up binaries from a repo into<br>
packages (not sure if true))<br>
Our lives easier (merging is hardly ever a problem due to this, and<br>
each module has it's own maintainers)<br>
<br>
So who does that leave?<br>
people who want to compile our code from git themselves for some<br>
reason. They should just use kdesrc-build and I don't care about these<br>
people anyway they're often annoying anyway (except alin and einar<br>
obviously)<br>
<br>
I agree we should "cut down slightly", that's why I'm merging my<br>
plasmoids into one place - but I don't see the need to go overkill the<br>
other direction either, especially as changing massively again is<br>
going to piss people off as much as help them.<br></blockquote><div><br></div><div>Of course we don't have to go the overkill way and merge just the plasma and utils stuff for example. It's not a finished plan in any way, just a brainstorming that I put up for discussion.</div>
<div><br></div><div><div>--</div><div><font color="#666666">Martin Klapetek | KDE Developer</font></div></div></div>