It's perfectly legitimate for paying subscribers to have a say in the direction they would like to see Slackware take. Yes, disk space is cheap, but skin is expandable as well. Does that mean you should keep eating all you like? Some of us just don't like having a 9GB default install. I start to retch when I see it go over 2GB.

The ideal for me would be three options: Server (no desktop environments, no multimedia, no games, no DE or X libraries in L); Workstation (choice of DEs, everything except multimedia and games); and Desktop (everything).

This is obviously a lot of work but people shouldn't be criticised over and over again for offering their opinion about how things could be done better. There's a little bit of Lord of the Flies bullying going on lately with Darth and it shouldn't be happening here.

3 members found this post helpful.

No, but I intend to prune every piece of Non-Plasma Qt5-based software, if/when Plasma 5 will be adopted by Slackware.

WHY? To go back to KDE4 and Qt4. Until I manage to compile them, using my experience of C/C++ programmer (working with/on Qt, specially)

Then, finally probably I will go full ahead to XFCE (or Mate?) realm, avoiding the Plasma 5 (and Qt5) like the devil avoid a church.

From the man page:Anyone is free to break one's system, of course.

Well, thanks for giving proof that this is not about the merits of software but purely a crusade born out of spite that you were not able to upgrade your old computer to Slackware 14.2.
Again, you do not have to use PLasma 5 when it geta added to Slackware. But KDE4 stopped being developed even before Slackware 14.2 was released and only the kdelibs package gets long term support (LTS) updates. Don't expect security or functional updates in that area.

I am of course biased as well. I have been compiling KDE sources for many many years and using KDE for even more years. I do think that released of KDE were awful efforts at cloning MS Windows and only with KDE4 there was a serious effort to create something unique. The KDE5 aka Plasma5 is progressing along that philosophical path. I really enjoy the speed and cleanliness of Plasma5. I would never want to go back to any of the older releases of KDE on any of my contemporary machines. The older machines I have, they work fine with XFCE or LXQT. Choice enough.

3 members found this post helpful.

Eric, I do no crusade. I just learned my lesson of not proper testing -current when was the proper times, nothing more.

And, BTW... My "old" computers includes two Phenom x4 and two Buldozer x8, with memory ranging from 16GB to 32 GB. As hard drives, I have several "terabytes", SDDs, the usual dish.

I know, I know, no Ryzens and DDR4 there, but I am very satisfied about their actual performances.

To be very clear, what is hard to understand that I love KDE4, as I love Windows 7 too?

And what is hard to understand that I hate with passion that KDE5, as I hate also Windows 8 (and beyond), for their Metro interface and chunked usability?

What you call cleanliness, for me is straight ugliness. I know, different persons with different tastes.

What I want? I just want KDE4 on my own computers, without to arrive to the need to "fork" Slackware, maintaining a very customized tree, because I have no time for.

I understand, the "modern way" is Metro, with smartphone like interface in your screen, and Plasma 5 developers tried their best to borrow from it.

BUT, is too much to ask that that Plasma 5 goodness and its dependencies to be isolated in future in their very own series, as for me to be easy to NEVER install them?

What if one wants XFCE only and he needs only Firefox and Thunderbird? What if one wants a web server or mail server? What if one have serious storage size constraints, where is no room for a crashing and fat desktop environment like your beloved Plasma 5?

I understand that you love Plasma 5, but please simplify the life of those who do not want it at all!

2 members found this post helpful.

You are perfectly right, but let's start with just having the KDEL series there.

How much work would it be to remove KDE4 and add KDE5. Considering Eric has already done all the leg work, it would be simple. To start separating them into various dependency folders is going to be extra work initially. Eric already pointed out that some of the KDE5 "dependencies" he has would be useful outside of KDE.

I'm not saying there's anything wrong with requesting, but there's nothing wrong with me questioning it. I'm fine if people question any suggestions I might have. We obviously aren't going to agree on everything.

And Darth brings it on himself. He specifically tries to be as outspoken as possible, calling plasma5 "filthy" and "crap" just in this thread. Not to mention all his demands, especially when they contradict each other. He despises new releases of software in -current thinking Slackware is going into a rolling-release type situation like Arch, but then he demands that Pat include PHP7 shortly after its release. He wants no changes to any of the software unless it is a specific piece of software he wants.

But having 3 options (server, workstation, and desktop) is historically something Slackware has never provided. It is a thing provided by other distros, but with Slackware, requires users to manually trim their installs. It seems like this would be better suited to user provided tag files rather than new options within the installer. Trying to break Slackware into specific molds isn't going to work for everyone anyway. What if a server needs X? What if a developer running the workstation wants to be able to play music, or watch youtube videos? They'd need multimedia.

I'm not saying it's bad to have multiple choices, but Slackware has always suggested to just install everything. This would create extra work for Pat to create and maintain.

But then, I've largely been a proponent of letting Pat decide what should be done with Slackware. If that involves taking Darth's (or your) suggestions and ignoring mine, I'm fine with that. But my opinions on how this should be handled are no greater or lesser than the opinions held by Darth, and we're both free to suggest what we feel should happen with Slackware. And we should both be prepared to take any criticism from others if they feel our suggestions shouldn't be done.

Then, in the (far?) future all our BDFL have to do is to grab Eric's KDE5 "dependencies" and to put them into KDEL.

Anyway, he will have to edit the tag files and so on from L series, then writing some new from scratch would be the same work as complexity.

Maybe are some news for you, but today is usual to use both PHP 5.6.x and 7.x. I repeat: both of them. The reasons we can discuss in another thread.

And it is very possible to have both on a Slackware system. For example, now I have installed PHP 7.1.9 on one of my computers and works well along with the Slackware original one.

BTW, if you remember, I have the habit to test myself very well the things, before to raise them on discussions here.

You are kidding, right? So, I can have the bright idea to watch Youtube videos on my server hosted in Japan, while me living on Romania? And I will gain what?

Please explain how adding the KDEL series in future, compared with dumping all Plasma 5 deps on L series, will creates extra work for Patrick Volkerding!

Feel free to enter in infinitesimal details, I know very well how to setup a package series in Slackware.

Just make a full Slackware installation, run depfinder on all packages to find their dependencies and use the output (either as .dep files or to feed the REQUIRED fields of a PACKAGES.TXT), and use that information to remove all packages that are not deps of packages you'd like to keep (depfinder can take care of Python deps but not yet of other scripts like Ruby ones, so you'll have do that yourself case occurring).

Now if you want that Slackware contributors do that for you... Good luck

Let's see the bright side: a KDEL series for the today KDE4 (from Slackware current, of course), would simplify also the life of the early adopters for the Plasma 5 goodness, when they do a fresh install.

See bottom.

I'm not saying you're wrong in requesting PHP7, just that you're inconsistent in requesting no updates to packages like mesa, but want PHP7.

Not all servers are remote. Some may prefer to use something other than the terminal with Ctrl+Alt+F something to get access to the other TTYs. I'm not saying everyone would have this requirement, but if Pat made a specific server version, you *know* someone is going to not be happy with what's included. It's either going to include too much or not enough. Same goes for any other specialized installs.

Not all the dependencies for Plasma5 are provided by Eric's ktown. They are only the additional dependencies he needed to make Plasma5 work on top of what's already provided in -current. There would have to be additional pruning of the l/ series to pull all the KDE/QT dependencies into a separate folder. That is more difficult than using a simple script or even a one-liner command to regenerate the tag files (which I'd imagine he already has). Even if he does do it manually, what would be easier? Adding tags for the 60ish (actually 50ish, since 10 are reworked versions from -current) dependencies from ktown or extracting all KDE/QT dependencies from l/ and create a new series?

Well, I am much more interested by PHP7, which is by example required by the latest major release of Laravel (5.5), and by the overall LAMP stack, than the Mesa.

To be honest, I have very modest requirements from the graphics stack: to bring up the KDE4, preferable with desktop effects, and I love when it not freeze the system at random intervals.

In the design proposed by @Gerard Lally, this could be covered well by the Desktop (everything) installation style. You are sure that you really read what he said?

Look, even on moving the packages on a separate series, I suppose that their priority (ADD/REC/OPT) will remain the same.

If we discuss about the today KDE4 deps, it is simple to write a script which check the the packages present on KDEL series, then process just the L series tag file and keep the ones present. Same on the L series.

I speculate that our BDFL already written and use something similar, to sync the tagfiles, if not, is rather trivial for him to write it in less than 10-20 minutes, considering that this is the time required for me to write it.

If we discuss about the future Plasma 5 deps, he already have to edit the packages, to sync the new ones. So, after using the initial script, in either case he will have to do this job. I suppose that should be done manually.

Again, I do not think that creating the KDEL series would be more complicated than updating the actual L series.

And other people have requirements which require newer mesa versions. Why are your requirements more important than theirs? Why should PHP be updated (or PHP7 added), but not have mesa updated? I'm not saying your suggestions to add things to Slackware shouldn't happen, just that you are very inconsistent in your logic behind it.

So, it's either a very minimal server install, or I need everything? What if someone just wants a database server? Then they may say that the "server" installation contains things they don't need... mail servers, ftp servers, https servers, etc. To them, it would all be bloat. This is exactly have having pre-determined tag files could be problematic. They will be either be too much or too few packages. And it would require extra work from Pat. Instead, if this is something enough people wanted, someone could create a community like SBo to handle tag files. People could answer a series of questions or check boxes of software they need, then it could spit out the tag file they could use. Me? I'll continue to just do a full install. I haven't had a 4GB harddrive since 2001 (which I quickly upgraded to a 20GB one).

If we were to just take the deps provided in Eric's ktown repo, then I would agree with you, it'd be trivial to add them to a separate series. But, those aren't the only KDE deps. Just the ones that weren't already satisfied in -current. Pat (or Eric) would need to go through all the packages in the l/ series and see if they are tied to KDE or something else. *That* is what I am saying would take extra time. And based on your comments in this thread, I imagine you'd much prefer the latter over the former (all KDE and QT deps in a separate series).

No offence to both KDE4 and 5 users, but I would like to ask: is there a definitive guide to remove KDE and Qt, with all their dependencies?

I am not against to one or other, but simply I have no space to spare for them and I need to go only with XFCE and GTK.

I guess you talk (for now) about removing KDE4 and Qt4. So, look what Eric says about this subject: http://bear.alienbase.nl/mirrors/ali...rrent/5/README

That KDE4 removing guide is to make room for Plasma 5, which is incompatible with KDE4, but you can apply it well.

Then, you can take care by kdelibs and Qt and its dependencies.