Re: [Geany-Devel] Scintilla Use
On 2018-03-21 09:04 PM, Lex Trotman wrote: I would think by the time older LTS distros are up to C++17, that Scintillua/LongTerm3 will have no reason to exist anymore. Old LTS will never get it, thats why they are called "stable", what I meant is the next LTS, for example Ubuntu is scheduled to release a new LTS next month which will have gcc7 in its repositories even if it uses gcc 6 by default. Yeah, I mean by the time those currently new LTS distros with C++17 are the old lowest common denominator LTS distros we support - in the future. Regards, Matthew Brush ___ Devel mailing list Devel@lists.geany.org https://lists.geany.org/cgi-bin/mailman/listinfo/devel
Re: [Geany-Devel] Scintilla Use
> I would think by the time older LTS distros are up to C++17, that > Scintillua/LongTerm3 will have no reason to exist anymore. Old LTS will never get it, thats why they are called "stable", what I meant is the next LTS, for example Ubuntu is scheduled to release a new LTS next month which will have gcc7 in its repositories even if it uses gcc 6 by default. IIUC it's only > the new maintainer of that branch needing to stick to C++11 for his own > project(s) for now and so can't use the tip of Scintilla code. I assume that > eventually the goal is to have Scintillua into the mainline or some future > external library containing all of Scintilla's lexers (see > scintilla-interest mailing list). Or at least support it there. Cheers Lex ___ Devel mailing list Devel@lists.geany.org https://lists.geany.org/cgi-bin/mailman/listinfo/devel
Re: [Geany-Devel] Scintilla Use
On 2018-03-21 08:01 PM, Lex Trotman wrote: On 22 March 2018 at 10:35, Matthew Brush wrote: Hi, Going forward, are we going to use the Scintilla LongTerm3 branch or v4.x releases? Interesting question. I had been assuming that once Neil settled it down and after the next round of LTS Linuxen were released with C++17 (at least in the repos if not by default) we would go with Scintilla 4, but so far I havn't seen any changes there that excite me greatly. I would think by the time older LTS distros are up to C++17, that Scintillua/LongTerm3 will have no reason to exist anymore. IIUC it's only the new maintainer of that branch needing to stick to C++11 for his own project(s) for now and so can't use the tip of Scintilla code. I assume that eventually the goal is to have Scintillua into the mainline or some future external library containing all of Scintilla's lexers (see scintilla-interest mailing list). If LongTerm3, is there any opposition to integrating the new LPeg lexer so we can use/write lexers in Lua/PEG? I might be interested in working on this if it's not going to be controversial. The downside to this is that we are then stuck on 3 until Mitchell ports Scintillua to 4, and if its in his Scintilla LTS 3 tree he has no incentive to port it to 4. And once its made available in Geany it See above. can't be removed. Even if technically it can be removed, once people start using it there will be no going back. Well yeah, but if one of our dependencies breaks, we break; be it Scintilla or GTK+ or whatever. And it makes Geany depend on Lua. Meh, it's small, stable and well maintained. I could imagine it being used for other stuff too, like maybe a core proxy plugin for first class Lua plugins. And does it compile Lua in, distros will probably not like that? It can be embedded or linked against a dynamic library, selected at build-time. This is easy to do in the build system. And will the Lua plugin still work given it has another copy of Lua. Should work fine. If the Lua versions are compatible, the plugin could even detect and optionally use the version inside Geany (if we exported it), its own, or some external dynamic library. Regards, Matthew Brush ___ Devel mailing list Devel@lists.geany.org https://lists.geany.org/cgi-bin/mailman/listinfo/devel
Re: [Geany-Devel] Scintilla Use
On 22 March 2018 at 10:35, Matthew Brush wrote: > Hi, > > Going forward, are we going to use the Scintilla LongTerm3 branch or v4.x > releases? Interesting question. I had been assuming that once Neil settled it down and after the next round of LTS Linuxen were released with C++17 (at least in the repos if not by default) we would go with Scintilla 4, but so far I havn't seen any changes there that excite me greatly. > > If LongTerm3, is there any opposition to integrating the new LPeg lexer so > we can use/write lexers in Lua/PEG? I might be interested in working on this > if it's not going to be controversial. The downside to this is that we are then stuck on 3 until Mitchell ports Scintillua to 4, and if its in his Scintilla LTS 3 tree he has no incentive to port it to 4. And once its made available in Geany it can't be removed. Even if technically it can be removed, once people start using it there will be no going back. And it makes Geany depend on Lua. And does it compile Lua in, distros will probably not like that? And will the Lua plugin still work given it has another copy of Lua. So I'm against adding it, especially as it is only needed for obscure languages and personal languages. Thanks Lex > > Regards, > Matthew Brush > ___ > Devel mailing list > Devel@lists.geany.org > https://lists.geany.org/cgi-bin/mailman/listinfo/devel ___ Devel mailing list Devel@lists.geany.org https://lists.geany.org/cgi-bin/mailman/listinfo/devel
[Geany-Devel] Scintilla Use
Hi, Going forward, are we going to use the Scintilla LongTerm3 branch or v4.x releases? If LongTerm3, is there any opposition to integrating the new LPeg lexer so we can use/write lexers in Lua/PEG? I might be interested in working on this if it's not going to be controversial. Regards, Matthew Brush ___ Devel mailing list Devel@lists.geany.org https://lists.geany.org/cgi-bin/mailman/listinfo/devel
Re: [Geany-Devel] Why has the universal ctags project to be merged?
Le 21/03/2018 à 17:17, Lars Paulsen a écrit : > Hello, > > I have got a maybe stupid question out of curiosity: > Why needs the universal ctags project be merged into Geany instead of > e.g. just copying it over? Lex covered it more thoroughly I guess, but it basically simply boils down to having two very similar yet different codebases. Ultimately we'd like to make it a lot easier to share work, but it requires work on both ends, as we also have features not present in U-CTags, mostly the ability to parse in-memory data (ctags is really a command-line tool, but we wanna use it as a library). So… in the end the only reason is that work is needed to get to the point where we can simply upgrade out U-CTags copy to get the new stuff. Regards, Colomban ___ Devel mailing list Devel@lists.geany.org https://lists.geany.org/cgi-bin/mailman/listinfo/devel
Re: [Geany-Devel] Why has the universal ctags project to be merged?
On 22 March 2018 at 07:17, Lars Paulsen wrote: > Hello, > > I have got a maybe stupid question out of curiosity: > Why needs the universal ctags project be merged into Geany instead of e.g. > just copying it over? > I guess the reason is not that we do not trust their test quality ;-) Its not really that we don't trust them, its because: 1) we don't use all of their parsers, so some have to be left out 2) we have parsers they don't have, which need to be combined into the same directory 3) universal ctags is the zombie of the old ubiquitous ctags come back to life, but during the dead period in between our parsers diverged from ctags ones, and although some of the changes have been merged back upstream, some have proved too complex, so changes in their parsers have to be merged into ours 4) our layout is not the same as ctags, there is a PR for this, but its a huge blob, and nobody has the several years to check it (needed no matter how much we trust the OP) and according to the OP it can't be done incrementally So it needs human intervention to merge, and: 5) we do like to check even if we trust their quality :) Cheers Lex > > Greetings, > Lars > > ___ > Devel mailing list > Devel@lists.geany.org > https://lists.geany.org/cgi-bin/mailman/listinfo/devel ___ Devel mailing list Devel@lists.geany.org https://lists.geany.org/cgi-bin/mailman/listinfo/devel
[Geany-Devel] Why has the universal ctags project to be merged?
Hello, I have got a maybe stupid question out of curiosity: Why needs the universal ctags project be merged into Geany instead of e.g. just copying it over? I guess the reason is not that we do not trust their test quality ;-) Greetings, Lars ___ Devel mailing list Devel@lists.geany.org https://lists.geany.org/cgi-bin/mailman/listinfo/devel