From kde-core-devel Tue Apr 26 17:21:04 2005 From: Mashrab Kuvatov Date: Tue, 26 Apr 2005 17:21:04 +0000 To: kde-core-devel Subject: Re: Spell Checking in KDE4 (KSpell2) Message-Id: <200504261921.22318.kmashrab () sat ! physik ! uni-bremen ! de> X-MARC-Message: https://marc.info/?l=kde-core-devel&m=111453612917223 MIME-Version: 1 Content-Type: multipart/mixed; boundary="--nextPart2093471.D4oLeg4rO7" --nextPart2093471.D4oLeg4rO7 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Hi all, though I'm not a KDE developer, but a KDE user who wishes a success to the project, I'd like to say a few words on this issue. Even at the moment, from the user's point of view KSpell2 is a very nice lib, IMHO. Why? Take KSpell2 plus ASpell-0.60+, there is no need of manually setting the encoding. All "conversation" with libaspell done in UTF-8, since starting with 0.60 it supports UTF-8 naturally. So, encoding item from Settings->Spellchecker has gone (usability +10). If we imagine that by defa= ult KSpell2 uses ASpell, there is no need of setting a backend. So, spellchecki= ng settings, or more likely language settings, will consist of one list to cho= ose the language from (usability +10). Look at AbiWord, for example. So, why ASpell? What about other *Spell? I fail to see a sense of going to tomorrow with a spellchecker which still uses 8-bit encoding. If we are talking about NG technologies, I do not understand at all. A note about HSpell, according to its homepage, those guys make ASpell dictionary every = new release. BTW, I'm using Mandriva Linux 10.2, it comes with 74 ASpell-0.60 a= nd 44 MySpell dictionaries. A number of ASpell dictionaries will raise, I know= =20 many i18n teams are working on them. Mdk dropped ISpell long ago, nothing happened people still use Mdk. I know Enchant is cool, Dom and Co made a great lib. Great thing about it is that no manual setting for backend and encoding is needed, just set the language and it will try to find the best backend available. No worries abo= ut=20 encoding as well, it uses UTF-8. To my knowledge (correct me if I'm wrong) it is used in AbiWord, the others are using GtkSpell (e.g. Evo) or GnomeSpe= ll. I know there are patches for all that, but not all of GNOME is using Enchant like it is the case with KSpell in KDE. If KDE uses Enchant the spellchecking will look like app->kspell->enchant->[a-z]spell if KDE uses KSpell2 it will look like app->kspell2->[a-z]spell So, obviously, it is preferable to go with kspell2. I'm skipping the=20 advantages of having a lib in KDE CVS, it is also obvious. IMHO, if one puts the same amount of care and love into improving KSpell2 as one has to put into adjusting Enchant for KDE's need, KSpell2 will rock as much as Enchant does. =46inally, I'd like to encourage the devs to improve, at least try, KSpell2 for KDE4. Like Heiko Evermann said, it is the best time for that. I'm sure there always will be people to help. Heiko Evermann has shown his interest, I'm also a little bit of help (do not expect much from my limited abilities= ). Cheers, Mashrab. =2D-=20 Mashrab Kuvatov Ph.D student University of Bremen, IUP Home-page: www.sat.uni-bremen.de/members/mashrab PGP key: www.uni-bremen.de/~kmashrab/kmashrab.asc --nextPart2093471.D4oLeg4rO7 Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.4 (GNU/Linux) iD8DBQBCbniSiSmHahjHWuoRArt5AJ9/mdU/FFPkoHPFv69QT2klwj4KzQCcCdOw Qn4OEfKJqQRXIyWy2tHuttY= =1sNp -----END PGP SIGNATURE----- --nextPart2093471.D4oLeg4rO7--