On Wed, Feb 19, 2014 at 3:55 PM, Rob Weir <robw...@apache.org> wrote: > On Wed, Feb 19, 2014 at 3:31 PM, jan i <j...@apache.org> wrote: >> On 19 February 2014 20:22, Louis Suárez-Potts <lsuarezpo...@gmail.com>wrote: >> >>> hi, >>> >>> >>> On 19 February 2014 14:05, Donald Whytock <dwhyt...@apache.org> wrote: >>> > On Tue, Feb 18, 2014 at 12:34 PM, Rob Weir <robw...@apache.org> wrote: >>> > >>> >> http://www.italovignoli.org/2014/02/language-support-of-office-suites/ >>> >> >>> >> Don't you love it when they come up with these false comparisons? >>> >> >>> >> If you look a little bit closer you see that they are releasing and >>> >> counting languages where the UI is only 15% translated. So yes, if >>> >> you are willing to release incomplete work then you can claim to >>> >> "support" more languages. But what kind of support is this? >>> >> >>> >> A specific example: Tartar (15% UI translated) >>> >> >>> >> I thought OOo had a requirement for 80% completion before releasing a >>> >> translation. With AOO we made the requirement be 100%. LO releases >>> >> 15% complete UI translations ?! >>> >> >>> >> Of course, we shouldn't judge their release criteria. That is their >>> >> business (and their users) not ours. But when they make false >>> >> comparisons in a table, comparing apples-to-oranges, then we ought to >>> >> note it. It is not fair to claim lower standards are the same as >>> >> greater results. >>> >> >>> >> Another example: They've released Hebrew support at 90% complete. We >>> >> have Hebrew support at 96% complete, but we have not released it yet. >>> >> >>> >> Another example: Our Icelandic translation (unreleased) is 95% >>> >> complete. Theirs (released) is only 88%. >>> >> >>> >> Another example: We have 36 languages at 100% complete UI >>> >> translation. LO has only 13. >>> >> >>> >> Look at the data and make your own comparisons: >>> >> >>> >> https://translate.apache.org/projects/aoo40/ >>> >> >>> >> https://translations.documentfoundation.org/projects/libo_ui/ >>> > >>> > >>> > More recently posted on the blog by the author: >>> > >>> > "italovignoli February 19, 2014 at 2:12 >>> > am< >>> http://www.italovignoli.org/2014/02/language-support-of-office-suites/#comment-6950 >>> > >>> > Although the comparison was between LibreOffice and Microsoft Office, I >>> > have updated the table to reflect the situation at AOO provided by that >>> > project." >>> > >>> > In all fairness, even if they're not complete, it's still an impressive >>> > list of languages LO is claiming. >>> > >>> >>> :-) Your language undoes their claim. This claim of theirs is not new. >>> When we were doing OOo, we claimed, too, >100 languages, until I tried >>> insisting that we really needed to clarify what language support >>> meant. (Something similar occurs with format support.) Further, it >>> does not matter much if a language is localized to, say, Klingon (as >>> we tried), only to have it be forgotten by tomorrow's children and >>> left unmaintained. It's a truth about open source that seldom goes >>> acknowledged, that what counts is not what you did yesterday or even >>> today but what others will do with all that over the stretch of >>> tomorrows to come. >>> >>> Let's be as ruthlessly real uras possible. Money decisions, not >>> marketing lard, are at stake. >>> >> >> before we we get too far out on theoretical claims. It is true we and LO >> have different release policies and so be it. >> >> But has anyone looked at their po files, I just spent a couple of hours >> doing so, and reality is that they have many languages that are far more >> complete that ours (see https://translations.documentfoundation.org/) >> > > How closely did you look? We only put languages in Pootle where a > volunteer has requested them. LO has 20 translations that have never > been edited. Some are at 0% complete. Many more that have not been > touched in over a year. > > If you want apples-to-apples comparisons then you should look at the > ones we have in SVN: > > http://svn.apache.org/viewvc/openoffice/trunk/extras/l10n/source/ > > Both projects have their share of incomplete, neglected translations. > Nothing wrong with that. But comparing the one project's list of > active translations with another project's list of inactive ones is > not very fair. >
Here's what I'd call a fair comparison. Use the criterion that OOo used for shipping a supported language -- 90% UI completion. By that criterion LO has 47 "complete" translations and AOO has 42. This is not a very large difference. -Rob >> They also have the web site templates in pootle, very elegant, something I >> would like us to have. >> > > We've been doing this in SVN, via a website template. If we had > MDText support for Pootle we could do more. > >> If we claim they compare oranges and apples we should not start doing the >> same, at least not without having looked at the facts. >> > > Maybe we're not looking at the same facts? > >> So, yes maybe LO takes the statement to the limit, but that is called >> marketing, and in general accepted. >> > > Again, if you are not looking at the same facts and do not know that > you are not looking at the same facts then you have been deceived , > not just subjected to marketing. > > Regards, > > -Rob > >> And yes I am still very frustrated about the fact that translators have to >> translate the same text twice. THAT would be a good answer to the blog. >> >> rgds >> Jan Iversen. >> >> >> >> >> >>> >>> >>> > Don >>> >>> best >>> louis >>> >>> --------------------------------------------------------------------- >>> To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org >>> For additional commands, e-mail: dev-h...@openoffice.apache.org >>> >>> --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org For additional commands, e-mail: dev-h...@openoffice.apache.org
