Lock access to Transifex temporarily?
27 views
Skip to first unread message
Tobias Killer
Nov 24, 2017, 2:41:23 PM11/24/17
to blacklight, Desobediente, qubes-translation
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256
Hello, blacklight,
hello, Desobediente,
There is the question if (A) we should lock access (making it
impossible to contribute) to the project on Transifex until a working
bi-directional process is running (as Alain-Oliver recommended in an
email you received from Andrew on 29.10.2017 21:17 CET) or if (B) we
spend more time on Transifex helping translators.
My answer/proposal to that problem is: Yes, we should indeed lock
access for the following reasons:
- - As long as we are lack of a specification of "how to translate", we
cannot solve translation issues reported by translators.
- - The strings being translated today are pretty old (from 2015?) and
differ heavily from the versions today.
- - Since a working translation infrastructure is not within reach yet,
the translators currently do hard but useless work, even if we would
upload the current texts.
- - Doing useless stuff, as pointed out in the previous item, may annoy
people. This is probably the most important fact to consider, also
mentioned by Alain-Olivier.
What are your comments, views and proposals on this?
BTW: On the "qubes-translation" mailing list, there are many emails
with many, many proposed steps, considerations, goals and so on. You
may have a look at these (Attention: A lot of text!).
Regards,
Tobias Killer
-----BEGIN PGP SIGNATURE-----
iQIzBAEBCAAdFiEE9VOPt2yY9xS+XKzULaXvb25AsygFAloYddMACgkQLaXvb25A
syhL+A/+PWXerap9QvRDcMRVWjsIS8U7Dp71IgwSF9/6GWq7FHFlkoLY0aOL0WXh
fPWNDkQnt7utqzKYf2R7lHChn5F86lkN1vEbQ83UVujkbveMYFWlyihQmAhszzVy
2yuSCxA+SdF6O7SOyynsAzQJTfkd1Krjp/OVdQji4DYYy+nPN4GIvvUQiJHIa9v+
2x7ObHOInyxz6IBsfBsSaaHgXi1iVB9qGaAJOzunNDEiuLusGbmsAMEe0OMf64oB
5YaZPzOERVPzbciXsOkqfGXMgIsdNX0ABxs7XCMqcJ2ZTTTxox/JRbAuIzVsBFU9
R20vN61B5MP58eBvaifu7F0Iub4wp1L/l3U0Amiw+jxHni5NAyOg3EfW3zjKCvKS
kCsbPpjcjkbR/rFY1w82VfOR6jhZvENlMNrqt4VbW/EbgqAsCUBpKOdya+g/Mwaz
jhVwvwyJMlR/Cs3PHRix/YdOUJR1lyPpVFNa8rVB2P2hW7+s5JHpaq+dlW2SB2/9
iR5Jgu/kZnDykRedI6hbHCYmT2uqMZ+0h9jfIC7UTdlltc/3cnK66Aaf7I5CQuFQ
JwMXv9hrYbqHG/kMYTgbt1pswEyhIbUvF9dBWyKP/AGfZRWPg5yR4No7jgU7j/y9
wNI0B3zJB60aOZ/FvPlBNyoqPe17Cf1aiyVFEELqO+QXX66X/c8=
=PDLf
-----END PGP SIGNATURE-----
Hash: SHA256
Hello, blacklight,
hello, Desobediente,
There is the question if (A) we should lock access (making it
impossible to contribute) to the project on Transifex until a working
bi-directional process is running (as Alain-Oliver recommended in an
email you received from Andrew on 29.10.2017 21:17 CET) or if (B) we
spend more time on Transifex helping translators.
My answer/proposal to that problem is: Yes, we should indeed lock
access for the following reasons:
- - As long as we are lack of a specification of "how to translate", we
cannot solve translation issues reported by translators.
- - The strings being translated today are pretty old (from 2015?) and
differ heavily from the versions today.
- - Since a working translation infrastructure is not within reach yet,
the translators currently do hard but useless work, even if we would
upload the current texts.
- - Doing useless stuff, as pointed out in the previous item, may annoy
people. This is probably the most important fact to consider, also
mentioned by Alain-Olivier.
What are your comments, views and proposals on this?
BTW: On the "qubes-translation" mailing list, there are many emails
with many, many proposed steps, considerations, goals and so on. You
may have a look at these (Attention: A lot of text!).
Regards,
Tobias Killer
-----BEGIN PGP SIGNATURE-----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=PDLf
-----END PGP SIGNATURE-----
tokidev
Nov 25, 2017, 7:00:41 AM11/25/17
to Desobediente, blacklight, qubes-translation, Andrew David Wong
Hello, Desobediente,
hello, blacklight,
hello, Andrew,
@Desobediente: Thank you for your reply. I'm somewhat confused by it and
have some questions. So, let's check it.
@Andrew: Since this thread is about a big decision/step, I want you to
be in the loop. See [10] for the root of this thread which I've just
forwarded to you in a separate email.
@Desobediente @blacklight: In general, I encourage you to use trimmed
inline-style replies in your emails to make it easier for the recipients
to understand the context of your writings. See
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
@all: Sorry for my long emails. Since texts are often ambiguous, I try
to make clear what I want to say as well as in any way possible, thus,
resulting in longer emails.
Now the main thing:
Am 25.11.2017 um 00:00 schrieb Desobediente:
> On the Tails translation effort, the put-it-togheter work is done all
> with git. This is better not just because it's a wonderful tool, but
> because ikiwki (the site engine) itself uses git.
Here, you write that the put-it-together with Git is "better". What do
you mean by "better"? Better than what?
I've read the Tails' website about translation. As far as I can see, it
uses other means than we do with the Qubes OS doc (see below), but it
also has its advantages and disadvantages (like the use of Git for
inexperienced translators). Git is nice, I know, but in general, there
is always a trade-off, thus, there is never a "best" solution. The Qubes
OS team made some decisions in the past, already well before I appeared.
So, we have to deal with these decisions/restrictions. One example is
that Qubes OS uses Jekyll instead of Ikiwiki.
> As for the transifex people collaborating, the idea is to bring people
> who want to help but can't or won't use gobby, poedit, git, etc.
I'm confused here. In the previous paragraph, you praised Git. But here,
you don't. So, what do want to say?
> So it's more important to have the expected/proposed workflow well
> defined than to just open access to anyone do things their own way.
> Otherwise we would end up trying to put things together or even redoing
> work.
This looks like my thoughts on the need of a well-defined translation
workflow: As long as there is no proven workflow, the Transifex
translators shouldn't proceed in order to avoid mistakes when
"translating URLs" and "translating YAML headers" etc.. So, I think we
have the same idea, don't we?
Please note: The translation effort via Transifex already started around
May 2016 [9] but without a dedicated workflow AFAIK. So, since then, the
translators have access to "do things their own way", as you
(Desobediente) said, over a year now. Many old strings have been
translated meanwhile, but also many mistakes happened like translating
untranslatable YAML header key words, wrong URLs and such. And not a
single translated word appeared and appears online on the Qubes website.
So, seen from the perspective of the translators, they translate in a
potentially wrong way and even if they translate correctly, their
results don't show up on the target (Qubes) website. Thus, it's very
likely that the translation folk turn sour, as Alain-Olivier pointed out
in his email I referred you to.
So, analogously to what Alain-Olivier proposed, I suggest to
1. temporarily close access for the translators (with a hint that this
is only temporarily),
2. specify a translation workflow,
3. build an appropriate infrastructure (automated upload/download of
source files to/from Transifex etc.),
4. test it and
5. finally reopen access for the translators with references to the
workflow specification.
@all: What are your thoughts on this?
Last but not least, some general hints:
1. Please keep in mind that Qubes OS is an OS with focus on security.
Thus, decisions were and will be made with respect to security.
2. The Qubes OS website uses Jekyll [1] which is a static website
generator written in Ruby.
Thus, the source files are Markdown files with predecessing YAML
headers [2]
and HTML/CSS files with Liquid instructions [3].
3. For security reasons, Jekyll plugins (like those for l10n) will not
be used.
That's why I programmed a prototype of a language-switch [4] some
time ago, written in Liquid.
For Andrew and Marek, it looks good [5][6] but it still needs some
improvements.
In [6], there are also necessary steps to build the corresponding
infrastructure.
4. I already wrote a proposal for a translation workflow here [7],
discussed there [8].
Note that such a workflow also depends on the infrastructure we build.
The aim is to automate steps as well as possible. Thus, the
referenced workflow is still in progress and could become an easier one.
Although all these written things may look final to you, you may comment
it and propose your ideas, too, of course, since they are all just
proposals. Once we've found a solution we all agree to, we could go for it.
Regards
Tobias
[1] https://jekyllrb.com/
[2] https://jekyllrb.com/docs/frontmatter/
[3] https://github.com/Shopify/liquid/wiki
[4] https://github.com/tokideveloper/langswitch-prototype
[5]
https://github.com/QubesOS/qubes-issues/issues/2930#issuecomment-317375022
[6]
https://groups.google.com/forum/#!searchin/qubes-translation/prototype/qubes-translation/N0jj2Eu2x-E/KIR2FnXyAgAJ
[7]
https://github.com/tokideveloper/qubes-doc/blob/translation-workflow/basics_user/translation-workflow.md
[8] https://github.com/QubesOS/qubes-issues/issues/2824
[9]
https://github.com/QubesOS/qubes-issues/issues/1452#issuecomment-220690437
[10]
https://groups.google.com/forum/#!searchin/qubes-translation/lock/qubes-translation/1-vaupTtINA/smz9F-kVBAAJ
hello, blacklight,
hello, Andrew,
@Desobediente: Thank you for your reply. I'm somewhat confused by it and
have some questions. So, let's check it.
@Andrew: Since this thread is about a big decision/step, I want you to
be in the loop. See [10] for the root of this thread which I've just
forwarded to you in a separate email.
@Desobediente @blacklight: In general, I encourage you to use trimmed
inline-style replies in your emails to make it easier for the recipients
to understand the context of your writings. See
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
@all: Sorry for my long emails. Since texts are often ambiguous, I try
to make clear what I want to say as well as in any way possible, thus,
resulting in longer emails.
Now the main thing:
Am 25.11.2017 um 00:00 schrieb Desobediente:
> On the Tails translation effort, the put-it-togheter work is done all
> with git. This is better not just because it's a wonderful tool, but
> because ikiwki (the site engine) itself uses git.
Here, you write that the put-it-together with Git is "better". What do
you mean by "better"? Better than what?
I've read the Tails' website about translation. As far as I can see, it
uses other means than we do with the Qubes OS doc (see below), but it
also has its advantages and disadvantages (like the use of Git for
inexperienced translators). Git is nice, I know, but in general, there
is always a trade-off, thus, there is never a "best" solution. The Qubes
OS team made some decisions in the past, already well before I appeared.
So, we have to deal with these decisions/restrictions. One example is
that Qubes OS uses Jekyll instead of Ikiwiki.
> As for the transifex people collaborating, the idea is to bring people
> who want to help but can't or won't use gobby, poedit, git, etc.
I'm confused here. In the previous paragraph, you praised Git. But here,
you don't. So, what do want to say?
> So it's more important to have the expected/proposed workflow well
> defined than to just open access to anyone do things their own way.
> Otherwise we would end up trying to put things together or even redoing
> work.
This looks like my thoughts on the need of a well-defined translation
workflow: As long as there is no proven workflow, the Transifex
translators shouldn't proceed in order to avoid mistakes when
"translating URLs" and "translating YAML headers" etc.. So, I think we
have the same idea, don't we?
Please note: The translation effort via Transifex already started around
May 2016 [9] but without a dedicated workflow AFAIK. So, since then, the
translators have access to "do things their own way", as you
(Desobediente) said, over a year now. Many old strings have been
translated meanwhile, but also many mistakes happened like translating
untranslatable YAML header key words, wrong URLs and such. And not a
single translated word appeared and appears online on the Qubes website.
So, seen from the perspective of the translators, they translate in a
potentially wrong way and even if they translate correctly, their
results don't show up on the target (Qubes) website. Thus, it's very
likely that the translation folk turn sour, as Alain-Olivier pointed out
in his email I referred you to.
So, analogously to what Alain-Olivier proposed, I suggest to
1. temporarily close access for the translators (with a hint that this
is only temporarily),
2. specify a translation workflow,
3. build an appropriate infrastructure (automated upload/download of
source files to/from Transifex etc.),
4. test it and
5. finally reopen access for the translators with references to the
workflow specification.
@all: What are your thoughts on this?
Last but not least, some general hints:
1. Please keep in mind that Qubes OS is an OS with focus on security.
Thus, decisions were and will be made with respect to security.
2. The Qubes OS website uses Jekyll [1] which is a static website
generator written in Ruby.
Thus, the source files are Markdown files with predecessing YAML
headers [2]
and HTML/CSS files with Liquid instructions [3].
3. For security reasons, Jekyll plugins (like those for l10n) will not
be used.
That's why I programmed a prototype of a language-switch [4] some
time ago, written in Liquid.
For Andrew and Marek, it looks good [5][6] but it still needs some
improvements.
In [6], there are also necessary steps to build the corresponding
infrastructure.
4. I already wrote a proposal for a translation workflow here [7],
discussed there [8].
Note that such a workflow also depends on the infrastructure we build.
The aim is to automate steps as well as possible. Thus, the
referenced workflow is still in progress and could become an easier one.
Although all these written things may look final to you, you may comment
it and propose your ideas, too, of course, since they are all just
proposals. Once we've found a solution we all agree to, we could go for it.
Regards
Tobias
[1] https://jekyllrb.com/
[2] https://jekyllrb.com/docs/frontmatter/
[3] https://github.com/Shopify/liquid/wiki
[4] https://github.com/tokideveloper/langswitch-prototype
[5]
https://github.com/QubesOS/qubes-issues/issues/2930#issuecomment-317375022
[6]
https://groups.google.com/forum/#!searchin/qubes-translation/prototype/qubes-translation/N0jj2Eu2x-E/KIR2FnXyAgAJ
[7]
https://github.com/tokideveloper/qubes-doc/blob/translation-workflow/basics_user/translation-workflow.md
[8] https://github.com/QubesOS/qubes-issues/issues/2824
[9]
https://github.com/QubesOS/qubes-issues/issues/1452#issuecomment-220690437
[10]
https://groups.google.com/forum/#!searchin/qubes-translation/lock/qubes-translation/1-vaupTtINA/smz9F-kVBAAJ
Andrew David Wong
Nov 25, 2017, 8:43:46 AM11/25/17
to tokidev, Desobediente, blacklight, qubes-translation
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
Sounds fine to me.
Just a reminder to everyone (quoting myself from [11]):
Based on internal discussions, we've decided to put all translated files
in a separate, untrusted repo (i.e., no signed commits or tags). The
main reason is that we can't evaluate content in languages that we can't
read, and we don't want to sign content that we haven't evaluated (also
see #2652).
[11] https://github.com/QubesOS/qubes-issues/issues/1333#issuecomment-282525224
- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----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=UlQ0
-----END PGP SIGNATURE-----
Hash: SHA512
Just a reminder to everyone (quoting myself from [11]):
Based on internal discussions, we've decided to put all translated files
in a separate, untrusted repo (i.e., no signed commits or tags). The
main reason is that we can't evaluate content in languages that we can't
read, and we don't want to sign content that we haven't evaluated (also
see #2652).
- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----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=UlQ0
-----END PGP SIGNATURE-----
Andrew David Wong
Nov 25, 2017, 7:51:36 PM11/25/17
to Desobediente, tokidev, blacklight, qubes-translation
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
On 2017-11-25 10:08, Desobediente wrote:
> 2017-11-25 10:00 GMT-02:00 tokidev <tok...@posteo.de>:
>
>> Hello, Desobediente,
>>
>
> Hey.
> clarify.
> let's do it
> Like I said, being the most important part the site being built
> with ikiwiki.
> transifex, it's the core people, familiar and comfortable with git
> which uses it.
> both are markdown compilers, it really matters most for the
> sysadmin trying to put things togheter in the backend (in this
> case, its' handled by github anyway)
> find transifex easier and the people who find poedit, gobby, git,
> etc. easier. That approach makes everyone translate the way they
> can/want and also there are defined workflows to get all work
> togheter for the website and the iso image.
>
> I didn't praised git on purpose, I said it is the best tool for the
> backend people. On the other hand, the transifex is the best tool
> for most non tech translators. Maybe I overtalked about git because
> I'm a git person.
> limit on their freedom to do what they want so things can stay
> sufficiently organized for everyone. Rephrase that sentence with
> your favorite philosophical group work's motto if you wish, but
> that's the main idea.
> explaining how the workflow should be done
> the site directly. Or any of us could send pull requests to the git
> repository as well, and it would be eventually accepted (see, git
> rocks! can't avoid it!)
> e-mail, will check later.
> can/will understand it various ways. And not always will the
> communication be perfect as to what was the reason for closing
> their access. If we could skip that step it would be smoother I
> think. But I don't have an answer as to what is the best way to do
> step 1 in a smooth way nor how can we do without it, altought I
> believe it can be done both ways.
>
IMO, it sounds like the main consideration regarding Step 1 is how
much time it would take to do the rest of the steps. If it'll only
take a week, for example, then closing access probably isn't worth the
trouble. But if it'll take more than a month, it's probably worth it.
>
> ---
>
> I've read some things you've already done and it looks like it
> needs more hands to continue it and not let it sit. I'm willing to
> jump in, but I'm also short on time ATM.
>
> This e-mail I'm using don't have write permission on the
> qubes-translation googlegroups.
>
Any idea why? I just verified the settings for qubes-translation.
Everyone is allowed to post without restriction, even non-members.
There are no messages caught in the spam filter, either, so I'm not
sure why your messages aren't appearing on the Google Groups
interface. It's possible that they're just delayed and will show up
there eventually.
- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----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=TzZ4
-----END PGP SIGNATURE-----
Hash: SHA512
On 2017-11-25 10:08, Desobediente wrote:
> 2017-11-25 10:00 GMT-02:00 tokidev <tok...@posteo.de>:
>
>> Hello, Desobediente,
>>
>
> Hey.
>
>
>> @Desobediente: Thank you for your reply. I'm somewhat confused by
>> it and have some questions. So, let's check it.
>>
>
> Indeed, I have trouble even if it's e-mail communication. but let's
>
>> @Desobediente: Thank you for your reply. I'm somewhat confused by
>> it and have some questions. So, let's check it.
>>
>
> clarify.
>
>
>> @all: Sorry for my long emails. Since texts are often ambiguous,
>> I try to make clear what I want to say as well as in any way
>> possible, thus, resulting in longer emails.
>>
>
> I was trying to avoid that, but since you're from my religion,
>
>> @all: Sorry for my long emails. Since texts are often ambiguous,
>> I try to make clear what I want to say as well as in any way
>> possible, thus, resulting in longer emails.
>>
>
> let's do it
>
>
>> Here, you write that the put-it-together with Git is "better".
>> What do you mean by "better"? Better than what?
>>
>
> Better for the people involved in the Tails translation project.
>
>> Here, you write that the put-it-together with Git is "better".
>> What do you mean by "better"? Better than what?
>>
>
> Like I said, being the most important part the site being built
> with ikiwiki.
>
>
>> I've read the Tails' website about translation. As far as I can
>> see, it uses other means than we do with the Qubes OS doc (see
>> below), but it also has its advantages and disadvantages (like
>> the use of Git for inexperienced translators).
>
>
> On the contrary, the inexperienced translators are presented to
>
>> I've read the Tails' website about translation. As far as I can
>> see, it uses other means than we do with the Qubes OS doc (see
>> below), but it also has its advantages and disadvantages (like
>> the use of Git for inexperienced translators).
>
>
> transifex, it's the core people, familiar and comfortable with git
> which uses it.
>
>
>> Git is nice, I know, but in general, there is always a trade-off,
>> thus, there is never a "best" solution. The Qubes OS team made
>> some decisions in the past, already well before I appeared. So,
>> we have to deal with these decisions/restrictions. One example
>> is that Qubes OS uses Jekyll instead of Ikiwiki.
>>
>
> Jekyll is as straightforward to collaborators than ikiwiki, because
>
>> Git is nice, I know, but in general, there is always a trade-off,
>> thus, there is never a "best" solution. The Qubes OS team made
>> some decisions in the past, already well before I appeared. So,
>> we have to deal with these decisions/restrictions. One example
>> is that Qubes OS uses Jekyll instead of Ikiwiki.
>>
>
> both are markdown compilers, it really matters most for the
> sysadmin trying to put things togheter in the backend (in this
> case, its' handled by github anyway)
>
>
>> I'm confused here. In the previous paragraph, you praised Git.
>> But here, you don't. So, what do want to say?
>>
>
> That there's a separation between work being done by people which
>
>> I'm confused here. In the previous paragraph, you praised Git.
>> But here, you don't. So, what do want to say?
>>
>
> find transifex easier and the people who find poedit, gobby, git,
> etc. easier. That approach makes everyone translate the way they
> can/want and also there are defined workflows to get all work
> togheter for the website and the iso image.
>
> I didn't praised git on purpose, I said it is the best tool for the
> backend people. On the other hand, the transifex is the best tool
> for most non tech translators. Maybe I overtalked about git because
> I'm a git person.
>
>
>> This looks like my thoughts on the need of a well-defined
>> translation workflow: As long as there is no proven workflow, the
>> Transifex translators shouldn't proceed in order to avoid
>> mistakes when "translating URLs" and "translating YAML headers"
>> etc.. So, I think we have the same idea, don't we?
>>
>
> I think that too.
>
>> This looks like my thoughts on the need of a well-defined
>> translation workflow: As long as there is no proven workflow, the
>> Transifex translators shouldn't proceed in order to avoid
>> mistakes when "translating URLs" and "translating YAML headers"
>> etc.. So, I think we have the same idea, don't we?
>>
>
>
>
>> Please note: The translation effort via Transifex already started
>> around May 2016 [9] but without a dedicated workflow AFAIK. So,
>> since then, the translators have access to "do things their own
>> way", as you (Desobediente) said, over a year now.
>
>
> Which I don't think it's a desirable thing. People must have a
>
>> Please note: The translation effort via Transifex already started
>> around May 2016 [9] but without a dedicated workflow AFAIK. So,
>> since then, the translators have access to "do things their own
>> way", as you (Desobediente) said, over a year now.
>
>
> limit on their freedom to do what they want so things can stay
> sufficiently organized for everyone. Rephrase that sentence with
> your favorite philosophical group work's motto if you wish, but
> that's the main idea.
>
>
>> Many old strings have been translated meanwhile, but also many
>> mistakes happened like translating untranslatable YAML header key
>> words, wrong URLs and such.
>
>
> This is such a thing that could be avoided if we had a chart
>
>> Many old strings have been translated meanwhile, but also many
>> mistakes happened like translating untranslatable YAML header key
>> words, wrong URLs and such.
>
>
> explaining how the workflow should be done
>
>
>> And not a single translated word appeared and appears online on
>> the Qubes website.
>>
>
> I suppose it's our task to send work to mr. Wong so he can update
>
>> And not a single translated word appeared and appears online on
>> the Qubes website.
>>
>
> the site directly. Or any of us could send pull requests to the git
> repository as well, and it would be eventually accepted (see, git
> rocks! can't avoid it!)
>
>
>> So, seen from the perspective of the translators, they translate
>> in a potentially wrong way and even if they translate correctly,
>> their results don't show up on the target (Qubes) website. Thus,
>> it's very likely that the translation folk turn sour, as
>> Alain-Olivier pointed out in his email I referred you to.
>>
>
> I'm not sure I got that e-mail. I haven't accessed my riseup
>
>> So, seen from the perspective of the translators, they translate
>> in a potentially wrong way and even if they translate correctly,
>> their results don't show up on the target (Qubes) website. Thus,
>> it's very likely that the translation folk turn sour, as
>> Alain-Olivier pointed out in his email I referred you to.
>>
>
> e-mail, will check later.
>
>
>>
>> So, analogously to what Alain-Olivier proposed, I suggest to
>>
>> 1. temporarily close access for the translators (with a hint that
>> this is only temporarily), 2. specify a translation workflow, 3.
>> build an appropriate infrastructure (automated upload/download
>> of source files to/from Transifex etc.), 4. test it and 5.
>> finally reopen access for the translators with references to the
>> workflow specification.
>>
>> @all: What are your thoughts on this?
>>
>
> Is step 1 necessary for the rest? Because when we close it, people
>
>>
>> So, analogously to what Alain-Olivier proposed, I suggest to
>>
>> 1. temporarily close access for the translators (with a hint that
>> this is only temporarily), 2. specify a translation workflow, 3.
>> build an appropriate infrastructure (automated upload/download
>> of source files to/from Transifex etc.), 4. test it and 5.
>> finally reopen access for the translators with references to the
>> workflow specification.
>>
>> @all: What are your thoughts on this?
>>
>
> can/will understand it various ways. And not always will the
> communication be perfect as to what was the reason for closing
> their access. If we could skip that step it would be smoother I
> think. But I don't have an answer as to what is the best way to do
> step 1 in a smooth way nor how can we do without it, altought I
> believe it can be done both ways.
>
IMO, it sounds like the main consideration regarding Step 1 is how
much time it would take to do the rest of the steps. If it'll only
take a week, for example, then closing access probably isn't worth the
trouble. But if it'll take more than a month, it's probably worth it.
>
> ---
>
> I've read some things you've already done and it looks like it
> needs more hands to continue it and not let it sit. I'm willing to
> jump in, but I'm also short on time ATM.
>
> This e-mail I'm using don't have write permission on the
> qubes-translation googlegroups.
>
Any idea why? I just verified the settings for qubes-translation.
Everyone is allowed to post without restriction, even non-members.
There are no messages caught in the spam filter, either, so I'm not
sure why your messages aren't appearing on the Google Groups
interface. It's possible that they're just delayed and will show up
there eventually.
- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----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=TzZ4
-----END PGP SIGNATURE-----
Andrew David Wong
Nov 25, 2017, 8:02:39 PM11/25/17
to Desobediente, tokidev, blacklight, qubes-translation
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
On 2017-11-25 10:09, Desobediente wrote:
Hash: SHA512
> I believe the qubes team could delegate signing people for each
> language?
>
Do you mean, e.g., designated a trusted person(s) who knows language X
to review and sign all translations in language X?
That might be an improvement, but it seems unlikely that we'd have
trusted person for every language. I suppose we could do this only for
the languages for which we have trusted people, which might be better
than not doing it at all.
Another consideration is that this could (depending on how we design
the workflow) delay translations from going live. For example, if we
only have one trusted person for language X, and that person is backed
up with dozens of translated doc pages to review in their spare time,
it could be months before the pages go live. (Alternatively, we could
push them live with a warning that they haven't been reviewed yet.)
> I'll find that issue and make a github comment on that eventually.
>
I'm not sure which issue you're referring to, but here are the two I
mentioned in my previous message (and the rest are linked from the
second one):
https://github.com/QubesOS/qubes-issues/issues/1333
https://github.com/QubesOS/qubes-issues/issues/2652
- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----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=9A5A
-----END PGP SIGNATURE-----
Tobias Killer
Nov 26, 2017, 4:55:55 AM11/26/17
to Andrew David Wong, Desobediente, blacklight, qubes-translation
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256
Am 26.11.2017 um 01:51 schrieb Andrew David Wong:> On 2017-11-25
>>> I've read the Tails' website about translation. As far as I
>>> can see, it uses other means than we do with the Qubes OS doc
>>> (see below), but it also has its advantages and disadvantages
>>> (like the use of Git for inexperienced translators).
>
>
>> On the contrary, the inexperienced translators are presented to
>> transifex, it's the core people, familiar and comfortable with
>> git which uses it.
@Desobediente: I don't get it due to English problems, sorry.
I guess you want to say that
- - the core developers of Tails are familiar with Git and use it and
- - the Git-inexperienced translators don't need to use Git and can
translate using the Transifex interface instead.
>>> Git is nice, I know, but in general, there is always a
>>> trade-off, thus, there is never a "best" solution. The Qubes OS
>>> team made some decisions in the past, already well before I
>>> appeared. So, we have to deal with these
>>> decisions/restrictions. One example is that Qubes OS uses
>>> Jekyll instead of Ikiwiki.
>>>
>
>> Jekyll is as straightforward to collaborators than ikiwiki,
>> because both are markdown compilers, it really matters most for
>> the sysadmin trying to put things togheter in the backend (in
>> this case, its' handled by github anyway)
@Desobediente: Okay.
>>> I'm confused here. In the previous paragraph, you praised Git.
>>> But here, you don't. So, what do want to say?
>>>
>
>> That there's a separation between work being done by people
>> which find transifex easier and the people who find poedit,
>> gobby, git, etc. easier. That approach makes everyone translate
>> the way they can/want and also there are defined workflows to get
>> all work togheter for the website and the iso image.
@Desobediente: I see. This is what I've guessed above.
>> I didn't praised git on purpose, I said it is the best tool for
>> the backend people. On the other hand, the transifex is the best
>> tool for most non tech translators. Maybe I overtalked about git
>> because I'm a git person.
@Desobediente: Maybe. ;-)
>>> This looks like my thoughts on the need of a well-defined
>>> translation workflow: As long as there is no proven workflow,
>>> the Transifex translators shouldn't proceed in order to avoid
>>> mistakes when "translating URLs" and "translating YAML
>>> headers" etc.. So, I think we have the same idea, don't we?
>>>
>
>> I think that too.
@Desobediente: Nice!
>>> Please note: The translation effort via Transifex already
>>> started around May 2016 [9] but without a dedicated workflow
>>> AFAIK. So, since then, the translators have access to "do
>>> things their own way", as you (Desobediente) said, over a year
>>> now.
>
>
>> Which I don't think it's a desirable thing. People must have a
>> limit on their freedom to do what they want so things can stay
>> sufficiently organized for everyone. Rephrase that sentence with
>> your favorite philosophical group work's motto if you wish, but
>> that's the main idea.
@Desobediente: Sorry, I've English problems again.
I just wanted to say that currently there is chaos at Transifex, due
to the lack of an appropriate workflow and of translation rules. So,
IMHO, it's better to pause that until we've both a workflow and
translation rules.
>>> Many old strings have been translated meanwhile, but also many
>>> mistakes happened like translating untranslatable YAML header
>>> key words, wrong URLs and such.
>
>
>> This is such a thing that could be avoided if we had a chart
>> explaining how the workflow should be done
Exactly. As I just mentioned.
>>> And not a single translated word appeared and appears online
>>> on the Qubes website.
>>>
>
>> I suppose it's our task to send work to mr. Wong so he can
>> update the site directly. Or any of us could send pull requests
>> to the git repository as well, and it would be eventually
>> accepted (see, git rocks! can't avoid it!)
I see, I see, you really love Git. And it rocks indeed, yeah! ;-) But
this here is not Tails. This is Qubes OS and things happen differently
here.
The current rough idea is that translations will be made *exclusively*
via Transifex. So, currently, Qubes OS won't accept translations that
have been made using Git branches (this is different from the Tails
workflow). Since Qubes OS doesn't have enough people involved in
translation, they cannot afford to review and accept/reject
Git-branch-made translations. If Qubes OS had a lot more of helping
hands, like Tails has, this could be an additional way. But currently,
they can't afford it.
>>> So, seen from the perspective of the translators, they
>>> translate in a potentially wrong way and even if they translate
>>> correctly, their results don't show up on the target (Qubes)
>>> website. Thus, it's very likely that the translation folk turn
>>> sour, as Alain-Olivier pointed out in his email I referred you
>>> to.
>>>
>
>> I'm not sure I got that e-mail. I haven't accessed my riseup
>> e-mail, will check later.
Indeed, it has been sent to your riseup account.
Subject is "Fwd: Re: L10n".
Date/time is 29.10.2017 21:17 (CET).
>>> So, analogously to what Alain-Olivier proposed, I suggest to
>>>
>>> 1. temporarily close access for the translators (with a hint
>>> that this is only temporarily), 2. specify a translation
>>> workflow, 3. build an appropriate infrastructure (automated
>>> upload/download of source files to/from Transifex etc.), 4.
>>> test it and 5. finally reopen access for the translators with
>>> references to the workflow specification.
>>>
>>> @all: What are your thoughts on this?
>>>
>
>> Is step 1 necessary for the rest? Because when we close it,
>> people can/will understand it various ways. And not always will
>> the communication be perfect as to what was the reason for
>> closing their access. If we could skip that step it would be
>> smoother I think. But I don't have an answer as to what is the
>> best way to do step 1 in a smooth way nor how can we do without
>> it, altought I believe it can be done both ways.
@Desobediente: Step 1 IS necessary for not to hurt the translators'
feelings.
Imagine you're a translator at Transifex. You spend many days in
thoroughly translating. After a couple of weeks later, you visit the
Qubes OS website and wonder why you can't see your work there. So, you
ask the Qubes OS developers where your hard work is. And the only
correct (and honest!) answer is that the stuff you translated is much
too old (there are much newer versions of the texts while the old
stuff you translated is totally outdated) or it contains a lot of
mistakes (like incorrect URL translating, wrong YAML front matter
translating etc.). How would you feel now?
Also note that closing access does NOT mean saying "We don't like your
work." or something. We just could thank them for their hard work and
say that due to unsolved problems we need some time to make
translations easier and errors avoid. Meanwhile the translators may
translate other projects on Transifex. When we're done, we can re-open
the access again.
I think that we are not the first project where such a step is
necessary. I could ask Transifex for help. I'm quite sure that they
know how to do it the smooth way.
> IMO, it sounds like the main consideration regarding Step 1 is how
> much time it would take to do the rest of the steps. If it'll only
> take a week, for example, then closing access probably isn't worth
> the trouble. But if it'll take more than a month, it's probably
> worth it.
@all: I think that it'll take more than a month, Andrew. Also remember
that the current stuff there is very old. So, current translations are
a waste of time anyway.
>
>
>> ---
>
>> I've read some things you've already done and it looks like it
>> needs more hands to continue it and not let it sit. I'm willing
>> to jump in, but I'm also short on time ATM.
>
>> This e-mail I'm using don't have write permission on the
>> qubes-translation googlegroups.
>
>
> Any idea why? I just verified the settings for qubes-translation.
> Everyone is allowed to post without restriction, even non-members.
> There are no messages caught in the spam filter, either, so I'm
> not sure why your messages aren't appearing on the Google Groups
> interface. It's possible that they're just delayed and will show
> up there eventually.
@Desobediente: Or do you mean access to a Google account?
Or is it due to blocked scripts (javascript etc.) or something?
How do you know that you don't have write permission? Are there error
messages? Where and when do they show up?
Regards
Tobias
-----BEGIN PGP SIGNATURE-----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=Qnst
-----END PGP SIGNATURE-----
Hash: SHA256
Am 26.11.2017 um 01:51 schrieb Andrew David Wong:> On 2017-11-25
10:08, Desobediente wrote:
>> 2017-11-25 10:00 GMT-02:00 tokidev <tok...@posteo.de>:
>
>>> Hello, Desobediente,
>>>
>
>> Hey.
>
>
>>> @Desobediente: Thank you for your reply. I'm somewhat confused
>>> by it and have some questions. So, let's check it.
>>>
>
>> Indeed, I have trouble even if it's e-mail communication. but
>> let's clarify.
>
>
>>> @all: Sorry for my long emails. Since texts are often
>>> ambiguous, I try to make clear what I want to say as well as in
>>> any way possible, thus, resulting in longer emails.
>>>
>
>> I was trying to avoid that, but since you're from my religion,
>> let's do it
>
>
>>> Here, you write that the put-it-together with Git is "better".
>>> What do you mean by "better"? Better than what?
>>>
>
>> Better for the people involved in the Tails translation project.
>> Like I said, being the most important part the site being built
>> with ikiwiki.
@Desobediente: I see, okay, it's a Tails thing. Not related to Qubes OS.
>> 2017-11-25 10:00 GMT-02:00 tokidev <tok...@posteo.de>:
>
>>> Hello, Desobediente,
>>>
>
>> Hey.
>
>
>>> @Desobediente: Thank you for your reply. I'm somewhat confused
>>> by it and have some questions. So, let's check it.
>>>
>
>> Indeed, I have trouble even if it's e-mail communication. but
>> let's clarify.
>
>
>>> @all: Sorry for my long emails. Since texts are often
>>> ambiguous, I try to make clear what I want to say as well as in
>>> any way possible, thus, resulting in longer emails.
>>>
>
>> I was trying to avoid that, but since you're from my religion,
>> let's do it
>
>
>>> Here, you write that the put-it-together with Git is "better".
>>> What do you mean by "better"? Better than what?
>>>
>
>> Better for the people involved in the Tails translation project.
>> Like I said, being the most important part the site being built
>> with ikiwiki.
>>> I've read the Tails' website about translation. As far as I
>>> can see, it uses other means than we do with the Qubes OS doc
>>> (see below), but it also has its advantages and disadvantages
>>> (like the use of Git for inexperienced translators).
>
>
>> On the contrary, the inexperienced translators are presented to
>> transifex, it's the core people, familiar and comfortable with
>> git which uses it.
I guess you want to say that
- - the core developers of Tails are familiar with Git and use it and
- - the Git-inexperienced translators don't need to use Git and can
translate using the Transifex interface instead.
>>> Git is nice, I know, but in general, there is always a
>>> trade-off, thus, there is never a "best" solution. The Qubes OS
>>> team made some decisions in the past, already well before I
>>> appeared. So, we have to deal with these
>>> decisions/restrictions. One example is that Qubes OS uses
>>> Jekyll instead of Ikiwiki.
>>>
>
>> Jekyll is as straightforward to collaborators than ikiwiki,
>> because both are markdown compilers, it really matters most for
>> the sysadmin trying to put things togheter in the backend (in
>> this case, its' handled by github anyway)
>>> I'm confused here. In the previous paragraph, you praised Git.
>>> But here, you don't. So, what do want to say?
>>>
>
>> That there's a separation between work being done by people
>> which find transifex easier and the people who find poedit,
>> gobby, git, etc. easier. That approach makes everyone translate
>> the way they can/want and also there are defined workflows to get
>> all work togheter for the website and the iso image.
>> I didn't praised git on purpose, I said it is the best tool for
>> the backend people. On the other hand, the transifex is the best
>> tool for most non tech translators. Maybe I overtalked about git
>> because I'm a git person.
>>> This looks like my thoughts on the need of a well-defined
>>> translation workflow: As long as there is no proven workflow,
>>> the Transifex translators shouldn't proceed in order to avoid
>>> mistakes when "translating URLs" and "translating YAML
>>> headers" etc.. So, I think we have the same idea, don't we?
>>>
>
>> I think that too.
>>> Please note: The translation effort via Transifex already
>>> started around May 2016 [9] but without a dedicated workflow
>>> AFAIK. So, since then, the translators have access to "do
>>> things their own way", as you (Desobediente) said, over a year
>>> now.
>
>
>> Which I don't think it's a desirable thing. People must have a
>> limit on their freedom to do what they want so things can stay
>> sufficiently organized for everyone. Rephrase that sentence with
>> your favorite philosophical group work's motto if you wish, but
>> that's the main idea.
I just wanted to say that currently there is chaos at Transifex, due
to the lack of an appropriate workflow and of translation rules. So,
IMHO, it's better to pause that until we've both a workflow and
translation rules.
>>> Many old strings have been translated meanwhile, but also many
>>> mistakes happened like translating untranslatable YAML header
>>> key words, wrong URLs and such.
>
>
>> This is such a thing that could be avoided if we had a chart
>> explaining how the workflow should be done
>>> And not a single translated word appeared and appears online
>>> on the Qubes website.
>>>
>
>> I suppose it's our task to send work to mr. Wong so he can
>> update the site directly. Or any of us could send pull requests
>> to the git repository as well, and it would be eventually
>> accepted (see, git rocks! can't avoid it!)
this here is not Tails. This is Qubes OS and things happen differently
here.
The current rough idea is that translations will be made *exclusively*
via Transifex. So, currently, Qubes OS won't accept translations that
have been made using Git branches (this is different from the Tails
workflow). Since Qubes OS doesn't have enough people involved in
translation, they cannot afford to review and accept/reject
Git-branch-made translations. If Qubes OS had a lot more of helping
hands, like Tails has, this could be an additional way. But currently,
they can't afford it.
>>> So, seen from the perspective of the translators, they
>>> translate in a potentially wrong way and even if they translate
>>> correctly, their results don't show up on the target (Qubes)
>>> website. Thus, it's very likely that the translation folk turn
>>> sour, as Alain-Olivier pointed out in his email I referred you
>>> to.
>>>
>
>> I'm not sure I got that e-mail. I haven't accessed my riseup
>> e-mail, will check later.
Subject is "Fwd: Re: L10n".
Date/time is 29.10.2017 21:17 (CET).
>>> So, analogously to what Alain-Olivier proposed, I suggest to
>>>
>>> 1. temporarily close access for the translators (with a hint
>>> that this is only temporarily), 2. specify a translation
>>> workflow, 3. build an appropriate infrastructure (automated
>>> upload/download of source files to/from Transifex etc.), 4.
>>> test it and 5. finally reopen access for the translators with
>>> references to the workflow specification.
>>>
>>> @all: What are your thoughts on this?
>>>
>
>> Is step 1 necessary for the rest? Because when we close it,
>> people can/will understand it various ways. And not always will
>> the communication be perfect as to what was the reason for
>> closing their access. If we could skip that step it would be
>> smoother I think. But I don't have an answer as to what is the
>> best way to do step 1 in a smooth way nor how can we do without
>> it, altought I believe it can be done both ways.
feelings.
Imagine you're a translator at Transifex. You spend many days in
thoroughly translating. After a couple of weeks later, you visit the
Qubes OS website and wonder why you can't see your work there. So, you
ask the Qubes OS developers where your hard work is. And the only
correct (and honest!) answer is that the stuff you translated is much
too old (there are much newer versions of the texts while the old
stuff you translated is totally outdated) or it contains a lot of
mistakes (like incorrect URL translating, wrong YAML front matter
translating etc.). How would you feel now?
Also note that closing access does NOT mean saying "We don't like your
work." or something. We just could thank them for their hard work and
say that due to unsolved problems we need some time to make
translations easier and errors avoid. Meanwhile the translators may
translate other projects on Transifex. When we're done, we can re-open
the access again.
I think that we are not the first project where such a step is
necessary. I could ask Transifex for help. I'm quite sure that they
know how to do it the smooth way.
> IMO, it sounds like the main consideration regarding Step 1 is how
> much time it would take to do the rest of the steps. If it'll only
> take a week, for example, then closing access probably isn't worth
> the trouble. But if it'll take more than a month, it's probably
> worth it.
that the current stuff there is very old. So, current translations are
a waste of time anyway.
>
>
>> ---
>
>> I've read some things you've already done and it looks like it
>> needs more hands to continue it and not let it sit. I'm willing
>> to jump in, but I'm also short on time ATM.
>
>> This e-mail I'm using don't have write permission on the
>> qubes-translation googlegroups.
>
>
> Any idea why? I just verified the settings for qubes-translation.
> Everyone is allowed to post without restriction, even non-members.
> There are no messages caught in the spam filter, either, so I'm
> not sure why your messages aren't appearing on the Google Groups
> interface. It's possible that they're just delayed and will show
> up there eventually.
Or is it due to blocked scripts (javascript etc.) or something?
How do you know that you don't have write permission? Are there error
messages? Where and when do they show up?
Regards
Tobias
-----BEGIN PGP SIGNATURE-----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=Qnst
-----END PGP SIGNATURE-----
Andrew David Wong
Nov 26, 2017, 6:34:37 AM11/26/17
to Blacklight447, tok...@posteo.de, aindate...@gmail.com, qubes-translation
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
On 2017-11-26 05:19, Blacklight447 wrote:
> Hello tokidev,desobiente.
>
> Ive looked into the problem that you described and i too, vote for
> temponarly locking translation efforts. I think a first good step
> would be to update the translation where it is needed, but then we
> hit a second issue, do we want to provide to old documentation or
> not, and if yes, how will do this? One option could be to provide
> both versions on the same page, and highlighting which
> documentation is for which version of qubes, or keeping legacy
> documentation on a seperate page. I would like to know your
> opinion on this matter.
>
Managing version-specific Qubes documentation is a distinct issue
from translating Qubes documentation into other languages. Let's keep
these topics separate, please.
The current plan is to display information for both Qubes 3.2 and
Qubes 4.0 in clearly-labeled sections on the same page.
P.S. - Please keep the list CCed and avoid top-posting.
>
> -------- Original Message -------- On Nov 24, 2017, 8:41 PM,
nxWjkqJEf7hNpYaJp1j9IR5tp8Ybcoo5U/xeaGCdK3GCrN5CLXO6ir0NDfJL7CL1
UVJ364oYwuYhk6PeyauamO5pYXOpqw2KvQDK7UKwPzpW0n4ox/jzU5GTpqHvy4oD
f0hLCUioQzB0GIkS08zaxYm+AMy0S41q23bYzU/HhtPmiI7boY9fXmm8QtCuiC+B
twP+dvMvBr/xD7zsUPtOGYRTtdlWTUhS2SaFe6YEPNpeDKoQ2AVvrYXSJ/tgxWEI
N02dBY8QXm6KQhzK8Cd4fxvK7Rn+pmXYXTz9WNRYbRRzjKj3pX6ZIN42uVJwQw66
DhwJiNjHzCqpW53aJx9/kpDIbGbxUFFEOTK+X9Fm8N+MAIwUhook69SpX6LZOxny
pcL3ZJMawfOTTUm/2Tu1u1HqHS7KCkcqSYn9KUG61wuZTdFWtQkpVsQfNHUXlgwM
NIg8ej4Om3dlk/7vZgtdOEU+Uxe1FTog5MZrsaqcenUj2SdKFlrMW4cEqQe0ayiO
kPCC7le6SY6bTcrvycTFXthGwE6iF7RWAt5aoupt2/nae54PBw6buA21f2Uko6wp
CodMaVkog11a+9WcsWg5gmci2jYJRgFrrgqi7qwd5myjkVC9/CZSvKS0qt/gOXWc
m0f9hjWhmLhbWGmYMxyk
=mbOY
-----END PGP SIGNATURE-----
Hash: SHA512
On 2017-11-26 05:19, Blacklight447 wrote:
> Hello tokidev,desobiente.
>
> Ive looked into the problem that you described and i too, vote for
> temponarly locking translation efforts. I think a first good step
> would be to update the translation where it is needed, but then we
> hit a second issue, do we want to provide to old documentation or
> not, and if yes, how will do this? One option could be to provide
> both versions on the same page, and highlighting which
> documentation is for which version of qubes, or keeping legacy
> documentation on a seperate page. I would like to know your
> opinion on this matter.
>
Managing version-specific Qubes documentation is a distinct issue
from translating Qubes documentation into other languages. Let's keep
these topics separate, please.
The current plan is to display information for both Qubes 3.2 and
Qubes 4.0 in clearly-labeled sections on the same page.
P.S. - Please keep the list CCed and avoid top-posting.
>
> -------- Original Message -------- On Nov 24, 2017, 8:41 PM,
> Tobias Killer wrote:
>
>> Hello, blacklight, hello, Desobediente, There is the question if
>> (A) we should lock access (making it impossible to contribute)
>> to the project on Transifex until a working bi-directional
>> process is running (as Alain-Oliver recommended in an email you
>> received from Andrew on 29.10.2017 21:17 CET) or if (B) we spend
>> more time on Transifex helping translators. My answer/proposal to
>> that problem is: Yes, we should indeed lock access for the
>> following reasons: - As long as we are lack of a specification of
>
>> Hello, blacklight, hello, Desobediente, There is the question if
>> (A) we should lock access (making it impossible to contribute)
>> to the project on Transifex until a working bi-directional
>> process is running (as Alain-Oliver recommended in an email you
>> received from Andrew on 29.10.2017 21:17 CET) or if (B) we spend
>> more time on Transifex helping translators. My answer/proposal to
>> that problem is: Yes, we should indeed lock access for the
>> "how to translate", we cannot solve translation issues reported
>> by translators. - The strings being translated today are pretty
>> old (from 2015?) and differ heavily from the versions today. -
>> Since a working translation infrastructure is not within reach
>> yet, the translators currently do hard but useless work, even if
>> we would upload the current texts. - Doing useless stuff, as
>> yet, the translators currently do hard but useless work, even if
>> pointed out in the previous item, may annoy people. This is
>> probably the most important fact to consider, also mentioned by
>> Alain-Olivier. What are your comments, views and proposals on
>> this? BTW: On the "qubes-translation" mailing list, there are
>> many emails with many, many proposed steps, considerations, goals
>> and so on. You may have a look at these (Attention: A lot of
>> text!). Regards, Tobias Killer
>> probably the most important fact to consider, also mentioned by
>> Alain-Olivier. What are your comments, views and proposals on
>> this? BTW: On the "qubes-translation" mailing list, there are
>> many emails with many, many proposed steps, considerations, goals
>> and so on. You may have a look at these (Attention: A lot of
>> text!). Regards, Tobias Killer
- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----
iQIcBAEBCgAGBQJaGqasAAoJENtN07w5UDAwkegQAKZ7Ac8Enr2jZKI8sFAYbEFQ
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----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=mbOY
-----END PGP SIGNATURE-----
Andrew David Wong
Nov 26, 2017, 3:47:34 PM11/26/17
to Desobediente, Tobias Killer, blacklight, qubes-translation
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
On 2017-11-26 14:30, Desobediente wrote:
> Almost forgot, I know I don't have write permission because every
> reply here I get a "Mail Delivery Subsystem -
> mail-...@google.com" standard message:
>
> Hello aindate...@gmail.com,
>
> We're writing to let you know that the group you tried to contact
> (qubes-translation) may not exist, or you may not have permission
> to post messages to the group. A few more details on why you
> weren't able to post:
>
> * You might have spelled or formatted the group name incorrectly.
> * The owner of the group may have removed this group. * You may
> need to join the group before receiving permission to post. * This
> group may not be open to posting.
>
> If you have questions related to this or any other Google Group,
> visit the Help Center at https://groups.google.com/support/.
>
> Thanks,
>
> Google Groups
>
Ah, I see now. The group was set to allow posts only from members. (I
had thought it was set to allow public posts, but it was actually only
set to allow public views and joins.) I've changed the settings so
that the list will behave like qubes-users. You should now be able to
post without joining.
- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----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=slyg
-----END PGP SIGNATURE-----
Hash: SHA512
On 2017-11-26 14:30, Desobediente wrote:
> Almost forgot, I know I don't have write permission because every
> reply here I get a "Mail Delivery Subsystem -
> mail-...@google.com" standard message:
>
> Hello aindate...@gmail.com,
>
> We're writing to let you know that the group you tried to contact
> (qubes-translation) may not exist, or you may not have permission
> to post messages to the group. A few more details on why you
> weren't able to post:
>
> * You might have spelled or formatted the group name incorrectly.
> * The owner of the group may have removed this group. * You may
> need to join the group before receiving permission to post. * This
> group may not be open to posting.
>
> If you have questions related to this or any other Google Group,
> visit the Help Center at https://groups.google.com/support/.
>
> Thanks,
>
> Google Groups
>
Ah, I see now. The group was set to allow posts only from members. (I
had thought it was set to allow public posts, but it was actually only
set to allow public views and joins.) I've changed the settings so
that the list will behave like qubes-users. You should now be able to
post without joining.
- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----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=slyg
-----END PGP SIGNATURE-----
Blacklight447
Nov 26, 2017, 3:53:55 PM11/26/17
to a...@qubes-os.org, aindate...@gmail.com, tok...@posteo.de, qubes-tr...@googlegroups.com
This too, was the reason i temponarly did not cc the group. I thought I was the only obe encountering the issue.
Sent from ProtonMail mobile
-------- Original Message --------
On Nov 26, 2017, 9:47 PM, Andrew David Wong < a...@qubes-os.org> wrote:
Sent from ProtonMail mobile
-------- Original Message --------
On Nov 26, 2017, 9:47 PM, Andrew David Wong < a...@qubes-os.org> wrote:
On 2017-11-26 14:30, Desobediente wrote:
> Almost forgot, I know I don't have write permission because every
> reply here I get a "Mail Delivery Subsystem -
> mail-...@google.com" standard message:
>
> Hello aindate...@gmail.com,
>
> We're writing to let you know that the group you tried to contact
> (qubes-translation) may not exist, or you may not have permission
> to post messages to the group. A few more details on why you
> weren't able to post:
>
> * You might have spelled or formatted the group name incorrectly.
> * The owner of the group may have removed this group. * You may
> need to join the group before receiving permission to post. * This
> group may not be open to posting.
>
> If you have questions related to this or any other Google Group,
> visit the Help Center at https://groups.google.com/support/.
>
> Thanks,
>
> Google Groups
>
Ah, I see now. The group was set to allow posts only from members. (I
had thought it was set to allow public posts, but it was actually only
set to allow public views and joins.) I've changed the settings so
that the list will behave like qubes-users. You should now be able to
post without joining.
Tobias Killer
Dec 9, 2017, 6:22:13 AM12/9/17
to Desobediente, Andrew David Wong, blacklight, qubes-translation
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256
Am 26.11.2017 um 21:27 schrieb Desobediente:
Hash: SHA256
> With regard to how people would feel about not having their
> translations accepted, I've saw it happening because a friend of
> mine sent work to the portuguese tails translation mailing list and
> so far nor me nor anyone else accepted it. She's an understanding
> person to a point, but eventually she won't contribute anymore
> because of those problems. So for my personal experience I can see
> how would it impact people, and I fully agree we should coordinate
> things to avoid those problems.
Fine!
BTW: Recently, I explained the problem (that probably a big amount of
the translators' work will probably end up in the trash) to other
non-involved, non-tech people. And they say that it's even better to
tell the truth right in the face than the case that the translators
spot it by themselves. They could feel deceived.
> As for everything I wrote in english, you got it all right Tobias.
> We seem to have an understanding even though we have to use this
> horrible intermediary language.
Yes. :-)
> Regarding the point on step 1 and the main concern here, I endorse
> Andrew's contributions, and I insist we should lock anything as a
> last resort, do it carefully and dedicate to unlock it as soon as
> possible.
I don't know the email (or whatever) you refer to. Could you please
cite Andrew?
The only email I found where Andrew talked about step 1 is this one
(excerpt):
Am 26.11.2017 um 01:51 schrieb Andrew David Wong:
> IMO, it sounds like the main consideration regarding Step 1 is how
> much time it would take to do the rest of the steps. If it'll only
> take a week, for example, then closing access probably isn't worth
> the trouble. But if it'll take more than a month, it's probably
> worth it.
"as a last resort", as you called it. As far as I understand Andrew,
it's just about if it's worth it.
And yes, I think it's worth it. The problem was first mentioned by
Alain-Olivier (the French Language Coordinator) back in October 2017.
So, it's almost two months ago, now. Also note that there're still
many difficult decisions to make, difficult scripts to write and much
stuff to test until a working bi-directional process between Qubes and
Transifex can run.
@all:
I have a new suggestion. My considerations for this are:
- - As far as I can see, now as Transifex project maintainers, we're
able to lock single resources, rather than all resources at once.
(Note that Alain-Olivier has suggested to lock at least resources that
change heavily.)
- - Also, we're able to post project announcements now.
So, my new idea now is not to lock single heavily-changing resources
(since this (1) would be difficult to find out and (2) doesn't solve
the problem of the lack of translation guidelines which are necessary
even for seldom-changing resources), but to post an announcement where
we thank the translators and the language coordinators for their hard
work, and that the coordinators may move the translators to other
projects temporarily until solutions to some internal issues could
have been found. We could also send this to the language coordinators
in private but I think it's better to go the public way.
So, what do you think about this?
@Blacklight:
AFAIK on 30.10.2017 you wrote some emails to qubes-translation and
addressed Alain in the email body but not in the "To" field. So, I'm
not sure if Alain got them at all. Could you just look it up? (Thank
you for your work back then and today!)
@Desobediente @Blacklight:
I didn't ask Nina (the Transifex supporter we asked in the past) yet.
Could one of you please describe the problem and ask her for help and
suggestions? Please don't forget to CC to the qubes-translation
mailing list.
Kind regards,
Tobias Killer
-----BEGIN PGP SIGNATURE-----
iQIzBAEBCAAdFiEE9VOPt2yY9xS+XKzULaXvb25AsygFAlorxykACgkQLaXvb25A
-----BEGIN PGP SIGNATURE-----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=F/zJ
-----END PGP SIGNATURE-----
Andrew David Wong
Dec 9, 2017, 2:39:42 PM12/9/17
to Tobias Killer, Desobediente, blacklight, qubes-translation
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
On 2017-12-09 05:21, Tobias Killer wrote:
> [...] @all: I have a new suggestion. My considerations for this
1. Stop all translation.
2. Make sure translators know why they can't continue translating and
that it's temporary.
3. Do the other stuff blocking translation.
4. Re-open translation.
5. Make sure translators know they can start translating again.
- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----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=VZAc
-----END PGP SIGNATURE-----
Hash: SHA512
On 2017-12-09 05:21, Tobias Killer wrote:
> [...] @all: I have a new suggestion. My considerations for this
> are:
>
> - As far as I can see, now as Transifex project maintainers, we're
> able to lock single resources, rather than all resources at once.
> (Note that Alain-Olivier has suggested to lock at least resources
> that change heavily.) - Also, we're able to post project
>
> - As far as I can see, now as Transifex project maintainers, we're
> able to lock single resources, rather than all resources at once.
> (Note that Alain-Olivier has suggested to lock at least resources
> announcements now.
>
> So, my new idea now is not to lock single heavily-changing
> resources (since this (1) would be difficult to find out and (2)
> doesn't solve the problem of the lack of translation guidelines
> which are necessary even for seldom-changing resources), but to
> post an announcement where we thank the translators and the
> language coordinators for their hard work, and that the
> coordinators may move the translators to other projects
> temporarily until solutions to some internal issues could have been
> found. We could also send this to the language coordinators in
> private but I think it's better to go the public way.
>
> So, what do you think about this?
>
No need to make things so complicated, IMO. Just:
>
> So, my new idea now is not to lock single heavily-changing
> resources (since this (1) would be difficult to find out and (2)
> doesn't solve the problem of the lack of translation guidelines
> which are necessary even for seldom-changing resources), but to
> post an announcement where we thank the translators and the
> language coordinators for their hard work, and that the
> coordinators may move the translators to other projects
> temporarily until solutions to some internal issues could have been
> found. We could also send this to the language coordinators in
> private but I think it's better to go the public way.
>
> So, what do you think about this?
>
1. Stop all translation.
2. Make sure translators know why they can't continue translating and
that it's temporary.
3. Do the other stuff blocking translation.
4. Re-open translation.
5. Make sure translators know they can start translating again.
- --
Andrew David Wong (Axon)
Community Manager, Qubes OS
https://www.qubes-os.org
-----BEGIN PGP SIGNATURE-----
EQoAnr/ppoUBN2FExAWb61vXeOfE2KIw0s5xncyqJvYTaOEiws27NKC0FoW9K8za
LBdMstfNyfsSvxUBtWcbpVj6YGV/AfNTpbb/FSgjxYEu4kGZ3a+2/Al6QhWaGlBJ
Mk7Vgu36zdzuNmOKVESPGntd/dhacMLUmSrX9HD677eNljm2Lzxfxk0obu61Y1qw
6nR8e41fF89A5HrWnDmbUwYC9d3ymx/L10Jdk2pqnPbl3tEKO+/gMESRJwg5sTyD
ihGdZpoZLICvZCuNhJdmIeeKH60opf81Yv6uWuep+WgMRCoEArxccJXkECUzYDbY
tUbf5xMgQ2tpP3TCAORC0ItaJ5naz4IJzuZd3/njiLNWcYWGo4MM0BB3kaNiKBkS
9ghiqC2sHe4QurBYVhNw51PnHBzvf5O875qu/AHoUPsODxgtufyg8Qysdr4oWeow
Fcy1S7ERgjuXdwXXaU1/ulLpsD8DTStWDZifYQdagCKW1BHceHPJqr2x746ltyiE
Fr/l44agHkWQD0mFYeystCCRS0Dmgd9g3SOJdRWTGIk/vmlmj47U678HWuROzNk3
li65Dry8vkaKdRkcLwHbOxGx+j462YPZvKPeBcQGbEERwGjTdw5tgEwUs/VFomYH
Q9NNKc2JSOA2PdVBm3bl
=VZAc
-----END PGP SIGNATURE-----
tokidev
Dec 10, 2017, 7:52:24 AM12/10/17
to Andrew David Wong, Desobediente, blacklight, qubes-translation
Am 09.12.2017 um 20:39 schrieb Andrew David Wong:
> On 2017-12-09 05:21, Tobias Killer wrote:
>> [...] @all: I have a new suggestion. My considerations for this
>> are:
>
>> - As far as I can see, now as Transifex project maintainers, we're
>> able to lock single resources, rather than all resources at once.
>> (Note that Alain-Olivier has suggested to lock at least resources
>> that change heavily.) - Also, we're able to post project
>> announcements now.
>
>> So, my new idea now is not to lock single heavily-changing
>> resources (since this (1) would be difficult to find out and (2)
>> doesn't solve the problem of the lack of translation guidelines
>> which are necessary even for seldom-changing resources), but to
>> post an announcement where we thank the translators and the
>> language coordinators for their hard work, and that the
>> coordinators may move the translators to other projects
>> temporarily until solutions to some internal issues could have been
>> found. We could also send this to the language coordinators in
>> private but I think it's better to go the public way.
>
>> So, what do you think about this?
>
>
> No need to make things so complicated, IMO.
I tried to find a compromise with Desci. But I think you're right here. :)
> On 2017-12-09 05:21, Tobias Killer wrote:
>> [...] @all: I have a new suggestion. My considerations for this
>> are:
>
>> - As far as I can see, now as Transifex project maintainers, we're
>> able to lock single resources, rather than all resources at once.
>> (Note that Alain-Olivier has suggested to lock at least resources
>> that change heavily.) - Also, we're able to post project
>> announcements now.
>
>> So, my new idea now is not to lock single heavily-changing
>> resources (since this (1) would be difficult to find out and (2)
>> doesn't solve the problem of the lack of translation guidelines
>> which are necessary even for seldom-changing resources), but to
>> post an announcement where we thank the translators and the
>> language coordinators for their hard work, and that the
>> coordinators may move the translators to other projects
>> temporarily until solutions to some internal issues could have been
>> found. We could also send this to the language coordinators in
>> private but I think it's better to go the public way.
>
>> So, what do you think about this?
>
>
> No need to make things so complicated, IMO.
Andrew David Wong
Dec 10, 2017, 4:09:20 PM12/10/17
to Blacklight447, Desobiente, tokidev, qubes-translation
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
On 2017-12-10 15:00, Blacklight447 wrote:
Hash: SHA512
> What would be the most appropriate way to let the translators
> know, maybe we can send them a message via transifex? We could also
> send out an announcement on the qubes-translation list, or both.
>
Send a message or announcement via Transifex that they can't miss. We
shouldn't assume that every Transifex translator monitors the
qubes-translation ML. Both is fine.
P.S. - Please keep the list CCed and avoid top-posting.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=xSg6
-----END PGP SIGNATURE-----
Tobias Killer
Dec 12, 2017, 8:44:53 AM12/12/17
to Andrew David Wong, Blacklight447, Desobiente, qubes-translation
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256
Hello @all,
Am 10.12.2017 um 22:09 schrieb Andrew David Wong:
> On 2017-12-10 15:00, Blacklight447 wrote:
>> What would be the most appropriate way to let the translators
>> know, maybe we can send them a message via transifex? We could
>> also send out an announcement on the qubes-translation list, or
>> both.
>
>
> Send a message or announcement via Transifex that they can't miss.
> We shouldn't assume that every Transifex translator monitors the
> qubes-translation ML. Both is fine.
I agree.
I've just written a proposal for such an announcement:
"
Dear translators,
Dear coordinators,
Thank you for your hard work on translating Qubes OS!
Due to open internal discussions of how to solve several translation
issues and due to the lack of integration of the translations on our
website, we've decided to lock access for translations temporarily.
This way, we all will be spared of waste of effort. Once these
problems are solved, we'll re-open access to the resources again. The
process of solving might take some months.
If there are concerns, you may contact as via email to
<qubes-tr...@googlegroups.com>.
Kind regards,
The Qubes OS Translation Team
"
What do you think about it?
Regards,
Tobias
-----BEGIN PGP SIGNATURE-----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=Uk5A
-----END PGP SIGNATURE-----
Hash: SHA256
Hello @all,
Am 10.12.2017 um 22:09 schrieb Andrew David Wong:
> On 2017-12-10 15:00, Blacklight447 wrote:
>> What would be the most appropriate way to let the translators
>> know, maybe we can send them a message via transifex? We could
>> also send out an announcement on the qubes-translation list, or
>> both.
>
>
> Send a message or announcement via Transifex that they can't miss.
> We shouldn't assume that every Transifex translator monitors the
> qubes-translation ML. Both is fine.
I've just written a proposal for such an announcement:
"
Dear translators,
Dear coordinators,
Thank you for your hard work on translating Qubes OS!
Due to open internal discussions of how to solve several translation
issues and due to the lack of integration of the translations on our
website, we've decided to lock access for translations temporarily.
This way, we all will be spared of waste of effort. Once these
problems are solved, we'll re-open access to the resources again. The
process of solving might take some months.
If there are concerns, you may contact as via email to
<qubes-tr...@googlegroups.com>.
Kind regards,
The Qubes OS Translation Team
"
What do you think about it?
Regards,
Tobias
-----BEGIN PGP SIGNATURE-----
iQIzBAEBCAAdFiEE9VOPt2yY9xS+XKzULaXvb25AsygFAlov3UMACgkQLaXvb25A
syggWQ//fnhQiyWBh9/hyi8BCRS5NVWYF/DhSorlbnwCL/1rUFhuoV2iN//AY9QH
FUxOpqNvxjvghfvrZRw2+8wcL6PpMxifmcg5CCmlfPvAFzgAurB9jCxp774TlN3c
C4hRVqWmfckTV02WMSpLLq8nY5eoWNLjvEQlsBQ9982zbPq46DUSUrEyBgae0sVK
mgMZo+l0rjuQjBfmihbDIX1zd3fCZCAskyq4OYOy/NT+vPaBOI0vKLEHzqo9eOQt
L0OEHykCONa4eB9Yz9nJKWmwORaOq+f3RhtEJBNv70Qrv6MUiE4ZK8hUSh1ZtZdU
6omPNUPWaxIOEdK/nVgycHoEcb86wbd2akcFg32dXc3g+w9lYffMT/f694zg09WV
qU55fEXB7XcNKbHWxBp/czIOM4dnfx+weHo2WGsxX0Bg2XercHeGNoFWtALfzKde
0oqA6nu0FFSFvZwQ4/+DsFiOYkzcYDPgxSrakmBcopqG4EoJ4rYghEmLQqBBuKgq
H9bOUoSbR4tIFDHYCP5qg4aIHtujOFgkh4SxrirWzWsyPynkDjAKaUNHWgKnc5ZM
DxKQ1UDKITALgSIsVgJz9CfYxo7AgWX1Bu5ARPJrrVMH5p6q0KUsivlWO3VZ+JMF
IoB3wn8AVGdgA5jPsU1r+4w2KntjqOm7pERy7HIp+a34LdueVrM=
=Uk5A
-----END PGP SIGNATURE-----
yaho...@gmail.com
Dec 15, 2017, 4:49:38 PM12/15/17
to qubes-translation
Hello to all,
Sorry for not answering any sooner, but I wasn’t aware of this post being here.
From what I gather from reading this long post, you decided to temporary lock access to the Transifex Qubes project. We can do that for you by archiving it, and we can also take care of making the announcement to the Qubes language teams.
I’ll contact Andrew David directly for him to confirm that this is what you want.
Best regards,
Alain-Olivier
French language coordinator
localizationlab.org | transifex.com/organization/otf
5C8F 0C0C 4812 7771 365C 262D 47DC 6187 47A9 22D2
Sorry for not answering any sooner, but I wasn’t aware of this post being here.
From what I gather from reading this long post, you decided to temporary lock access to the Transifex Qubes project. We can do that for you by archiving it, and we can also take care of making the announcement to the Qubes language teams.
I’ll contact Andrew David directly for him to confirm that this is what you want.
Best regards,
Alain-Olivier
French language coordinator
localizationlab.org | transifex.com/organization/otf
5C8F 0C0C 4812 7771 365C 262D 47DC 6187 47A9 22D2
Reply all
Reply to author
Forward
0 new messages