[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Use of meaningful (and probably non-verbal) hyperlinks



-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi,

I've been told (thank you Andrei), that the subject I just raised [0]
should also be announced on debian-www, debian-doc and
debian-l10n-english, so let me try to fix the initial mess. My thought
was focused on the DPN, thus the choice of debian-publicity, but maybe
debian-l10n-english is a better place to follow up this thread.

[0] hu37d4$ule$1@dough.gmane.org">http://lists.debian.org/hu37d4$ule$1@dough.gmane.org

Cheers

David

Le 01/06/2010 14:01, Justin B Rye a écrit :
> David Prévot wrote:
>> I raised an issue last week-end, while the DPN was already freezed, and
>> got only one person on my "side", so I reverted my changes [287]. In
>> fact, I think everybody focused on the "verb" part, and my poorly chosen
>> example from W3C,
>
> I blame the W3C tip-writer.
>
>> so I'll try to make myself clearer in order to
>> eventually evolve in another (hopefully better) way to link stuff from
>> the web-based DPN (I think that the mail edition is better since there
>> are no true hyperlinks).
>>
>> [287]http://svn.debian.org/viewsvn/publicity/dpn/en/2010/04/index.wml?r1=285&r2=287
>>
>> For example (I picked a short and easy one), in the just published DPN,
>> we have :
>>
>> "Aurélien Jarno *reported* a bit about Debian's switch from GLIBC to
>> EGLIBC."
>>
>> I think that the hyperlink on *reported* is meaningless, and misguiding
>> (why should I follow a link just called "reported", am I going to be
>> reported somewhere?).
>
> The meaning I tend to assume is being conveyed is that such a link
> will lead to a location where the communicative event in question is
> archived.  Links on non-informational verbs, such as "Aurélien Jarno
> *lives* nearby", would be much less natural.
>
>> The purpose of Aurélien's report is about the *Debian's switch from
>> GLIBC to EGLIBC*, or at least the *switch from GLIBC to EGLIBC*, so I
>> really believe that the hyperlink should rather be on this phrase.
>
> That would make sense, though there's a slight semantic gap to be
> bridged either way, since his report did not itself constitute a
> "switch from GLIBC to EGLIBC".  Other kinds of post, such as for
> instance "Aurélien Jarno asked whether GLIBC should be removed
> from Squeeze", would be trickier.  It's also quite common for
> secondary links to be contained within the material quoted.
>
> Incidentally, I'm assuming there's an unspoken rule that where
> possible we prefer short link anchors, so that the page is a sea of
> normal text containing small islands of underlined blue clickability
> rather than the reverse.  But none of those tips pages say anything
> about this, so maybe it's just me.
>
>> Eventually, in some not so easy cases, I think that linking on the name
>> of the "reporter" (here *Aurélien Jarno*), should be far more meaningful
>> that the usual past tense verb "reported" (usually "announced"), since
>> the hyperlink will link to something about this person.
>
> What you're implying is that in a certain web-centric sense, that
> URL _is_ "Aurélien Jarno".  This works well enough for blogs, but
> less well for mailinglists and so on.
>
>> I'm well aware that the meaning of the total sentence is the best way to
>> have a clue about what the hyperlink is about, but I sincerely believe
>> that the hyperlink itself should better be meaningful if possible.
>
> Absolutely.  And probably I should be asking around to see what
> random strangers think instead of arguing for my own prejudices.
>
>> P.-S.: Alexander proposed to eventually raised this sort of discussion
>> to other more general lists, I think it could be fine, but why not
>> starting around here, and if we have a consensus, try to raised the
>> discussion to debian-www for example. Some people answered to the
>> previous thread that was crossed-linked with debian-l10n-english, maybe
>> we could do it again for this one in order to gather more opinion too.
>
> (I've already been sucked in from d-l-e.)

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iEYEARECAAYFAkwFvy8ACgkQ18/WetbTC/qWbACfZsJ9J3NS+ScVSo8rORJPx476
Cb8An1y7Wn1ZHqZKfWNNg+5W2sDR9kb3
=6TQz
-----END PGP SIGNATURE-----


Reply to: