Re: [bug-GNUnet] GNUnet PO file bugs

[Christian Grothoff]

Dear Clytie,

First of all, I don't know what the problem is/was, address@hidden 
is the correct e-mail address.  

I've made all of your suggested changes to SVN HEAD (and they will be in the 
next release), except for the (various variants of)

msgid "# bytes received via HTTP"

I fixed a few other places where transport names were not %s-ed, and the 0.8.0 
tree had already a few improvements in that area.  However, this very 
particular instance cannot be fixed -- the string is sent over the network 
untranslated and the receiver then does the translation to the current 
locale.  So we need to embedd the transport type in the specific string -- 
and format strings are not supported (and will not for various technical 
reasons).  In good news, TCP6/UDP6 are gone (integrated into TCP/UDP) anyway, 
so this impacts only 12 short strings overall at this point.

All of the other ones that you mentioned below (and a few I found by looking 
manually) have been %s-ed.

Thanks for your feedback & the translations!!!

Christian


On Saturday 03 May 2008 02:25:34 am Clytie Siddall wrote:
> Hello :)
>
> I have just translated the file gnunet-0.7.3 for the Translation
> Project. In doing so, I've had this email first bounced by your bug
> address, as quoted twice in the file, encountered one fatal error,
> occurring in four strings, which needs fixing ASAP, one efficiency
> issue for which I have made a suggestion, and some typos which I
> thought you might like to weed out in a future release.
>
> *** <address@hidden> bounces.
>
> "Report-Msgid-Bugs-To: address@hidden"
>
> #: src/server/startup.c:233
> #, c-format
> msgid ""
> "Filesystem `%s' of partition `%s' is unknown. Please contact gnunet-"
> "address@hidden"
>
> Please fix this.
> ____
>
> A. Fatal error: the translated PO file will not pass gettext msgfmt
> check
>
> #: src/applications/stats/statistics.c:228
> #, c-format
> msgid "% of allowed network load (up)"
> msgstr "% trọng tải mạng được phép (chạy)"
>
> #: src/applications/stats/statistics.c:230
> #, c-format
> msgid "% of allowed network load (down)"
> msgstr "% trọng tải mạng được phép (không chạy)"
>
> #: src/applications/stats/statistics.c:231
> #, c-format
> msgid "% of allowed cpu load"
> msgstr "% trọng tải CPU được phép"
>
> #: src/applications/stats/statistics.c:232
> #, c-format
> msgid "% of allowed io load"
> msgstr "% trọng tải V/R được phép"
>
> Naked percentage symbols in c-format strings are invalid in the
> translation. You can get around this by escaping the percentage symbol
> (%%), by stating no-c-format, or by using the word "percentage" instead.
>
> Pearl:TP clytie$ msgfmt -cv -o /dev/null gnunet-0.7.3/gnunet-0.7.3.vi.po
> gnunet-0.7.3/gnunet-0.7.3.vi.po:3168: 'msgstr' is not a valid C format
> string, unlike 'msgid'. Reason: In the directive number 1, the
> character 'r' is not a valid conversion specifier.
> gnunet-0.7.3/gnunet-0.7.3.vi.po:3173: 'msgstr' is not a valid C format
> string, unlike 'msgid'. Reason: In the directive number 1, the
> character 'r' is not a valid conversion specifier.
> gnunet-0.7.3/gnunet-0.7.3.vi.po:3178: 'msgstr' is not a valid C format
> string, unlike 'msgid'. Reason: In the directive number 1, the
> character 'r' is not a valid conversion specifier.
> gnunet-0.7.3/gnunet-0.7.3.vi.po:3183: 'msgstr' is not a valid C format
> string, unlike 'msgid'. Reason: In the directive number 1, the
> character 'r' is not a valid conversion specifier.
> msgfmt: found 4 fatal errors
> 1062 translated messages.
> ________________________
>
> B. Efficiency
>
> A suggestion
>
> There are a LOT of repeated strings where the only difference is the
> protocol name. To reduce the size of the file, and thus attract more
> translator time (and spend it more effectively), you could substitute
> a placeholder for the protocol. For example:
>
> #: src/transports/tcp6.c:531
> msgid "# bytes received via TCP6"
>
> #: src/transports/udp6.c:500
> msgid "# bytes received via UDP6"
>
> #: src/transports/http.c:2300
> msgid "# bytes received via HTTP"
>
> #: src/transports/smtp.c:831
> msgid "# bytes received via SMTP"
>
> #: src/transports/tcp.c:594
> msgid "# bytes received via TCP"
>
> #: src/transports/udp.c:560
> msgid "# bytes received via UDP"
>
> all become
>
> msgid: # bytes received via %s"
>
>
> C. Typos
>
> 1.
>
> #: src/setup/ncurses/wizard_curs.c:342
> msgid ""
> "You can limit GNUnet's resource usage here.\n"
> "\n"
> "The \"upstream\" is the data channel through which data is *sent* to
> the "
> "internet. The limit is the maximum amountwhich GNUnet is allowed to
> use. If "
> "you have a flatrate, you can set it to the maximum speed of your
> internet "
> "connection. You should not use a value that is higher than what your
> actual "
> "connection allows."
>
> - amountwhich
> + amount which
>
>
> 2.
>
> #: src/setup/ncurses/wizard_curs.c:406
> msgid ""
> "For security reasons, it is a good idea to let this setup create a
> new group "
> "for the chosen user account.\n"
> "\n"
> "You can also specify a already existant group here.\n"
> "\n"
> "Only members of this group will be allowed to start and stop the the
> GNUnet "
> "server and have access to GNUnet server data.\n"
>
> - existant
> + existent
>
>
> 3.
>
> #: src/setup/lib/wizard_util.c:127
> msgid "Cannot write to the regisitry"
>
> - regisitry
> + registry
>
>
> 4.
>
> #: src/setup/gnunet-setup.c:74
> #, c-format
> msgid ""
> "Invalid synatx, argument to 'set' must have the format SECTION:"
> "OPTION=VALUE.\n"
>
> #: src/setup/gnunet-setup.c:95
> #, c-format
> msgid ""
> "Invalid synatx, argument to 'get' must have the format SECTION:OPTION.
> \n"
>
>
> - synatx
> + syntax
>
>
> 5.
>
> #: src/applications/gap/gap.c:2261
> msgid "# gap re-try ttl difference (cummulative)"
>
> - cummulative
> + cumulative
>
>
> 6.
>
> #: src/applications/chat/gnunet-chat.c:49
> msgid "set the nickname to use (requred)"
>
> #: src/applications/chat/gnunet-chat.c:52
> msgid "set the chat room to join (requred)"
>
>
> - requred
> + required
>
>
> 7.
>
> #: src/server/connection.c:3660
> msgid "# total number of bytes we were allowed to sent but did not"
>
> #: src/server/connection.c:3664
> msgid "# total number of bytes we were allowed to sent"
>
>
> - allowed to sent
> + allowed to send
>
>
> 8.
>
> #: src/server/handler.c:437
> #, c-format
> msgid "Received corrupt message from peer `%s'in %s:%d.\n"
>
> - `%s'in
> + `%s' in
>
>
> 9.
>
> #: src/util/threads/mutex.c:187
> #, c-format
> msgid "Lock aquired for too long (%llu ms) at %s:%u\n"
>
> - aquired
> + acquired
>
>
> 10.
>
> #: contrib/config-daemon.scm:166
> msgid ""
> "How many minutes is the current IP valid?  (GNUnet will sign HELLO
> messages "
> "with this expiration timeline. If you are on dialup, 60 (for 1 hour)
> is "
> "suggested. If you are having a static IP address, you may want to set
> this "
> "to a large value (say 14400).  The default is 1440 (1 day). If your
> IP "
> "changes periodically, you will want to choose the expiration to be
> smaller "
> "than the frequency with which your IP changes."
>
> - If you are having a static IP address
> + If you have a static IP address
>
> - you will want to choose the expiration to be smaller
> + you will want to choose an expiry period smaller
>
>
> 11.
>
> #: contrib/config-daemon.scm:350
> msgid ""
> "Use space-separated list of the modules, e.g.  \"udp smtp tcp\".  The "
> "available transports are udp, tcp, http, smtp, tcp6, udp6 and nat.\n"
> "\t\t\n"
> "Loading the 'nat' and 'tcp' modules is required for peers behind NAT
> boxes "
> "that cannot directly be reached from the outside.  Peers that are NOT
> behind "
> "a NAT box and that want to *allow* peers that ARE behind a NAT box to "
> "connect must ALSO load the 'nat' module.  Note that the actual
> transfer will "
> "always be via tcp initiated by the peer behind the NAT box.  The nat "
> "transport requires the use of tcp, http, smtp and/or tcp6 in addition
> to nat "
> "itself."
>
> - Use space-separated list of the modules
> + Use a space-separated list of modules
>
>
> 12.
>
> #: contrib/config-daemon.scm:396
> msgid ""
> "This option can be used to tell gnunetd not to open the client port.
> When "
> "run like this, gnunetd will participate as a peer in the network but
> not "
> "support any user interfaces.  This maybe useful for headless systems
> that "
> "are never expected to have end-user interactions.  Note that this
> will also "
> "prevent you from running diagnostic tools like gnunet-stats!"
>
> #: contrib/config-daemon.scm:445
> msgid ""
> "This option maybe useful during testing, but turning it off is
> dangerous! If "
> "in any doubt, set it to YES (which is the default)."
>
> #: contrib/config-client.scm:323
> msgid ""
> "This option is useful to eliminate files that the user already has
> from the "
> "search.  Naturally, enabling this option maybe confusing because some "
> "obviously expected search results would no longer show up.  This
> option only "
> "works if the URI_DB_SIZE option under FS is not zero (since the URI
> DB is "
> "used to determine which files the user is sharing)"
>
> - maybe
> + may be
>
>
> 13.
>
> #: contrib/config-daemon.scm:647
> msgid ""
> "Minimum number of connected friends before this peer is allowed to
> connect "
> "to peers that are not listed as friends"
>
> - Minimum number of
> + The minimum number of
>
>
> 14. Do we need both these strings?
>
> #: contrib/config-daemon.scm:876
> msgid ""
> "You can use 'make check' in src/transports/upnp/ to find out if your
> NAT "
> "supports UPnP.  You should disable this option if you are sure that
> you are "
> "not behind a NAT.  If your NAT box does not support UPnP, having this
> on "
> "will not do much harm (only cost a small amount of resources)."
>
> #: contrib/config-daemon.scm:942
> #: contrib/config-daemon.scm:1092
> msgid ""
> "You can use 'make check' in src/transports/upnp/ to find out if your
> NAT "
> "supports UPnP. You should disable this option if you are sure that
> you are "
> "not behind a NAT.  If your NAT box does not support UPnP, having this
> on "
> "will not do much harm (only cost a small amount of resources)."
>
> The only difference seems to be a double space between the first and
> second sentences in the first string, where there's a single space in
> the second string.
>
>
> 15.
>
> #: contrib/config-daemon.scm:954
> msgid ""
> "Use this option if your firewall maps say port 80 to your real HTTP
> port.  "
> "This can be useful in making the HTTP messages appear even more legit "
> "(without the need for running gnunetd as root due to the use of a "
> "priviledged port)."
>
> - if your firewall maps say port 80
> + if your firewall maps, say, port 80
> OR
> + if your firewall maps (e.g.) port 80
>
> - without the need for running gnunetd as root
> + without needing to run gnunetd as root
>
> - priviledged
> + privileged
>
>
> 16.
>
> #: contrib/config-daemon.scm:1363
> msgid ""
> "Which CPU load can be tolerated.  Load here always refers to the
> total "
> "system load, that is it includes CPU utilization by other processes.
> A "
> "value of 50 means that once your 1 minute-load average goes over 50%
> non-"
> "idle, GNUnet will try to reduce CPU consumption until the load goes
> under "
> "the threshold.  Reasonable values are typically between 50 and 100.  "
> "Multiprocessors may use values above 100."
>
> #: contrib/config-daemon.scm:1376
> msgid ""
> "Which IO load can be tolerated.  Load here refers to the percentage
> of CPU "
> "cycles wasted waiting for IO for the entire system, that is it
> includes disk "
> "utilization by other processes.  A value of 10 means that once the
> average "
> "number of cycles wasted waiting for IO is more than 10% non-idle,
> GNUnet "
> "will try to reduce IO until the load goes under the threshold.
> Reasonable "
> "values are typically between 10 and 75."
>
> #: contrib/config-daemon.scm:1388
> msgid ""
> "Which CPU load can be tolerated.  This is the hard limit, so once it
> is "
> "reached, gnunetd will start to massively drop data to reduce the
> load.  Use "
> "with caution."
>
>
> - Which CPU load can be tolerated.
> + The highest tolerable CPU load.
>
> - Which IO load can be tolerated
> + The highest tolerable IO load.
>
>
> 17.
>
> #: contrib/config-daemon.scm:1425
> msgid ""
> "Basic bandwidth limitation (YES) means simply that the bandwidth
> limits "
> "specified apply to GNUnet and only to GNUnet. ..[snip]...\n"
> "\n"
> "If either of these two scenarios applies, set BASICLIMITING to NO.
> Then set "
> "the bandwidth limits to the COMBINED amount of traffic that is
> accepable for "
> "both GNUnet and other applications.  GNUnet will then immediately
> throttle "
> "bandwidth consumption if the short-term average is above the limit,
> and it "
> "will also try to ensure that the long-term average is below the
> limit.  Note "
> "however that using NO can have the effect of GNUnet (almost) ceasing "
> "operations after other applications perform high-volume downloads
> that are "
> "beyond the defined limits.  GNUnet would reduce consumption until the
> long-"
> "term limits are again within bounds.\n"
> "\n"
> "NO ...[snip]... irrelevant)."
>
> - accepable
> + acceptable
>
> ___
>
> I hope this is useful. Good luck with everything. :)
>
> Clytie Siddall
> Vietnamese Free-Software Translation Team
> http://vnoss.net/dokuwiki/doku.php?id=projects:l10n