- Aur-dev - lists.archlinux.org

[aur-dev] [PATCH v2] Fixed Null issue in git-interface/git-serve.py
by Mark Weiman 26 Nov '15

26 Nov '15

Fixes a regression introduced in 8c87b1d (git-serve: Add support for setting keywords, 2015-10-22). Signed-off-by: Mark Weiman <mark.weiman(a)markzz.com> --- git-interface/git-serve.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/git-interface/git-serve.py b/git-interface/git-serve.py index 8bd6aa8..62410e9 100755 --- a/git-interface/git-serve.py +++ b/git-interface/git-serve.py @@ -37,7 +37,7 @@ def pkgbase_from_name(pkgbase): return row[0] if row else None def pkgbase_exists(pkgbase): - return (pkgbase_from_name(pkgbase) > 0) + return pkgbase_from_name(pkgbase) is not None def list_repos(user): db = mysql.connector.connect(host=aur_db_host, user=aur_db_user, -- 2.6.2

2 1

[aur-dev] [PATCH] Fixed Null issue in git-interface/git-serve.py
by Mark Weiman 26 Nov '15

26 Nov '15

Found while setting up a brand new instance of aurweb, in git-serve.py, if the PackageBases table is empty, it generates a TypeError. "TypeError: unorderable types: NoneType() > int()" Signed-off-by: Mark Weiman <mark.weiman(a)markzz.com> --- git-interface/git-serve.py | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/git-interface/git-serve.py b/git-interface/git-serve.py index 8bd6aa8..6ba19e8 100755 --- a/git-interface/git-serve.py +++ b/git-interface/git-serve.py @@ -37,7 +37,10 @@ def pkgbase_from_name(pkgbase): return row[0] if row else None def pkgbase_exists(pkgbase): - return (pkgbase_from_name(pkgbase) > 0) + if pkgbase_from_name(pkgbase): + return (pkgbase_from_name(pkgbase) > 0) + else: + return False def list_repos(user): db = mysql.connector.connect(host=aur_db_host, user=aur_db_user, -- 2.6.2

2 1

[aur-dev] [PATCH] Change "File Request" to be more explicit.
by Eli Schwartz 26 Nov '15

26 Nov '15

Apparently it confuses some people, and there appears to be enough space in the Package Actions section anyway. Fixes: https://lists.archlinux.org/pipermail/aur-general/2015-November/031776.html Because a 37-post flame war is fun. :rolleyes: Signed-off-by: Eli Schwartz <eschwartz93(a)gmail.com> --- web/html/pkgreq.php | 2 +- web/template/pkgbase_actions.php | 2 +- web/template/pkgreq_form.php | 4 ++-- 3 files changed, 4 insertions(+), 4 deletions(-) diff --git a/web/html/pkgreq.php b/web/html/pkgreq.php index 1e671d0..5f5aca4 100644 --- a/web/html/pkgreq.php +++ b/web/html/pkgreq.php @@ -13,7 +13,7 @@ if (isset($base_id)) { header('Location: /'); exit(); } - html_header(__("File Request")); + html_header(__("Delete/Orphan/Merge Request")); include('pkgreq_form.php'); } elseif (isset($pkgreq_id)) { if (!has_credential(CRED_PKGREQ_CLOSE)) { diff --git a/web/template/pkgbase_actions.php b/web/template/pkgbase_actions.php index 198f7ab..f70ddf9 100644 --- a/web/template/pkgbase_actions.php +++ b/web/template/pkgbase_actions.php @@ -32,7 +32,7 @@ <?php endif; ?> <li><span class="flagged"><?php if ($row["RequestCount"] > 0) { echo _n('%d pending request', '%d pending requests', $row["RequestCount"]); } ?></span></li> - <li><?= html_action_link($base_uri . 'request/', __('File Request')) ?></li> + <li><?= html_action_link($base_uri . 'request/', __('Delete/Orphan/Merge Request')) ?></li> <?php if (has_credential(CRED_PKGBASE_DELETE)): ?> <li><?= html_action_link($base_uri . 'delete/', __('Delete Package')) ?></li> diff --git a/web/template/pkgreq_form.php b/web/template/pkgreq_form.php index 623c247..9fd3cd5 100644 --- a/web/template/pkgreq_form.php +++ b/web/template/pkgreq_form.php @@ -1,5 +1,5 @@ <div class="box"> - <h2><?= __('File Request: %s', htmlspecialchars($pkgbase_name)) ?></h2> + <h2><?= __('Delete/Orphan/Merge Request: %s', htmlspecialchars($pkgbase_name)) ?></h2> <p> <?= __('Use this form to file a request against package base %s%s%s which includes the following packages:', '<strong>', htmlspecialchars($pkgbase_name), '</strong>'); ?> @@ -58,7 +58,7 @@ <textarea name="comments" id="id_comments" rows="5" cols="50"></textarea> </p> <p> - <input type="submit" class="button" name="do_FileRequest" value="<?= __("File Request") ?>" /> + <input type="submit" class="button" name="do_FileRequest" value="<?= __("Delete/Orphan/Merge Request") ?>" /> </p> </fieldset> </form> -- 2.6.2

5 11

[aur-dev] [PATCH] Change File Request To Submit a Request
by Mark Weiman 25 Nov '15

25 Nov '15

It looks like my last attempt to post this here failed, so I am trying again. Ignore this if the first one appears. Here is a patch that makes it Submit a Request. I believe I also got the strings in the translation files. Mark Weiman

3 2

Re: [aur-dev] [aur-general] [PRQ#4382] Deletion Request for b43-firmware
by Xavion 22 Nov '15

22 Nov '15

> > It looks this thread is on it's way to become spam... > Why? Is it because you don't like watching me romping to victory? > You got a couple of spam requests this month. I went back a couple of > months and coudn't find any requests related to you. > That indicates exponential growth. > If it really was a Problem, I guess the TUs would want to change that > themselves. Afterall they are the ones that would be really affected by > that problem. > Prior to my opening of this thread, I don't think the TUs realised why they were getting these "file requests". This is the main reason I opened it. > I don't think english speakers are the problem here. For all the others, > there is something called different translations. And whatever you do, if > you have some kind of input form, you'll get spam. > I am a native English speaker and even I had to do a double-take when I first saw "File Request". For the record, both natives and ESLs have reported having difficulty interpreting it. Since there's a new "Gospel of Mark" outside of the Christian Bible > that I have not written and is somehow about me (that I have not read > yet, please send me a copy), it would probably say to read > documentation and understand what you are using before you use attempt > to use it. If this person is being serious, he should be made aware of > the documentation available. See my earlier email about leading a horse to water. > Also what does the gun laws in the United States have to do with any > part of this? Usually the people who use those guns are familiar with > how to use one. They are both systemic flaws that need to be fixed. > This is a case where ignorance is causing an issue and I only think it > is a trivial problem (or lack of). IMHO, the fix (i.e. "Lodge a Request") is even more trivial than the problem itself. Following the argumentation that every written expression must be possibly > put in a correct english sentence, the whole table would have to be changed > to: _The_ Package Actions > View _the_ PKGBUILD / View _any_ Changes > Download _the_ snapshot > Search _for_this_package_in_the_ wiki > Flag _this_ package out-of-date > Vote for this package | Remove _my_ vote > Notify _me_ of new comments | Disable _any_ notifications > [ Manage _the_ Co-Maintainers ] > File _a_ Request > [ Disown _this_ package ] No, only the confusing ones should be changed. I thought this would've been pretty obvious. Some changements may not be strictly necessary to form a correct english > sentence, however the extracted part of that sentence would then lack the > respective syntactical context, e.g.: > "Dear Server, would you please disable notifications completely for me?" > Probably this sentence could be simplified while maintaining its > correctness, but stripping of some semantically necessary information, more > precisely which notifications are to be disabled. I don't see the point in all of that. We just need to change "File Request" to "Lodge a Request" (or similar). Though I am not a native speaker either, so any improvements or corrections > are welcome. I don't know why anyone would bother. It's all collateral that can be flushed away. ... You hardly need to have a degree in anything to know more about English > than a native speaker. In fact, native speakers tends to know less about > their language than non-native speakers. Assuming that you are, in fact, a > native speaker of English, you're proving yourself to be a prime example of > this. It's true that ESLs are forced to pay more attention to specifics than natives are. However, there are also times when ESLs miss the boat completely (through lack of know-how). > It is most definitely not. A double negative is a grammatical construct > where two forms of /grammatical/ negation is used in the same sentence. > Using a semantically negative word in the same sentence as a grammatically > negated one does not in any way or form count as a double negative. That's where you're wrong, Pal. The (English) Wikipedia article on the "double negative" lists exactly the example I provided for you ("I'm not feeling bad" versus "I'm feeling good"). So, back to my original point: "Unless you have at least a master's degree in linguistics, I wouldn't recommend thinking you know more about English than a native speaker does". > Hardly better. "File a Request" would be better though, but I don't feel > like rebuilding the translation catalogues, so I'll let Lukas change it. You seem to be about as clued up on this as Marky Mark is. As mentioned in an earlier email: "The word "file" has a double meaning with respect to the AUR, while "lodge" does not. This is because the AUR's central topic is "files", but it has nothing to do with "lodges".

1 0

Re: [aur-dev] [aur-general] [PRQ#4382] Deletion Request for b43-firmware
by Xavion 21 Nov '15

21 Nov '15

> > I can understand someone getting confused when they first click on it; and > any of the suggested name changes sound pretty logical. But, how does > someone click on that, then select deletion or merge to some other request > type, enter text and then click submit without realizing 'This probably > wasn't a download link' > They probably go there thinking it's the only way to "request a file". Once they're there, they have difficulty choosing between "deletion", "merge", and "orphan". They don't want to proceed, but they know of no other way to request that file :-). So, they randomly choose one of those options and send their unrelated request away. Essentially, they only proceed because they think the AUR doesn't have an inbuilt file requesting facility. I think people would be surprised at just how often this happens. Less than a fortnight ago, someone else did the same thing with another of my packages. That guy filed both deletion and orphan requests with the following comments: * Deletion: "Hi, Can anybody provide me this package ? Git not found. Use for bcm4312 . Thanx" * Orphan: "L can't find this package , can you upload it somebody ? Thanx" I liken this problem to that of gun laws in the US. If nothing is done about them, the mass shootings will continue. Similarly, if "File Request" isn't at least changed to "File a Request", the occasional Arch newbie will annoy/worry package maintainers with these bogus requests. Still, according to the Gospel of Mark, it doesn't need to be changed because *he* has no trouble interpreting the meaning of the link in its current form.

1 0

Re: [aur-dev] [aur-general] [PRQ#4382] Deletion Request for b43-firmware
by Xavion 21 Nov '15

21 Nov '15

> > As a native English speaker, I have never misinterpreted the meaning of > that link. There are plenty of words in the English language that have > multiple meanings and it doesn't prevent me or others from confusing > the intent. > Actually, that second sentence of yours seems pretty confusing to me. It looks like you've used a double negative, which gives the opposite meaning than the one you intended. I just hope you're never tasked with contributing to any of the Arch Wiki articles. > Lodge also has a noun and verb meaning. As a noun, it is a temporary > home and as a verb, it can also mean to live in a temporary home. By > your logic, this makes the word even worse of a choice. > I'm starting to wonder whether you might have Asperger syndrome. Just because the dictionary says something doesn't mean it fits all contexts. The word "file" has a double meaning with respect to the AUR, while "lodge" does not. This is because the AUR's central topic is "files", but it has nothing to do with "lodges". > Furthermore, I am a firm believer of reading documentation and the Arch > Wiki has a spot on the Arch User Repository page that explains how to > file requests [1] and how to grab the PKGBUILDs then how to install > them [2]. If the link is changed, I think Jiachen Yang has a better > solution. > My view is that leaving the AUR in a confusing state and forcing Archers to go to the Wiki for answers is a pretty stupid idea. Several people have already said that they were nearly tricked by "File Request". How many more will need to speak up before you'll finally admit that it's necessary to make a change?

1 0

Re: [aur-dev] [aur-general] [PRQ#4382] Deletion Request for b43-firmware
by Xavion 21 Nov '15

21 Nov '15

If we wish to continue using the word "file" in this context, "File Request" should be changed to "File a Request". This would prevent native English speakers from misinterpreting our use of "file" as a noun. Having said that, I still feel that "lodge" is a more appropriate term in this situation. The main reason, and I hope you're concentrating this time Mark, is that there's no double meaning (i.e. noun vs verb). For people who don't like "Lodge a Request", there's always "Make a Request". In fact, I imagine most similar websites would use the latter term in preference to either of the two mentioned alternatives. Regardless, clinging to the textbook and saying that nothing needs to be changed is too rigid and impractical. We're not doing a maths test here! This is all about practicality and cutting down on confusion.

1 0

Re: [aur-dev] [PRQ#4382] Deletion Request for b43-firmware
by Xavion 20 Nov '15

20 Nov '15

On 20 November 2015 at 13:22, <notify(a)aur.archlinux.org> wrote: > fckinshitag [1] filed a deletion request for b43-firmware [2]: > > i want this file > > [1] https://aur.archlinux.org/account/fckinshitag/ > [2] https://aur.archlinux.org/pkgbase/b43-firmware/ Okay, I see what's going on here. This guy has interpreted the "File Request" link (i.e. in the "Package Actions" section of each package 's AUR webpage ) as being a way of requesting a file (i.e. the package in this case). To prevent this kind of mixup in future, I would recommend that said link be renamed to "Lodge Request". I don't think anyone would be dumb enough to interpret this as a way of requesting lodging at the maintainer's residence.

2 2

[aur-dev] [PATCH] Describe how to omit "have" lines
by Lukas Fleischer 14 Nov '15

14 Nov '15

A new feature in Git allows for omitting "have" lines corresponding to refs outside the current Git namespace. Explain how to enable this feature in the INSTALL instructions and in the Git interface documentation. Signed-off-by: Lukas Fleischer <lfleischer(a)archlinux.org> --- INSTALL | 2 ++ doc/git-interface.txt | 13 +++++++++++++ 2 files changed, 15 insertions(+) diff --git a/INSTALL b/INSTALL index d68fa26..d956911 100644 --- a/INSTALL +++ b/INSTALL @@ -46,6 +46,8 @@ Setup on Arch Linux # mkdir /srv/http/aurweb/aur.git/ # cd /srv/http/aurweb/aur.git/ # git init --bare + # git config --local transfer.hideRefs '^refs/' + # git config --local transfer.hideRefs '!refs/' # ln -s ../../git-interface/git-update.py hooks/update # chown -R aur . diff --git a/doc/git-interface.txt b/doc/git-interface.txt index 9ded20f..4a24eef 100644 --- a/doc/git-interface.txt +++ b/doc/git-interface.txt @@ -80,3 +80,16 @@ request. An example configuration for nginx and fcgiwrap can be found in the INSTALL instructions in the top-level directory. + +Further Configuration +--------------------- + +When using Git namespaces, Git advertises refs outside the current namespace as +so-called "have" lines. This is normally used to reduce traffic but it has the +opposite effect in the case of aurweb: Many essentially useless lines are +transferred to the Git client during `git push` operations. + +In order to omit these advertisements, add the strings "^refs/" and "!refs/" to +the transfer.hideRefs configuration setting. Note that the order of these +patterns is important ("^refs/" must come first) and that Git 2.7 or newer is +required for them to work. -- 2.6.2

1 0