<div dir="ltr"><br><div class="gmail_extra"><br><div class="gmail_quote">On Fri, Mar 6, 2015 at 3:25 PM, Norbert Preining <span dir="ltr"><<a href="mailto:preining@logic.at" target="_blank">preining@logic.at</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">Hi Luigi,<br>
<br>
(reincluding tlbuild)<br>
<span><br>
<br>
> Compiled with zlib 1.2.8; using 1.2.8<br>
<br>
</span>Yes, that is the optimum, but not necessarily reality.</blockquote><div>for luatex and zlib</div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"> <br></blockquote><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><span>
> The problem is 1.2.11 vs 1.2.12 and I don't like it<br>
<br>
</span>No, the problem is<br>
1.2.5 versus 1.2.6<br>
etc. Both should be API compatible, and luatex should ignore the .5/.6<br>
part.<br>
<br></blockquote><div>no, for me the problem is</div><div>1.2.11 vs 1.2.12 because with 5 digits the check is ok while the number differs.</div><div>1.2.5 versus 1.2.6 is not the problem because the check fails </div><div> and this is ok for me.</div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">
The last digits in SO names are normally API compatible changes,<br>
if not this is a failure of maintainers (I don't say this doesn't<br>
happen, but is very rare)<br>
<span><br></span></blockquote><div>and this exactly what I have seen (with libpng )</div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex"><span>
> So i'm for keep 5 digits, maybe even to move to 8.<br>
<br>
</span>In fact it should not be digits, but:<br>
- parse into \d+\.\d+\.\d+<br>
- check the first two numbers, ignore the third<br></blockquote><div>assuming a sane numeration <br></div><div><a href="http://aa.bb.cc" target="_blank">aa.bb.cc</a> (i.e. 8 digits) for zlib is enough --- otherwise it's a sign that something wrong is going on </div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">
<br>
For peace of soul one could give a warning when the third differs,<br>
but don't stop.<br>
<br></blockquote><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left-width:1px;border-left-color:rgb(204,204,204);border-left-style:solid;padding-left:1ex">I am patching this in luatex since years in Debian, on abut 15 architectures,<br>
with several zlib upgrades.<br>
<br>
I don't udnerstand what you are scared? That upstream makes something<br>
stupid in the minor version?<br>
<br></blockquote><div>For the upcoming luatex 0.80 for TL I can put 4 instead of 5</div><div><br></div></div><div><div>/* make sure header and library version are consistent */</div><div> const char* version = zlibVersion();</div><div> if (strncmp(version, ZLIB_VERSION, 4))</div><div> {</div><div> lua_pushfstring(L, "zlib library version does not match - header: %s, library: %s", ZLIB_VERSION, version);</div><div> lua_error(L);</div><div> }</div></div><div class="gmail_extra"><br></div><div class="gmail_extra">so we check a'.'b'.'</div><div class="gmail_extra">assuming a and b a single char --- as also all the strings with the same prefix of length 4, of course.</div><div class="gmail_extra">A next luatex will have a better <major>.<minor> checker.</div><div class="gmail_extra"><br></div>-- <br><div>luigi<br></div>
</div></div>